PDA

View Full Version : BUILD LOG: It's begun....



Washout
25-09-2012, 08:53 PM
Hi all,

As the first materials for my CNC Router have started arriving, here are the results so far:

The second pic shows the re-purposed 4U Server case housing the PS806-5 PSU and a 240v AC to 12V DC PSU for the fans (two 80mm fans will extract the hot air out the back). The 3 x PM752 drivers are int eh drive bay behind the blue 120mm fan in the lower left corner.

My wiring routing will be largely as shown:

Red= AC in to both PSU's - will probably wire both of of the same 3 pin input socket, which will be housed in a plate to be fixed to where the old expansion card slot are.
Orange = the route I will take for the driver power from and to the PS806
Yellow = the stepper wires and any other wiring e.g. limit switches, estop etc - I'll be using 4 pin XLR connectors
Green = earths from everthing back to the yellow star point and then back out the case via the three pin connector

The 12V DC motor wires will be roughly wher eyou can see them now running at the bottom of the case to the 2nd PSU.

The main power switch will be in line with AC live and will be a two button start/stop similar to what I have on the front of my pillar drill.

The breakout board will likely be mounted to the left of the PS806, with its feeds to the drivers going in the same side as the yellow stepper feeds.

Oh and there is also a pic of where I have got to with the mechanical end of things (ignore the undersized work surface - that is being replaced with something large enough and additional supports added). The thin bed ali profile was a fail purchase on my part and will be replaced with the 30mm profile you maybe able to see sitting on top of it and the thin profile is going to become wall mounting for my guitars and maybe tools etc, so it won't be wasted.

Got some more electrical stuff arriving soon to complete the enclosure and I have my ballscrews, bearings, supported rails etc, together with a 2.2kw watercooled spindle hopefully arriving from Chai soon.

3NM Nema 23's are off camera until I have worked through the rail and ballscrew mounting.

Cheers


Chris

PS. I am thinking that I should the VFD's earth back to the star point shown in the case in the pic and not back to its own 3 pin plug's earth (both will be on the same 32A spur, along with the PC that's going to run Mach3 so the whole rig is on its own RCD)?

Fivetide
25-09-2012, 08:59 PM
I have to say that the first pic is a piece of art on its own :)

Washout
25-09-2012, 09:08 PM
Thanks Fivetide - wish I had bought a self-reversing tapping chuck for my pillar drill - hand tapping all that profile was tiresome..

Fivetide
25-09-2012, 09:14 PM
Washout whats your buget for the complete unit ? Or should I say what are you aiming for as a buget lol ? I'm not going to attemt one for along time, although I have started to collect parts, i have the Mach3 controller.

Washout
25-09-2012, 09:29 PM
I was aiming for about 2 to 2.5K to start with, which I should just about squeek within once I settle on which CAM software I'm going to buy.

I'll then probably blow that budget when the itch for a 4th axis gets too much and I intend to replace some of the extrusion with plate, but I'll be happy just to get cutting something for now and do the upgrades and addons, when time and funds allow. The PC side of things isn't an issue, as you can probably tell by looking further up the rack in the enclosure pic, so I'm lucky in that respect....

Fivetide
25-09-2012, 11:15 PM
I think your rig looks awesome, and same here I want 4 ways but to be honest I need to get the one I have working first lol but I will follow the tread because I think the next one will be alot like yours.

Washout
02-10-2012, 11:24 AM
Quick update (no pics yet) - two parcels arrived from China today and I now have my rails, ballscrews, VFD and 2.2Kw spindle plus ancilliaries :-) I can, along with others recommend Chai for these components - very easy to deal with and will accomodate most requests (I needed FK bearing blocks rather than BK for example).

The rails are a little over sized, but that's better than undersized, so its out with the angle grinder to cut the stainless (hack saw is more eroding than cutting ;) ).

I'm going to need to make a few adapter plates for the ballscrew bearing housings and rail bearing blocks, so I'll need to track down a cheap supply of aluminium plate.

I'll get some pics when the bits start to go on the frame in the pic above.

Jonathan
02-10-2012, 12:27 PM
I'll need to track down a cheap supply of aluminium plate.

Aluminium Supplier | Aluminium Stock | Aluminium Warehouse (http://www.aluminiumwarehouse.co.uk)

Or if you just need a small bit let me know...

Washout
02-10-2012, 02:24 PM
Thanks Jonathan - I'll grab a decent sized sheet from them, which I can use for adapter plates and also the Z axis - I'm thinking 15mm thick should be sufficient (I can also make the spindle mounts from this).

Cheers


Chris

Washout
08-10-2012, 01:04 PM
Hi All,

Whilst waiting for my bed extrusions and some other components to arrive, I have been busy festooning the rack mount case with wiring (see pic below).

Stuff wired up are:
- the small PSU, which is powering the case fans (tested this and all is running fine).
- The 240v AC input, which goes to a front panel mounted On/Off switch (same kind you see on pillar drills etc)
- 240v AC from the On/Off switch to the two PSU's
- The CY cables for X, Y and Z from the drivers to the XLR sockets (shielding grounded to the star point) - I'm wiring my steppers in parallel, so am following the Zapp wiring diagram here: http://www.zappautomation.co.uk/downloads/PDF%20Datasheets%20and%20manuals/Breakout%20Boards/ZP5AINT%20wiring%20Model%20%281%29.pdf
- The high voltage DC side of the drivers to the transformer (running it parallel off of one PSU output (hope that is right)

Still to do are:
- Low voltage DC side from the breakout board to the drivers (waiting on some lower guage wire for this job)
- Limit switch, spindle, e-stop wiring from the breakout board to the spare XLR connectors
- Find out if the earth lead I have on the main PSU is really where it needs to be, as there doesn't seem to be anywhere else to attach it to, or discard it if its not needed?
- Perhaps get some QA on my wiring, because although I am following the diagrams etc to the letter, a fresh set of eyes is always useful before throwing the power switch for the first time....

The mechanical side of things is on hold for the moment pending delivery of some extrusion channel nuts to start mounting the rails and some ali plate for mounting adapters.

More to come........

irving2008
08-10-2012, 02:40 PM
Are you working to a circuit diagram or making it up as you go?

Be careful with that central bolt on the transformer. Have a look at the Printed Circuit board underneath. Does the bolt connect to a copper plane at all? If it does you should attach your earth to that side not the top or better, dont attach to that bolt at all, but use metal mounting posts for the PCB to earth it to the chassis. If it doesn't, i.e. its insulated from the circuit board, then you don't need an earth there at all. The way you have it wired could be a problem if the bolt should touch the chassis, or if its connected to something on the circuit board that could end up connecting to the chassis, because then what you will have created is a shorted turn on the transformer which will at best give some loss of power and at worst could cause overheating and melt things...

7087

Washout
08-10-2012, 03:24 PM
Thanks Irving

I'm working off of the circuit diagram in the pdf in my post, but that only shows the driver circuitry essentially and not the PSU specifically.

The PSU is this one Zapp Automation Ltd - Stepper motors & Stepper Motor Drivers - Servo motors and Servo Motor Drivers - Ballscrews & linear motion products. : PS806-5 Linear power supply [PS806-5] - 102.00 (http://www.zappautomation.co.uk/ps8065-linear-power-supply-p-518.html?cPath=10)

From what I can see the toroidal coil is bolted via that central bolt to a mainboard, which doesn't look like it has any copper substrate to it. The daughter board, which has the capacitors etc (rectifier?) is a standard PCB, which is wired to the coil(s) and that is also mounted on the main board by brass spacers. I currently have nylon spacers from the main board to the chassis, but could certainly replace those with brass also, if that's how its supposed to be earthed or maybe as you say it shouldn't have an earth connection at all (which I find a little difficult to believe, but I'm no expert).

I'll try and PM Gary @ Zapp and see if he can help (the data sheet for the PSU is next to useless, but that's normal for stuff like this I find).

Thanks for the sanity check in any case.


Chris

PS. I have removed the earth lead for now until I know how it should be wired.

Washout
09-10-2012, 01:47 AM
Quick update:

Gary came back to me and the PSU doesn't have an earth, so I have removed the wire to the main bolt.

Just need some more bits to arrive now.....

Washout
09-10-2012, 10:23 AM
Quick question for any of the more electrically minded amongst you all:

Looking at Irving2008's diagram showing an earth loop, when I had the earth wire from the transformer's bolt to my star point. Do I not also have an earth loop with the CY cable shielding going to my star point and that shielding also being grounded where the XLR sockets come into the chassis as the outside of the sockets is the earth conductor?

If that's the case, should I isolate the XLR sockets (or the plate they are mounted into) from the chassis? - I could do this with rubber O-Rings.

Apologies for all the electronics questions, as whilst I have some electronics experience (555 timer circuits and various PC input/simulation projects), those are generally a two wire solution i.e. +ve and 0 or -ve and little risk of zapping myself with dangerous currents ;-)

irving2008
09-10-2012, 10:35 AM
Generally you should only ground shielding at one point. If there is a good ground for the shield at the XLR then good practice says you shouldnt ground it at the star point either.

However the issue is minor here as there is little likelyhood of induced current in that short a run. The real issue is where you run a long shielded cable alongside, say, an unshielded power cable. Grounding both ends of the shield allows the magnetic field from the power cable to induce a current in the shield which can cause strange effects. In one site I worked on this problem caused the 'ground' of the remote device a few metres away to be 7v AC relative to the controller. With no opto isolation it wasn't surprising that the logic circuits were having some difficulty. Disconnecting the shield reduced the offset to <0.1V and all was well...

Washout
09-10-2012, 02:26 PM
Great - thanks for that, I'll leave that side as it is then.

My bed extrusion has turned up along with the M6 connectors I needed this morning whilst I was on one of my rare visits to the office, so I can also crack on with some of the mechanics as well (will make a welcome change from wiring) :-)

Washout
10-10-2012, 05:34 PM
Just a quick pic showing rails going on:

Washout
11-10-2012, 04:52 PM
Another pic showing the beginnings of the Z axis.

I must be doing at least something right here as it moves nice and freely on its Y axis rails (as my crushed finger will attest to when I tilted it to put it back on the bed ;-) ).

Also I guess I should relate some areas of "interest" (read frustration) getting this far:

Firstly I had a problem with the profile from metallin having a slight ridge along the centre "flat area" on the 60x60 profile. This combined with Chai using hex machine head screws meant the rails would not sit flat to the profile bars but instead rocked side to side. This was easily fixed with replacment button head screws, which are lower profile and therefore the ridge in the profile can sit inside the rail supports and they can lie flat. Mine were very tight, but didn't have the stripped threads I see in another build log....probably air tools or something used to screw them in with too much torque.

Secondly screwing the extrusion together can be a pain, but I found that using a couple of turns on each screw using a similar pattern familiar to anyone who has torqued a cylinder head down seems to work well and almost pulls the extrusion into alignment itself (provided its cut at a nice right angle, which I'm using a TCT bosch blade in a chop saw for).

Thirdly I highly recommend a spiral flute tap (I think that's the correct name) for tapping the aluminium. The alloy for the extrusions seems to be stickier than most and was clogging up my normal taps after only a few turns. Even backing out to break the "chip" after every 1/4 or 1/2 turn didn't seem to help on my normal taps. IIRC I got the spiral flute tap from Arc Euro trade (is that John S's company?) - highly recommended and I'll be ordering more goodies from there in the future.

Of course there is much more to do, but it is nice to have a moving part operating correctly on the machine.

My ali plate also arrived today from Aluminium Warehouse (thanks for the link Jonathan), so I can get on with making the mounting plates to fix the X Axis bearing blocks to the gantry.

I am also having to do an on the fly redesign of the z axis, as the 120mm extrusion is not going to be wide enough for the two supported rails and ballscrew mount etc. Luckily I ordered a 1000x500x15mm ali plate so I have at least got plenty of material to make a wide version from.

More as I progress......

Cheers


Chris

Washout
24-10-2012, 04:30 PM
Hi All,

Build has been on a bit of a pause since my last post, partially due to waiting for a new circular saw (read more powerful and able to take my Bosch 210mm metal cutting blade), which I also needed for some other wood projects and also due to finally selling my house, which means I will (fingers crossed) be moving to premises with a nice 30x15 foot workshop :-)

Anyway here's a quick couple of pics of the wiring work in the control box (re-purposed rack mount servre case) progressing and its now pretty much complete, with the remaining XLR sockets wired to the BoB, which will be used for limit switches and spindle relay (need to find out how the latter is wired i.e. at the VFD directly if it has one built in or whether I need to grab a seperate unit). The second pic shows the mains on/off switch, which was a nice fit into a 3.5inch drive bay blanking plate.

Stuff I need to do in the control box are decide if I need the short CY cables shielding earthed (probably not) and work out where the parallel lead will go.

Which also leads me to a few questions:

1. What kind of parallel lead will I need, as there are data transfer and printer types and IIRC 1 is 25 pin straight through and the other has a pin 1/2 swap like a nul modem (been a while since I've used a parallel lead on anything).

2. Also has anyone run a 3metre parallel lead and is this likely to be an issue?

3. Finally has anyone used this unit? Zapp Automation Ltd - Stepper motors & Stepper Motor Drivers - Servo motors and Servo Motor Drivers - Ballscrews & linear motion products. : PLCM-E3 CNC Controller (For use with Mach3) [PLCM-E3] - 150.00 (http://www.zappautomation.co.uk/plcme3-cnc-controller-for-use-with-mach3-p-776.html?cPath=15)

From reading up on it, it doesn't seem to do more than provide conversion from ethernet/USB to an existing Parallel port BoB (or BoBs), but using my structured cabling to connect the PC to the controller box is appealing.

Tomorrow I will likely attempt to chop my ali plate for the Z axis and the spacers to mount the gantry to the X axis bearing units.....will report back after that...

Later


Chris

Washout
24-10-2012, 04:37 PM
Ooops nearly forgot.

As I tend to cover all soldered connections with heatshrink I have kept a spreadsheet of conections and wire colours etc, so that if I am troubleshooting or adding new wiring I can work out what goes where and its function.

Might be of use to somebody, so attached.

Cheers


Chris

m.marino
24-10-2012, 04:46 PM
You might want to look at www.ascmetals.com (http://www.ascmetals.com) and remember that Aluminium can have both a crown and warp to the sheet. This means that unless you are buying engineering grade that you will most likely have to face the metal or have it faced to bring the surfaces into true parallel. Depending on how large the section of plate you are using this might have no effect, on the other hand the wider and long a plate is that has not been brought to tolerances the more likely the piece will have warp or crowning and that means surfacing/facing needed to be done to bring it into tolerances.

Just a few points that you might want to consider when looking at buying as the cost in material does have a bit of a reason behind it.

Michael

Washout
24-10-2012, 04:58 PM
Thanks Michael,

Link added to my list of suppliers.

The ali plate I have seems to be pretty accurate over the size I'm cutting it to, but I'm expecting some shimming being necessary, as my cutting of extrusion is likely not to the same tolerance. Hopefully I can get the machine true to itself to a certain degree and then work on replacement parts machined to better tolerances than the current tools I have been making it with, or of course issue some RFQs here for bits...

Cheers


Chris

Washout
26-10-2012, 12:20 AM
A quick update, as I got some time to work on the machine tonight, but no pics (will post some tomorrow):

I have my first ballscrew mounted, as I figured before I attach the Z-mounting plate the Y Axis needed to be finished. Used a Bosch 17mm bi-metal hole cutter to bore the holes in the gantry sides for the ballscrew and I find these tend to cut ~1mm over size (possibly due to my bench drill being a little wobbly). This is fine however as I need a little clearance to allow for alignment adjustment and have ~1mm rather than the intended 0.5mm.

The floating end is attached using an FK bearing block and circlip and the fixed is also a larger FK type which goes in the stepper motor mount and that is bolted to the gantry side also.

I have come across one of those annoying gotchas where the supplied ballnut mount protrudes ~1mm too far outside the gantry, so I need to work out whether I can "pocket" the ali plate for the z axis by 1mm using the tools I have. I'm thinking of very gently/slowly milling the pocket using my bench drill and an 8mm TC End Mill with the z axis plate clamped to my small cross slide table (normally only used for drilling holes in a straight line) - not sure I should mention a foolish move like this here, but it's a problem that needs solving (I have also toyed with using a low speed on my hand wood router, but I don't trust myself with that method.....).

Hopefully tomorrow I will also get try out the new circular saw on the ali plate (didn't finish work till late and the neighbours would probably not appreciate the tortured machining noises;-) ).

I must also try to resist connecting up the Y Axis just to get something moving under its own steam.....

Washout
26-11-2012, 06:22 PM
Afternoon gents (and ladies),

Not posted for a little while but I have not been idle ;-)

Here are a few pics of latest progress and despite a few "rough" bits (z axis plates and spindle mounts), I am finally approaching the end of the mechanical side of things.

I do a have some deflection in the gantry, which is due to my crude mounting solution, but have an RFQ out to help resolve this and will soon order a second ballscrew plus pulley/belt to make sure both sides get equal x axis driving force.

In the meantime, I am now working on the cable runs for the steppers and then the water cooling system and VFD/spindle wiring.

Good news is I have the rest of the year off from work, so I'm hoping I can get at least some light material cutting underway before Christmas.

Lastly before finishing I had better relate some frustrations:

1. I had underestimated the weight of the gantry once all the "stuff" is on it and care is needed when assembling the z axis, as its easy to drop the spindle/mounts etc and if your fingers are in there pain can result - propping things up when securing bolts etc is a must

2. Constructing the z axis is a pain the first time you do it - everything had to go on in order to avoid not being able to get tools onto bolts/nuts etc.

3. Like a berk I assembled the gantry reversed as you can see in one pic, but luckily I needed to almost completley disassemble the gantry to get the y axis ballscrew mounting block in, so put it back the right way.

4. to align the x axis ballscrew I torqued down the fixed end (stepper motor end) then ran the gantry by hand to the far end and tied down the ballscrew and floating end bearing. Hopefully that is the correct method, but it turns and moves easily by hand (I'm assuming turning a stepper by hand is OK).

5. I had a job getting the rails into the bearing blocks on the z axis and actually stripped a bunch of ball bearings out of one of the block, so had to get some replacements, but they are relatively inexpensive and I've sorted out a method which involves only bolting one end of the rails to the front plate, which gives some play to allow for alignment. I then put the remaining bolts onto the rails once they are located in their bearing blocks - not a hard job just needs doing with the front plate in the right place, so I can get access from the rear.

Oh I should (fingers crossed) be moving house soon and I'm really looking forward to the new workshop :-)

Hopefully my next update will not be too far away.

Later



Chris

Washout
30-11-2012, 01:26 AM
Quick update and some questions someone maybe able to help with.

I have thrown an e-chain on for the x axis movement and wired up the 3 steppers with cables going back to my control box (shield connected at the control box end to my earth point and not at the machine end).

I figured I would get the stepper motors working first before moving to the spindle and VFD and to prepare for the first control box power up I have done the following:

- Manually moved the spindle to the centre of x, y and z travel so that if for some reason the machine decides to shoot to the extremes of my travel I have time to hit the estop.
- Loaded Mach3 and connected the PC I'm running that on to the ZP5A-INT controller board/BoB but not in the control box so I can set up the software according to this pdf manual without risk of the machine moving in an undesired fashion: http://www.zappautomation.co.uk/downloads/PDF%20Datasheets%20and%20manuals/Breakout%20Boards/ZP5A-INT%20axis%20interface%20board%20manual.pdf

After running through that manual for setup I have loaded the Roadrunner.tap and run the GCode - the LEDs for the axis on the board are flickering nicely in sync with the movement on the Mach3 3D display, which is encouraging.

All good so far, but here comes the questions:

Having wired my steppers and drivers in bi-polar parallel mode (wiring diagram: http://www.zappautomation.co.uk/downloads/PDF%20Datasheets%20and%20manuals/Breakout%20Boards/ZP5AINT%20wiring%20Model%20%281%29.pdf) I am now stuck with what dip switch settings to use on the drivers (PM752's - http://www.zappautomation.co.uk/downloads/PDF%20Datasheets%20and%20manuals/Stepper%20Drivers/PM752.pdf) with the following Steppers: SY60STH88-3008B (Data sheet: http://www.zappautomation.co.uk/downloads/PDF%20Datasheets%20and%20manuals/Stepper%20Motors/SY60STH86-3008BF.pdf)

My first guess is the following in order from SW1 to 8: OFF, ON, OFF, OFF, ON, ON, ON, ON which should translate to Peak Current 4.09A, RMS Current 3.32A, Standstill=Half Current, Microstep=4, Steps/rev=400 - I have no idea if this is correct although one of the manuals mentions setting the Amp rating to the closest for the motor and 4.09 seems to fit that bill.

Any help would be gratefully received before I wire the BoB back in and throw the switch just to see what happens :-)

Oh if it helps I have 1605 ballscrews on all axis.

Cheers


Chris

Jonathan
30-11-2012, 04:29 AM
My first guess is the following in order from SW1 to 8: OFF, ON, OFF, OFF, ON, ON, ON, ON which should translate to Peak Current 4.09A, RMS Current 3.32A, Standstill=Half Current, Microstep=4, Steps/rev=400 - I have no idea if this is correct although one of the manuals mentions setting the Amp rating to the closest for the motor and 4.09 seems to fit that bill.

That is correct. Also depending on the feedrate you get and the kernel frequency your computer will support you may be able to use 800 or 1600 step/rev which will help the machine run more smoothly.

Washout
30-11-2012, 07:22 PM
Great Jonathan, I will throw the power switch tonight and all being well should get some movement out of the machine.......consider me excited now :-)

Washout
30-11-2012, 09:20 PM
Quick update - powered up the control box and nothing has sparked or smoked, which is nice.

The steppers seem to have power to then as they are making a slight humming noise and the ballscrews are now solid (i.e. I can't hand turn them). However, I can't seem to get any movement from the jog control or up and down cursor keys in the motor tuning dialogue. Also I have green LEDs lit on the drivers, which I also consider a good sign.

I've probably missed something silly in Mach3 so will watch some more tutorials and see if anything jumps out at me.

Washout
01-12-2012, 12:49 AM
Hmmm I think I am stuck now:

- I have checked my wiring several times (including continuity testing to make sure I haven't crossed any on the signal wires in error) and everything seems to check out.
- My drivers have their green lights lit.
- The breakout board seems to be responding on the axis correctly i.e. I use a jog key in Mach 3 or run the Roadrunner.tap GCode and the activity lights on the axis outputs seems to be lit when they should.
- I've tried the enables 1 to 3 ticked in the mach3 Outputs tab on Pin No.1 using Active Low ticked and unticked and using Pin 1 as suggested in the BoB instructions (when I have these checked the diagnostic screen in Mach3 is showing 3 red "LEDs" for each
- Tried 25000 and 35000 Kernel speeds (35000 is the setting suggested in the instructions)
- Checked to make sure the parallel Port's address is correct
- Checked and re-entered the Motor profiles for each axis as per the BoB manual, although I notice my G's value doesn't match theirs

Anyway all of that and each time I switch the machine on the steppers lock in place with a kind of high pitch hum and refuse to move, as if there is constant power going to them but no direction.

If anyone has any suggestions I'd be grateful , as I think I'm either missing something very fundamental, got some defective parts or some arcane setting in Mach3 is eluding me and I'm running out of things to try.

Cheers


Chris

Washout
01-12-2012, 10:45 AM
Hi Guys,

Having slept on the problem of my stuck machine can someone with wiring knowledge check a theory I have:

I've been wiring up according to this diagram on the Zapp website: http://www.zappautomation.co.uk/downloads/PDF%20Datasheets%20and%20manuals/Breakout%20Boards/ZP5AINT%20wiring%20Model%20%281%29.pdf

Looking at it and taking the x axis as an example, the wiring diagram shows all the +ve connectors for Pulse, Direction and Enable wired together and going back to XN (which I am assuming is X -ve?). Shouldn't this be the other way round i.e. the Pul-, Dir- and Ena- should be wired to XN?

If so I can re-wire and try, but some wiser head than mine with a view would be appreciated.

Thanks in advance.


Chris

mocha
01-12-2012, 11:56 AM
try this one... :-)
http://www.mycncuk.com/forums/drivers-controllers/5030-m542-wiring.html

Washout
01-12-2012, 12:20 PM
Aha then I think my theory is right - I have the +ve's tied to XN on the BoB, not the -ve's.

Going to rewire now and hopefully having them the wrong way round the first time hasn't damaged the drivers. A classic Meccano moment if you've ever tried to build that stuff as a kid from the instructions (they deliberately put errors in their diagrams to make you think for yourself ;) ).

I'll report back in an hour or so.

Cheers


Chris

Washout
01-12-2012, 12:39 PM
Well it didn't take an hour - the diagram is wrong and the polarities did need swicthing.

Happy days - my machine is moving.

I do have motor stalls though when the steppers have accelerated to a certain speed, so I assume I need some tuning.

Not felt quite like this since I first solo'ed a light aircraft :-)

More as I progress and thanks for the help chaps......

Washout
01-12-2012, 12:54 PM
OK, next challange:

I'm now trying to tune my motors and am thinking the following:

My steppers are 1.8degs and I am assuming this is 1.8degs per step, which would mean 200 steps per revolution (360/1.8=200)

My ballscrews are 1605, so again an assumption that 1 revolution of the ballscrew is going to translate to 5mm linear travel

So I am thinking that to get the number of steps per mm for Mach3 I am going to have to set the motors to 200/5=40 steps per mm?

Not sure what the velocity should be set to or acceleration in that case, but I guess I can play with the sliders and get the motors running "smoothly" and not stalling.

Again any advise welcome.

Cheers


Chris

Washout
01-12-2012, 01:13 PM
Oops forgot about the driver dip switch setting of 400 steps per rev, so that will be double what the above calc comes to per mm

Washout
02-12-2012, 05:53 PM
Quick update and yet another question:

Spent most of yesterday playing with motor tuning, setting soft limits etc and just having fun jogging my machine about the place :-)

Today however I have put in a seperate 32Amp spur from the office workshop consumer unit to a double socket for the VFD/machine and also wired up the VFD. So far so good, as whilst I type my VFD is whirring away with a flashing F00.00 LED display, which I think means its passed its self-diagnostics and is now waiting for me to set it up.

Having burrowed into a few threads I have picked up these settings from a woodworking forum, which seem to apply to a water cooled spindle setup:

Change PD001 Source of Run Commands to '1'
Change PD005 Max operating frequency to 400hz
Change PD004 base frequency to 400hz
Change PD003 Main frequency to 400hz
Change PD006 Intermediate frequency to 2.5
Change PD008 max voltage to 220v
Change PD009 Intermediate voltage to 15v
Change PD010 Min voltage to 8
Change PD011 frequency lower limit to >100hz (120hz to start)
Change PD142 7 Amps
Change PD143 2
Change PD144 3000

Thread here: Tips for Newbie Huanyang VFD users - Page 2 (http://www.woodworkforums.com/f170/tips-newbie-huanyang-vfd-users-96380/index2.html)

All fo those look sensible?

However, I am pretty sure the 2.2Kw spindle uses up to 11Amps and I have no idea what the PD144 3000 setting is.

Can someone post their 2.2Kw spindle and VFD settings for me or confirm those above are correct? I dearly like to connect the spindle up and get some rotation along with axis movement :-)

Thanks again in advance.


Chris

PS. Will try and get some video up soon, as pics don't do an active machine justice:-)

AdCNC
02-12-2012, 07:20 PM
Chris, did you get your brackets sorted how we descused?

Washout
02-12-2012, 09:34 PM
Ah yes, thanks - used some 15mm plate again, but this time wide enough to reach the gantry sides - 3 tapped holes and they are now tied into the gantry sides as well as the bottom of the gantry.

Of course I have yet to cut anything to see if they work as intended, but I'm getting closer. Just let the spindle run up to 12,000rpm using the settings above (apart from 11Amps which the VFD was already set for) by setting PD003 to 200.

The spindle is very quiet (certainly quiet compared to hand held routers I have).

Chris

Washout
03-12-2012, 03:25 PM
Hi Gents,

Stuck again I'm afraid:

For some reason I am getting an e-stop condition (whether the e-stop is connected to the BoB or not) when I turn my spindle on. Its consistent as well i.e. happens every time.

Can someone confirm my earthing setup:

I have VFD earthed to mains via its own earth connector and also have the shielding for the VFD to Spindle cable earthed to that point (only at that end i.e. shield is not connected to the spindle or machine, only the VFD earth point).

All other control cable shields are earthed back to the star point in my control box (again no connections for the shield on the machine).

I have good mains earth (tested with a socket tester).

Only thing I am not sure about is the spindle's earth (pin 4) is not connected to anything and I haven't connected a lead to it back to anywhere. Also I have earthed the machine via a seperate cable back to the control box star point and have good continuity all the way from the spindle casing down to the frame bed (my machine is nearly all metal).

Oh and I have the Z axis CY cable on the other side of the Z axis plate from the VFD/Spindle cable (about as much separation as I can get.

Any ideas?

Jonathan
03-12-2012, 03:45 PM
If your spindle case isn't earthed strange things can happen. It's easy to unscrew the endcap and solder an earth wire in on pin 4, or you could just connect the earth wire to the spindle mount assuming it's metal.

The minimum frequency setting you found, PD11=100Hz, is meant for air cooled motors which burn out when run at too low a speed, due to insufficient airflow. For the water cooled spindles this can be set much lower, however don't expect to get much torque below about 50hz.

PD144=3000 is just a scale factor which makes the rpm display show the correct numbers. You can also set PD14 and PD15 to increase the acceleration/deceleration. They're the ramp-up/down times in seconds, so just decrease them until it's fast enough without faulting.

Washout
03-12-2012, 04:24 PM
Thanks for the speedy reply Jonathan,

Should I also disconnect the earth lead from the bed to my star point? I figure if I earth the spindle and the bed to different points I may get a loop (my mounts and pretty much everything else are metal, although the bearing blocks may not be so good at continuity). This earthing thing is a little confusing and I'm allergic to large currents ;-)

Washout
03-12-2012, 05:54 PM
Quick update and firstly thanks to all those on here who have helped me out, both on PM and in my Build Log.

I have just successfully completed a couple of sprial drilling operations into some ali extrusion and ye gods it feels good. Holes look to be nice and circular and in the right place and the inside finish on them is excellent. I used GWizard to give me some feeds and speeds and so far the machine is withstanding the forces, so again thanks to those who helped me turn my original design into something stronger.

Time for some celebrationary dinner and then contemplate which part I should have a go at making, out of the pile of CAD I have been building up over the last few months.

BTW - I may have missed it but is there a way of jogging in Mach3 to a specified distance on an axis other than writing GCode for it?

Cheers


Chris

Jonathan
04-12-2012, 12:51 AM
Ideally you shouldn't rely on the linear bearings to provide the current path to earth, so earth both.

You can jog a specified distance by using the MDI tab and typing the Gcode commant, which is easy enough since you can see the co-ordinates the machine is at, so just type G0 Xsomething Ysomething etc to move to where you want. You can also set jogging to moving set distances instead of continuously, so for instance every time a jog key is pressed the axis moves 1mm (or whatever you set).

Washout
04-12-2012, 01:56 AM
Thanks Jonathan,

Reason for the question is I am zeroing by jogging up to the work piece on each side and setting that axis to zero and wanted to jog by 2.5mm afterwards in x/y to take account of my 5mm end mill, but I guess I should be using the tool offset for this.

I have also seen people using lasers (collet mounted and offset mounted) and web cams (with cross hairs) for zeroing and of course the more "pro" ruby tipped probes - any idea who might sell the probes and/or lasers? I have the aspiration to try probe scanning at some point as well, so the latter might suit me more than the laser technique.

Gotta love being on the learning curve.....its half the fun.


Chris

JAZZCNC
04-12-2012, 02:36 AM
Do your self a favor and go buy Gerry's screen set here: The CNC Woodworker - Mach3 2010 (http://home.comcast.net/~cncwoodworker/2010.html)
Then make a simple touch probe plate from single sided PCB. Connect to a spare input and set the probe Input inside Mach to look at it.
Then enter the thickness of the PCB into the screen set and it will take care of everything else.
It will save you hours of time just with setting the Z height alone never mind all the other built in probe routines which will find corners, edges, circle centres, parallel edges best 15 you'll ever spend.!!

Oh and it has MDI on every screen which is priceless if like me you use it a lot.!!

Washout
04-12-2012, 10:41 AM
Perfect - thanks for that - more like Excel/Word rather than Lotus123/Wordperfect :-)

Chris

Jonathan
04-12-2012, 01:14 PM
Reason for the question is I am zeroing by jogging up to the work piece on each side and setting that axis to zero and wanted to jog by 2.5mm afterwards in x/y to take account of my 5mm end mill,

After zeroing you can go on to MDI, raise the tool up obviously, type G0 X2.5 Y2.5 and zero the axes again. Assuming you zeroed both in the positive direction, the machine is now zeroed. You will probably often want to add a little bit, (G0 X3 Y3) to make sure any cutout pass does make a cut all the way round and not just miss the edge.

JAZZCNC
04-12-2012, 06:12 PM
After zeroing you can go on to MDI, raise the tool up obviously, type G0 X2.5 Y2.5 and zero the axes again. Assuming you zeroed both in the positive direction, the machine is now zeroed.

Even easier way than that if your using the default 1024 screen set. First set the Safe Z in Config (which you really should do for safety reasons regardless) and make sure it works in Work coordinates.
Then all you have to do is type the offset (-2.5 in your case) direct into the X,Y DRO by clicking on each DRO (Must press Enter to apply it) then simply clicking GOTO ZERO button just below DRO's.
You will still need to have the Z axis Zero set equal to the top or above material. The Z axis will retract first to Machine Zero coords because of SAFE Z then move the X,Y axis then return Z axis to Zero work coords.

Washout
05-12-2012, 08:41 PM
Thanks for that,

The new interface is much cleaner and I especially like the new jog tab rather than the old one hiding most of the DROs.

I've now had the machine cut the holes to attach its bed extrusions to the frame and it did that really well apart from on mishap where for some reason the mchine decided it was going to rapid and ramp to the first drilling spot. It made a nice gouge in the extrusion before I realised what was happening and hit my e-stop (which worked perfectly). I didn't snap the tool which is a blessing and I guess the 5mm carbide end mill went into the material by about 4mm before I caught it. A quick reset and zero and all was well , so not quite sure what happened as it had just cut the countbores on a seperate G-Code job before hand.

Anyway after resetting and cutting a bunch of holes to finish the bed without mishap, I have just "air cut" my first 3D part and it seemed to work OK. I do have another question though:

I'm using GWizard to help with feeds and speeds and for the drilling operations and counterbores the 18,000rpm, 750mm/min feed and 515 plunge rate GWizard suggested seemed to work really well. For the 3D roughing operation I'm about to try I'm using an 8mm Carbide End mill (3flute) and it is suggesting 13,000rpm, 2,000 mm/min and 750 plunge - does this sound right guys, as it seems to be very aggressive?

If there is ever a UK CNC meet up, I am going to owe you guys a small brewery.....

Cheers


Chris

JAZZCNC
05-12-2012, 09:18 PM
Anyway after resetting and cutting a bunch of holes to finish the bed without mishap, I have just "air cut" my first 3D part and it seemed to work OK. I do have another question though:

I'm using GWizard to help with feeds and speeds and for the drilling operations and counterbores the 18,000rpm, 750mm/min feed and 515 plunge rate GWizard suggested seemed to work really well. For the 3D roughing operation I'm about to try I'm using an 8mm Carbide End mill (3flute) and it is suggesting 13,000rpm, 2,000 mm/min and 750 plunge - does this sound right guys, as it seems to be very aggressive?

Can't answer really without knowing material, DOC.

Warning thou regards G-wizzard.? The calcs are based on using a milling machine which is much stronger than what you'll probably be using so I'd always knock a bit off to account for this.! How much mostly depends on your machine but 25% would be a good safe start.

Washout
05-12-2012, 10:03 PM
Ah yes DOC - I tried a 1mm and 3mm DOC and its 6082 T651 Ali

JAZZCNC
05-12-2012, 11:54 PM
Ah yes DOC - I tried a 1mm and 3mm DOC and its 6082 T651 Ali

You'll need a very strong machine with powerful spindle to cut at 3mm DOC and those feed rates Gwiz suggested with 8mm cutter.

Try 1mm @ 800mm/min 10,000rpm and take it from there.!!. . . . I cut @2.5-3mm Doc @800mm/min 9-10,000 rpm 8mm 3 flute ripper using blown air.
Each machine is different so you'll need to play with settings to find the sweet spot for yours. If you push too hard to start with you'll just cause premature wear and possible damage machine so start off easy with DOC and Feed rate then use the Feed override to change feed rate on the fly while cutting, also play with spindle rpm has well.
By doing this you'll find the sweet spot quicker and get the best finish and tool wear.

Washout
06-12-2012, 03:32 AM
Thanks,

I thought that was a little aggressive - I'm not too fussed about the length of the job (or at least won't be once I have my mist working and don't have to stand there puffing WD40 at the tool and brushing chips away for hours). I'll report back with results and must get some pics or video up.

Cheers


Chris

Washout
07-12-2012, 08:05 PM
Update:

I finished cutting my first part today after a bit of trial and a fair bit of error.

The part was a replacement gantry mounting bracket, which I had hand cut originally and thought I would get something a little more precise on the machine. I am using 15mm thick 6082 T651 Ali.

Firstly I quickly ran up the part on Vectric's Cut 2D and placed 8x5mm through holes to fix to the bearing blocks and 4x6mm holes with 4mm deep and 11.5mm dia counterbores, which fix to the bottom of the gantry. The previous mishap when drilling holes in extrusion meant I was out of 5mm carbide end mills, so opted for a 4mm carbide 3 flute which was running at 18,000 rpm 0.5mm DOC, 750m/min feed and 515m/min plunge (these were suggested by GWizard). The cuts were inside edge profiles, so spiral cutting.

This operation ran fine with no mishaps.

Then I modelled an outside profile cut to cut the part from the plate, using a 6mm end mill and tried 10,000 rpm, 0.8mm DOC, 750 m/min feed, 515 m/min plunge.

This ran fine until it got to ~10mm into the material and then the tool bogged down (probably on chips), went a little hysterical where it was stuck and before I could hit the estop. Once I had bought the spindle to a rest I figured it was another stepper/driver glitch and looking at the tool it appeared to be fine with no gummed up ali on it. So I rewound the job and started again. This time, once the job had reached material it ran fine for a couple of cycles and then again bogged down, and jammed, then snapped the tool.

I was using WD40 every so often and also using my airbrush to try and clear the chips, which was only partially successful and looking back on it I reckon the tool got bogged down in chips in the slot and this caused the jam/snapped tool.

So I figured I would tone stuff down and with Jazz's advice on feeds and speeds in previous posts, tried again with an 8mm end mill and this time used the airbrush exclusively, with some coolant to act as a mister. I also reran GWizard and with this time tried 12165 rpm, 0.2mm DOC, a 750m/min feed and 250m/mm plunge (although this time I opted for a ramp rather than stright plunge). I found the conservative/aggressive slider this time in GWizard and went for the former.

This time the job completed without mishap, but I do need to do a few things in the future and the first of those is to rig up the airbrush/mister so its on constantly rather than me chasing the tool around and trying not to tangle the airline from the compressor.

I do have chatter lines on the sides of the job, so I think I will have to bite the bullet and retro-fit a second ballscrew on the x axis, which should help with this.

I did a quick check over of the machine when I was done and the X and Y locking nuts had come loose. Not sure the set screws/grub screws are man enough for the job. so will have to think about solving that as I don't want one of the ballscrews breaking free or breaking the coupler. I was thinking about replacing the helical cut ones with the interlocking kind (don't know the name for them but have seen them on various sites).

Also my PC is far enough away from the machine that its a pain to have to keep going backwards and forwards if I want to feedhold and restart, but I'll likely live with it until I move home in the next week or so.

Once I have got a better hang of feeds and speeds and the new ballscrew in I think I'll have a go at cutting some gantry sides out of plate as well, although I'm eager to have a crack at something not fo rthe machine itself....

More to come...


Chris

Jonathan
07-12-2012, 08:32 PM
0.2mm DOC is disappointing. 12165rpm sounds a bit fast for an 8mm carbide cutter in aluminium. I use around 9800rpm (600mm/min, 1.5 DOC,8mm single flute). Looks like that was a 3 flute cutter since you've set the plunge to a third of feedrate which is what Gwizard suggests? If so then 750mm/min is slow, you wont be getting a big enough chipload. Use more like 1800mm/min if it's 3 flute.

If you want a quick way of guessing the feeds and speeds then scale it proportionately off one you know works. For instance I know that 9800rpm, 600mm/min, 1.5 DOC works well with the 8mm single flute, so if I used a 2 flute cutter I would just use twice the feedrate (1200mm/min), as its got twice the number of flutes and maybe decrease the depth of cut to reduce the force on the cutter. Using that logic, your 4mm 3fl should be run at 19600rpm and 1800mm/min...clearly this isn't the most accurate way to determine the settings, but it's a good start.

Sounds like you're describing and oldham coupling. They're pretty good, but you should be able to mount the motor accurately enough to use a solid coupling.

You could try using a finishing pass to eliminate the chatter marks - just take off 0.1-0.2mm at full depth.

Washout
07-12-2012, 09:20 PM
Thanks for the feedback Jonathan,

I went a bit shy on the DOC on the 8mm, after breaking the 6mm earlier and you're right it was 3 flute. I have the other mounting bracket to do tomorrow, so I'll pluck up the courage to up the feed rate, but I'm a bit wary of the sturdiness of my machine until I get that 2nd ballscrew on, so maybe a lighter DOC than 1.5mm (that and get an order in for more end mills ;-) ).

Another thing I need to figure out is an estop for the spindle/VFD and two pole that into the big red button for the control box.

For these pieces I'm not fussed about the chatter marks, they were quite light anyway, but a finishing pass will be in order for the other parts I'm making.

Solved my locking nut problem by binning the supplied M4 grub screws and chopping some M4 button head bolts to size (they had hardly any "hex" on the head and seemed to be made of silly putty). Ah yes Oldham Couplings - will go find some now.

Cheers


Chris

Washout
09-12-2012, 04:03 PM
Hi all,

Since my last update I've been doing some research on feeds and speeds and trying to understand the relationship between the two (amongst all the other variables).

I've used GWizard and also FSWizard to compare results and tried to compare to the advice I've had and I certainly need to use 2 flute (or even single flute) cutters to bring the speeds and feeds down (ordered a bunch and they should be here Monday or Tuesday).

Also I am a bit mystified on one thing: HSS vs Carbide end mills - Carbide F/S results from the wizards above seem to be way way higher than HSS and I know that carbide is alot stiffer, but do the carbide end mills actually need to be run at the multiples higher than HSS given the same geometry should result in the same chip loading? Or is there some other factor I'm not grasping?

Off to try and make a mount for my airbrush/mister :-)

Cheers


Chris

JAZZCNC
09-12-2012, 04:36 PM
Chris,

Just gone back and looked at the first post in this thread and I see the basic frame but can you post pics of the finished machine.?
Reason I ask is that you mention single screw and from the pic I'm thinking you must just have it down one side and not central to the gantry.? If I'm correct then your wasting your time using any calculator has the machine is so compromised with this set up they will be massively off target. Esp in aluminium.!!

EDIT: scrap that request looked a bit further up and found it.!! . . . Give it up untill you have the other screw you've no chance.!! It will turn into another beast with 2 screws.!!!

Looking at the machine and it's stubby size and strength then with twin screws it will easily cut aluminium with 3 flute cutter at 2.5mm DOC and 1500mm/min or more.
My machine is much wider and I easily cut at these figures and just today using 8mm size of this cutter (E5711 series) http://www.cutweltools.co.uk/files/ww/SM%20A160.pdf I was playing with DOC and cutting at 5mm Doc @900mm/min 9-10k rpm full slot without any trouble what so ever. I will try even deeper with 45% step over when pocketing and the tool spec says ok upto 1.5 dia @700mm/min 8K rpm.

HSS or carbide is mainly down to strength and the fact Carbide tends to work better when loaded up and can handle much greater heat and forces. To be honest I find I get better finish from HSS in Alu and actually prefer to use them for jobs where don't want to apply finish pass. With Carbide I can hog away put always have to do a finish pass for an acceptable finish.!

Washout
09-12-2012, 05:36 PM
Yeah I figured the single x axis ballscrew was causing weakness - I'll get a mail over to Chai for a new ballscrew with appropriate machine end and go the pulley/belt route (seems less prone to error than two steppers). I'm looking at Zapp/Gary for the belt and pulleys if I can decipher what length belt to order (whether the length is for the whole belt or the loop if folded in half - if that makes sense).

As always thanks for the sage advice and assesment of the machine once its been upgraded (I like the sound of "beast" :-) ).

Chris

JAZZCNC
09-12-2012, 06:46 PM
I'm looking at Zapp/Gary for the belt and pulleys if I can decipher what length belt to order (whether the length is for the whole belt or the loop if folded in half - if that makes sense).

Determine the screw centre distance and pulley pitch/number teeth then use a belt calculator (see attached) to find the number of teeth. (Or looped belt length) But if you order by number of teeth there's no room for error.!

The machine will be a beast at this size and strength. To give some idea my machine Y axis travel is 1100mm, so technically more likely to be weaker or more flexy than yours.
Smack bang in the centre at 3/4 Z extension so possible 120-130mm extension below Z axis. I've just cut a oval pocket 70mm x 30mm length single pass at 8mm Doc 45% step over @800mm/min 12Krpm in 6082 T6 Aluminium just using blown air no misting or coolant(quick initial squirt of WD40).
In doing this, it first spiralled down to depth, it then had to cut full slot for about 40mm and while it moaned slightly more than the when cutting 45% stepover it still handled it no problem and the actual 45% stepover was a doddle and I'm sure I could have easily pocketed at 45% step the full 10mm depth of material.

While bit extreme on DOC and not something I do often at full slot width. I Can't see why your machine couldn't do this kind of cutting and like mine will most certainly easily handle half this DOC at higher feeds.

Washout
09-12-2012, 08:57 PM
Wow, if I get half that performance I'll be ecstatic :-)

Thanks for the tool - that's a great help.

Now to see if I can get a ballscrew macined and delivered before Christmas.....

Chris

Washout
10-12-2012, 11:00 PM
A quick pic of the machine for those interested.

Doesn't look like I can get a ballscrew machined before Christmas, so the build/upgrade etc is going to have to go on hold. I'm also moving house next week, which means taking the gantry off or seeing the removal guys slip a disc or two and this will also be a good opportunity to fit the second ballscrew when it arrives.

For the sharper eyed amongst you I didn't have my new gantry mounting brackets on when I took this pic, but they are ready. Also for some reason the flash on my crappy camera phone has made the ali extrusion look like I have pimped it with blue LEDs, but the coolant is indeed green. I've been using PC watercooling pump and fluid, as I have been water-cooling my PC's for years without any gunk forming and using a spare radiator/fan unit means I don't need much of a resevoir for water. So far I've yet to feel my spindle get warm after a continuous 54min session.

BTW - the new house has a 31x15 foot workshop and I'm going to be building a workbench down one wall out of steel box section, which means learning to MIG weld - does anyone have any recommendations for a welder and whether gasless or gas is better?

Cheers


Chris

JAZZCNC
10-12-2012, 11:25 PM
BTW - the new house has a 31x15 foot workshop and I'm going to be building a workbench down one wall out of steel box section, which means learning to MIG weld - does anyone have any recommendations for a welder and whether gasless or gas is better?

Gas less is mickey mouse stay away if you want to weld seriously. If your mainly going to weld box section or thicker stuff then just buy a cheap stick welder because to weld 3mm box section you'll need a decent sized Mig. . . . Cheap Migs are rubbish and pretty much useless for thick stuff.

Also to get decent welds from a Mig you wan't CO2/Argon mix (Boc Argo sheild) and the little cannister bottles and regulators that usually used with cheap Migs are rubbish/expensive and last about 40mins on a good day.!! Plain CO2 or pub gas spits like crazy.
The real Mcoy bottles are difficult to get hold of unless you know somebody with an account at Gas supplier and even then you'll have to pay for the bottle rental plus Gas.!!

Even the cheapist stick welder will give better welds in 3mm box than an average power Mig.!!

AdCNC
11-12-2012, 12:42 AM
For the sharper eyed amongst you I didn't have my new gantry mounting brackets on when I took this pic

First thing i looked for. :playful:

Id go with Jazz's recommendations, i was in the same boat as to what welder to go for a few years back and last year i bought a old oxford oil filled welder off e-bay probably the best arc welders you can buy IMO. then from there i got a new Oxford S-Mig 330-1 even tho the new Oxfords are not made by Oxford but Technical Arc. They are very good welders, and for the money even the cheaper ones they do will be good.

But id look on e-bay for a old oil filled arc before you get a cheap buzz box arc welder!

D.C.
11-12-2012, 04:02 AM
Got to agree with Jazz that arc is much better than mig for this sort of thing.

Something you may want to consider is having a look on the skill swop websites to see if you can find someone local offering free welding or just find someone that you can pay to do the one off welding job for you. They will probably do a much better job in a couple of hours for beer money, so if you only have this one job that needs welding it might be a lot cheaper and easier to outsource it.

Buying your own welder even second hand, some sticks, gloves and a mask mean you are probably going to drop 50 - 150 on gear easily and then you will spend hours researching how to use the thing, hours practicing and then do a not great job of welding the frame.

Of course if you want to learn to weld and have lots of ideas for projects that need welding, it's lots of fun and very satisfying so go for it! :)

Washout
11-12-2012, 07:56 PM
Thanks guys,

I'll do a little more research before settling on something, but I'd like to learn the skill myself, as there will be other projects that require it (I need other sturdy work surfaces run up for example).

I was initially looking at MIG, as it seemed the easiest to pick up and from a few tutorials I have seen looks to be a matter of timing and motions to get a penetrating and consistent weld, but like decent soldering (or any other skill for that matter) will be a matter of practise, practise and more practise I expect.

Thanks again


Chris

D.C.
12-12-2012, 04:11 PM
In that case have a good read of this site, (Especially the safety page!) get hold of a few pieces of scrap that is the same thickness that you will be using and start practising.

Arc Welding Tutorial (http://www.mig-welding.co.uk/arc-tutorial.htm)

If you are welding anything that needs precision placement you will need to clamp it in position and put a few tack welds on or the welding will pull the work pieces totally out of alignment. It is also worth buying a pair of overalls to wear when welding so that you don't set fire to yourself.

Clean up your work space so there are no bits of saw dust or plastic bags etc lying around on you will burn the house down, make sure you are well ventilated and make sure the kids aren't around staring at you while you weld, the light can cause permenant blindness and even if you only suffer temporary blindness it is extremely painful and means a trip to A&E. This is also a very good reason not to start your weld off and then drop your mask, if you can't start the weld without binding the stick either practice more or stop being a tight arse and buy a reactive mask. :)

Also be careful after you have finished the weld don't lift your mask and stare at it, the slag that forms on top of the weld can suddenly explode off the weld ( I've had a 1cm long bit of very hot slag burn itself into my neck from doing this)

[end safety lecture]

JAZZCNC
12-12-2012, 06:13 PM
I was initially looking at MIG, as it seemed the easiest to pick up and from a few tutorials I have seen looks to be a matter of timing and motions to get a penetrating and consistent weld, but like decent soldering (or any other skill for that matter) will be a matter of practise, practise and more practise I expect.

Yes pretty much correct but with MIG's the kit and Esp the Gas make a massive difference to how easy.? A cheap low power MIG with wrong Gas is no fun believe me.!

Wobblycogs
13-12-2012, 06:54 PM
I've just been looking into welding so that I could weld up my frame and heartily recommend the site DC pointed you towards. The guys on the forums over there are friendly and knowledgeable.

I reckoned that it would probably cost around 500 to get a MIG welder and all the other kit needed to allow me to weld up my frame which is mostly 4mm box section. You'll need (at least): welder, apron, gloves, mask, wire, gas and g-clamps. When it comes to PPE I tend to go with something quality so that cost includes a half way decent auto-darkening mask and the welder is new rather than an e-bay special. I had a look around on e-bay, if you want a dinky little 90amp job then e-bay is the place to go but you'll be wanting more like a 125 to 150amps. I'm sure they turn up and a bargain can be had if you are willing to wait long enough though.

The real sticking point for me was gas, I found a local supplier that was cheap enough but I didn't really want to be signing up for a contractor and storing gas bottles in the house. I'd also need to find somewhere to store about 50kg of welding machine which would be problem for me at the moment.

wilfy
13-12-2012, 07:26 PM
any vids of your machine cutting?? it looks awesome

Washout
14-12-2012, 12:15 AM
Thanks for all the welding advice chaps, that should give me enough to go on.

@Wilfy - no video yet, as I'm going through the usual pain of 11th hour moving hurdles at the moment, so the gantry is off the machine and ready to shift to its new home. Of course I will need to get new home up and running once I land there (I have a task list that is several pages long, but luckily the CNC machine is quite high up, as it needs to be cutting MDF for a custom desk for my office along with other stuff (Sketchup pic attached).

I will get some video of it though once I have the second ballscrew on.

Later


Chris

wilfy
14-12-2012, 12:36 AM
i know the moving pain too well.. last house we were in was our longest stay since we've been together an that was 3 years, we lived in 3 houses previously and just moved again 6 months ago.

oh and forgive me for being stupid or blind.. or both.. but what is the cutting area?

Washout
14-12-2012, 12:51 AM
Cutting area is currently ~770x280x110mm, so quite small compared to many machines, although once in the new workshop I'll be making a few changes to increase the area if needed:

1. I am going to make the side support extrusions removable so if I need larger pieces I can index them and cut in sections (for wood/MDF though) just need to think about how to mount them without the need to remove the x-axis rails each time I want to switch function....

2. If I bolt the machine to the steel framed workbench and its rigid enough I should be able to "drop" the bed down to increase my Z axis (a 4th axis may require this dependant on the diameter of work piece and size of 4th axis I need or get).

Colin Barron
04-01-2013, 01:17 AM
A while ago i was looking at mig welders and gas. BOC were doing a deal through the gas supplies head office (0161)930 6822. I think it was half size argon at a decent price and the deal was called "volkszone" years rent 35 plus fill up 25. Worth a ring?

Washout
25-08-2013, 10:18 PM
Hi All,

Seems like years since I last posted (it has almost been ;) ).

Thought I'd give an update on progress, as I've finally managed to get near to the CNC machine on a long dependency chain not least of which meant rendering the workshop and infiltrating the entire property with structured cabling and updating the entire electrical layout in both the main house and my barn (some may recall a house move in January).

The machine is now assembled to where it was when I moved and in its new home and I am about to retrofit the second ballscrew on the X Axis (I've had the parts for over 6 months....or at least most of them as I find I need a bore reducer for one of the pulleys to fit it to the back of the stepper motor (8mm to 6.35mm). Also need to drill out and thread some grub/set screw holes to the pulleys, which I'm a little nervous about as they'll be 4.2mm through a fair chunk of steel.

Sharper eyes may notice the new gantry mounts are fitted and also I have Oldham couplers all round in place of the spiral cut ali versions the ballscrews came with and to help with that here are two photos of one half of the workshop and the machine itself. Now I'm getting on top of the property, rather than vice versa, I should have more updates and also a vid or two when things are moving again.

Oh and I've already got plans to build a second machine, which I hoping to build a kitchen with, as I've decided I'm very fed up buying stuff designed to a price rather than a spec ;)

Later


Chris

jcb121
25-08-2013, 11:30 PM
i have a question about your gantry, your Y axis rail are quite far apart but the X axis bearing blocks look quite close together.

does the height of your gantry make a negative effect on your machine? i'm thinking the tallness would make it like a big leaver?

Washout
26-08-2013, 12:20 AM
Hi JCB

Thanks for commenting.

I'm by no means an expert, so I'll defer to one of the J's on this, but my reasoning was that when the cutter is moving through material in the X-Axis (either towards or away from the camera in that shot), then the leverage force would be at the cutter opposite to direction of travel.

The fulcrum point would therefore be at the lower Y axis bearing and would be trying to deflect the upper bearing mounting point (or the top of the gantry). As the distance between the lower bearing and the cutter tip is going to be considerably shorter than the distance between the lower and upper bearing point (the other side of the fulcrum), then the deflection should be minimal/reduced.

Of course the forces may cause the gantry mounting brackets to flex, but I have yet to see any by leaning into the gantry at the top - I wish the same could be said for the yaw in the gantry, due to only having one X-axis ballscrew however, but this is to be shortly remedied :)

At least that's the theory ;-)

Washout
30-08-2013, 11:19 PM
Hi All,

A bit of an update and hopefully if I can work it out a vid to follow:

Its fun to be back on the learning curve and earlier in the week I retrofitted my second X Axis ballscrew. The ballscrew itself was not much of an issue, but I had some "fun" getting the pulley onto the back of the stepper motor and not wobbling about.


Anyway after doing that it was obvious I have some slack in my drive belt, which needs taking up so just for testing purposes I found something "suitable" to take some of the slack out it (see pic). I then ran a few full speed jogs around the table and all seemed well, but there's a few things I need to do:


1. Add some support to the 2nd ballscrew where it extends to the same length as the stepper motor (there's some flex in there when the axis changes direction) - a suitable bearing has been ordered.
2. Add a proper 3rd pulley to provide belt tension rather than the hasty affair you can see in the pic, although I'm not sure how much play should be in the belt - parts ordered anyway


After that I decided to push my luck, being an impatient bugger, and run a short job "cutting air", just to see how it goes. That ran OK from start to finish, which was nice although you'll hear that the belt (or maybe ballscrew alignment) is out if I can get the vid posted. The job is a 3D roughing operation from Cut3D that took ~15 minutes and I also ran the other "side" of it, which also completed OK (~15 minutes).


Flush with success I loaded up a longer job (50+ minutes) and ran that. Again all was well for 20 minutes or so, but I noticed that heat was building up in the X Axis stepper motor. Then all of a sudden during a longish X Axis run (@1200 mm/min) the stepper made what can only be described as a "graunching" sound, like it had met sudden resistance and I hit the feedhold and then the estop. After that I checked that the x axis could be moved by hand (as the motors were no longer "holding position" in the e-stop state) and it could.


I'm assuming therefore that my gantry hadn't crabbed and bound up due to the ballscrews getting out of sync, but I could be wrong as hitting the estop could have "relaxed" the binding. Also I noticed that the stepper motor was hot to the touch i.e. hot radiator temps, but I'm not sure what behaviour an over heating stepper usually exhibits and whether the stepper (3Nm SY60 Nema 23) is maybe not man enough to drive 2 ballscrews?


Anyway once I get the other parts I'll tidy up the belt and pulley arrangement and then work out next steps, which I guess might mean getting and then trying to fit a larger stepper on that axis (the other two axis motors were cold/barely warm)? Perhaps Nema 34?


Later


Chris
9917

Washout
30-08-2013, 11:28 PM
And hopefully here's the video:


http://www.youtube.com/watch?v=By8L1yb1ifQ

JAZZCNC
31-08-2013, 12:07 AM
What voltage are you running those steppers at.?
How are motors wired, Series or parallel.?

Regards heat then steppers do get Hot can you hold your hand on them for a minute without burning you.? If so then they'll be fine. They can take much higher heat and still work ok it just shortens there life.! . . They would have to be very very hot to lockup.

Your problem will almost certainly be coming from that belt setup causing misalignment and binding. Being honest now your always going to struggle with that long extension like that on ball-screw without some proper sturdy support or holding fixture. Even then it will need to be aligned spot on.
I'd seriously look at running 2 motors . . or . . Dumping the existing motor bracket then have 2 pulleys on one screw with belt on one across to other with shorter belt on other pulley to motor. This way you could apply some gearing and keep the Nema 23 which should do this job ok provided you have decent amount of volts.

That size machine with 2 x 3Nm motors running decent volts even with 5mm pitch screws should be running easily 6mtr/min at minimum.
I built a very similar machine with probably heavier gantry for Michael Marino running 5mm pitch screws with 2 x 3nm motors on AM882 drives with approx 70Vdc and I'm pretty sure he's had it up in the dizzy heights of 16-17mtr/min when pushing for it's limits.!! . . . Sure he's regularly running it around 8-9mtr/min with high acceleration.!!

IMO dump that belt setup you have now it's a serious bottle neck.!! . . . .Use belts by all means but needs to be better setup than that.!!

Washout
31-08-2013, 01:17 AM
Point taken - I knew I was pushing my luck running it like that, but was eager to get some movement after the machine had been dormant for so long.

I'm running the motors in bipolar parallel and the drivers are PM752's and IIRC the PSU (Zapp PS806-5) produces 68V.

Funnily enough I have ordered another stepper (wanted a spare and could always use it for 4th axis assuming I can get the 3 running first), so I could cut the extension down and run 2 motors - would I need another driver or wire them both in parallel off of the same X axis driver? I also thought running 2 motors was a bit risky if one starts losing steps for some reason?

I could also go the two belt method you suggest - I have a 3rd pulley coming also, which I was going to use for tensioning. The problem I foresee with that is finding the right length belts will likely mean some kind of custom motor mounting plate, which I'm not setup yet to make - if only I had a CNC machine ;)

Good to know that the heat isn't immediately fatal to the drives, thanks for that.

I'll have a think over the weekend and see which option for "sorting" out the drive to the second ballscrew I should go for (read easiest).

8-9 m/min sounds scary enough, let alone 16-17.........

JAZZCNC
31-08-2013, 02:55 AM
Funnily enough I have ordered another stepper (wanted a spare and could always use it for 4th axis assuming I can get the 3 running first), so I could cut the extension down and run 2 motors - would I need another driver or wire them both in parallel off of the same X axis driver? I also thought running 2 motors was a bit risky if one starts losing steps for some reason?

2 x motors will be your easiest route with how it's setup already.

Most common cause of missed steps is motors being tuned too high with folks trying to have both Acceleration and velocity too high.
If you keep tuning to sensible speeds then you generally won't have a problem, only other time you can run into issues is if your running very long jobs like large 3D work that can have cycle times well in excess of 20hrs. Because 3D work is usually lots of small fast moves and higher accelerations are often used then steps can get missed over time period.

Keep sensible motor tuning and you'll be fine with slaved motors but unfortunately you'll need another drive.

Run 2 x motors and with 68V you'll easily see 6mtr/min with good Acceleration without any issues provided everything is aligned properly.

Washout
01-09-2013, 08:43 PM
Thanks as ever Dean,

I could probably do with another Driver in any case, so will add that to my order with Zapp - will have to see if the 806-5 transformer will be good for a 4th Driver or I'll have to get another of those as well (4th Axis would have needed it anyway I expect).

I did a bit of tinkering over the weekend and one of the x axis ballscrews was out of alignment, so have also rectified that whilst waiting for parts.

More as I work on it.....


Chris

JAZZCNC
01-09-2013, 10:56 PM
I could probably do with another Driver in any case, so will add that to my order with Zapp - will have to see if the 806-5 transformer will be good for a 4th Driver or I'll have to get another of those as well (4th Axis would have needed it anyway I expect).


Erm . . well 6A is low for 4 drives and personally I would.!! . . . But worth trying first and will probably only cause you problems if pushing hard.

Being honest with you and Gary probably won't thank me for this.! . . Instead of buying another then I'd build one to suit your requirements exactly. It will be cheaper, take just the same room in control box has the one you have now. If you need the 5V then buy a cheap linear supply for 5.
If you then sell the existing 806-5 it will end up costing you next to nothing.!!

Washout
02-09-2013, 11:23 PM
LOL would you believe it, just as I think "yep, I'll go with the dual motor setup on the X-Axis" Sod's law kicks in and the PM752's are not easily available (ebay from China only), but there are comparable digital drivers available and my guess is I'll have to switch out the Z Axis driver to be the 2nd X Axis and run the new driver in its place. That's assuming MACH3 has no problem with a mix of drivers (provided they are tuned to the same steps per rev etc.?)

JAZZCNC
02-09-2013, 11:32 PM
That's assuming MACH3 has no problem with a mix of drivers (provided they are tuned to the same steps per rev etc.?)

Mach doesn't know or care what drives you use it just sends pulses to the drives based on your preferences in motor tuning.
You'll see a difference between digital and Analogue drives but it will be positive not negative and the smoother Digital drive will work good on Z axis anyway.!

m.marino
07-09-2013, 05:39 PM
JAZZ, Not quite that fast but have gotten it up to 12.5m/min rapids with an acceleration of 1.5m/sec^2. These days I am running it most the time at rapids of 7m/min on the x and y axis and 5.2m/min on the z axis with a acceleration of 1m/sec^2. Which with the mentioned motors and drivers and a well set up CV mode gives me cutting speeds possible of 4.8m/min with a 4mm Single Flute Lapped end mill with a 6mm shank. As JAZZ asked what voltage are you running at. I could run in crazy land IF I moved acceleration down to about 500mm/sec^2 but let me tell you 12.5m/min gets scary on my machine. -Michael

Washout
12-09-2013, 04:41 PM
Hi Michael

Sorry I missed your reply - I'm pretty sure I'm running the steppers at 68V (that's the output voltage of the PSU).

Those are some pretty hairy speeds you're getting on your machine and I'll be very happy to get near that.

I also had some bits arrive today:

a. some components to brace and tension the belt between the two ballscrews (and support the "wobbly " extension. I had ordered these anyway so will try them out, as I redid the second ballscrew alignment and I must have missed a step in my haste and it was well out (2mm or more). Chances are that's where my overheating stepper was coming from and gantry binding.
b. Also a spare M752 came from ebay (China), so will use that as plan b, should a. above not work out. I'll also likely need a custom motor mount as the T-slots don't line up with the mounting bolts on that side of the machine.

More as I get on with stuff.....


Chris

meerkat
12-09-2013, 08:49 PM
Hi.
First pic looks like a mechanical beast,but good!
A little tip for tapping holes. I use my B&d pivot driver saves a lot of work and no aching hands!
meerkat.

Washout
13-09-2013, 10:42 AM
Hi Meerkat,

Thanks - "beast" is the target after Dean' comments further back :-)

I made a little progress last night after putting a bracing structure and bearing onto the 2nd ballscrew extension - this has cut down flex considerably (i.e. hardly any movement even when leaning into it) and I did a test by jacking my motors up to 4000mm/min in MACH3 and had no stalling on full speed jogs. So a step in the right direction methinks. I'll likely push the speeds up again tonight and see where the stalling starts, but I'm quite happy with 4m/min, as in its original guise the machine wouldn't go above 2.5m/min reliably with the single ballscrew configuration.

To do tonight will be to put the belt tensioning pulley on and take the slack out of that (the bracing has improved that also anyway). Then run some more tests to see if there's been improvements.

I'll then think about plan b dependant on the results of that.

Chris

m.marino
13-09-2013, 10:59 AM
Just for giggles I set up my machine to where it stalled and then backed it off. Now this is a rough set up and I have NOT gone back and tuned the drives so that would change things as well. Given that I was able to get x(x/c) axis up to 17m/min V with 450mm/sec^2 A; Y axis was gotten to 11m/min V with 350mm/sec^2 A and Z to 7.5m/min V with 250mm/sec^2 A. Let's just say that is a bit scary in speed.
-Michael

Washout
13-09-2013, 11:32 AM
Wow - effective limit switches are a must at those speeds I expect ;)

Even 4m/min was hairy compared to what I've been used to previously when I tested it last night.

Oh and a quick question or two: I'm running the same test job "cutting air" to see what progress I'm making as I make mods etc. to the machine and I notice that whatever max speed I set the motors to is what the test run seems to use as its rapids - I'm assuming that's normal behaviour?

The other thing is when I first power up and load MACH3 the motors always seem to need one axis setting up in MACH before they "remember" the settings from the last session?

Washout
20-09-2013, 10:43 AM
Hi All,

Quick update (no pics or vids yet).

Last night I completed the mechanical option for running two ballscrews (I still have parts for plan b if needed), by adding some extrusion bracing and additional bearing to the long "bendy" extension on the back of the 2nd ballscrew. Alignment was surprisingly easy, as it was a case of using the already aligned ballscrew and using a transfer punch to mark, drill and tap the new bearing mounting holes. Things are much more sturdy there now.

I have also added a 3rd pulley mounted on more extrusion and an axle, which I can adjust up and down to increase or decrease tension on the belt (better than what my car engine uses methinks :) ).

So ran some tests - 4,000mm/min was no problem at all, so upped it to 6,000 mm/min and its still fine, but getting quite scary at that speed (I'm wondering if I should back off of 500mm/s/s acceleration?).

I'm going to start some cutting jobs, once I have solved my VFD issue (see other thread in Electronics section) and then its back to the feeds and speeds learning curve :)

Cheers


Chris

Jonathan
20-09-2013, 11:31 AM
(I'm wondering if I should back off of 500mm/s/s acceleration?).

No, you want to increase the acceleration if anything. I run my machine at around 1m/s^2, so twice what you've set. Part of the reason m.marino got such high feeddrates in post #91 is that the acceleration was set very low, and it's a compromise between high acceleration and speed.

For example, in this video I set the acceleration lower and put a bigger pulley on the motor, and got 60m/min:
https://www.youtube.com/watch?v=k3MG7ww49A8
... but it's not terribly useful.

mekanik
20-09-2013, 02:42 PM
Love those Ball nuts

JAZZCNC
20-09-2013, 06:54 PM
No, you want to increase the acceleration if anything. I run my machine at around 1m/s^2, so twice what you've set.

Be very careful here because while Jon's right acceleration is useful it's also the biggest cause of missed steps.!! . . . .6mtr/min and 500m/s/s on 5mm pitch screws is not unreasonable speed for first use and unless your doing lots of 3D work or lots of short moves then it will be fine for most work.

Like Jon's says it a compromise of velocity or Accel so if your happy 6mtr/min then wind the acceleration up until motors stall and then back off 10-20% and you should be fine.

Jonathan
20-09-2013, 07:18 PM
Yes, clearly if you set the acceleration higher you want to test the system thoroughly to make sure the motors wont stall. After doing what Jazz suggests, I suggest running some code from here:

http://www.mycncuk.com/forums/gcode/5843-random-g-code-testing-motor-tuning.html

Then you can change from 'should be fine' to almost certainly fine.

JAZZCNC
20-09-2013, 07:32 PM
Yes, clearly if you set the acceleration higher you want to test the system thoroughly to make sure the motors wont stall. After doing what Jazz suggests, I suggest running some code from here:

http://www.mycncuk.com/forums/gcode/5843-random-g-code-testing-motor-tuning.html

Then you can change from 'should be fine' to almost certainly fine.

That's handy Jon not seen that before and with G61 exact stop mode at G0 speeds with high acceleration it will double to test of how anchored the machine is.!! . . . . . Again careful has with high settings the Bloody thing will try to walk off. . Lol

Washout
24-09-2013, 01:03 AM
Some progress tonight, whilst I had some "workshop time".

I decided to do a complete re-wire of the VFD a. to give me some more space with the machine and b. to ensure that the old stuff was not at fault. I also replaced the RCD in the workshop and now rely on the RCD at the consumer unit which feeds the barn/workshop where its split off from the main house supply.

The VFD fired up first time and still had my old settings in it, which was nice. So I then ran a quick 2D job on some 6082 aluminium, which was very light work i.e. one central 3mm drill mark at 0.1mm deep and then 4 surrounding 6mm "dowel hole" marks at 2mm deep using ramping. Used a 3mm 2flute carbide bit at 16800rpm and 800mm/min. The job itself ran a treat apart from 3 spurious e-stops (tried a debounce setting in MACH3 of 1000 and 3000, but still had a phantom e-stop). I'll have to check earthing continuity on all the CY going back to the control box, as I know my spindle is earthed back to the VFD and the CY carrying the spindle cables is also earthed at the VFD end only i.e. no loop.

Interestingly I checked continuity/resistance from the spindle to various parts of my machine and I'm getting negligible readings from the spindle, through the gantry and all the way to the main frame.

Still some progress and light possibly at the end of the tunnel - at least the job I did complete showed very little grumbling from the tool and ripped through what it did have to do, with nice chips, which gives me some hope that I can start doing something other than work on the machine and make the machine do some work for a change ;-)

More as I get time....


Chris

Washout
29-09-2013, 02:24 PM
Another quick update:


I still have spurious e-stops, which I thought I might get rid of by running the VFD off of a circuit in the workshop which does not go through the dado trunking next to the control box supply, but to no avail. Also checked all my earthing for the control cables and that is OK i.e. CY shielding going back to the control box star point, with good continuity. Also tried Mach3 de-bounce settings up to 3000. Only thing I have left I think is I replaced a big chunky, but very long (3mtr) parallel cable with a shorter one, so will switch that back to see if there's an improvement.


On the bright side, to do the testing for the spurious e-stops I ran a job in beech block just to try and the following pic resulted, which despite the e-stops probably losing steps when incurred, is at least something and came out nice and clean apart from where the wood won't hold the very thin details. The 3mm end mill went through the wood like it was not even there (ran at 18600 rpm at 800mm/min on a 3flute), so aluminium next :-)

Washout
29-09-2013, 03:16 PM
Even quicker update: looks like all parallel cables are not equal, as I have just replaced that for the old one and all is now well - test job ran without any e-stops :-)

Now for some cutting.......

JAZZCNC
29-09-2013, 08:04 PM
Even quicker update: looks like all parallel cables are not equal, as I have just replaced that for the old one and all is now well - test job ran without any e-stops :-)

Now for some cutting.......

When having any problems with signals first place I always go is the Parallel Cable and swap it for a known good one.

Pleased your sorted.!

Washout
30-09-2013, 11:36 AM
Thanks Dean,

I ran a couple of jobs yesterday: one was some text in plywood for a sign, which was long but ran fine. The other was a first go in aluminium doing some roughing 3D work (I'm using Vectric 2 and 3D for my toolpaths).

The aluminium didn't do too well and I'm pretty sure I'm being a plank with regards feeds and speeds. Material is 6082 as per previous posts and I was running a Carbide 2Flute End Mill @ 12000rpm and 1200mm/min with 2mm DOC and 30% stepover (maybe a bit aggressive on the DOC?) and was blowing air. Anyway the machine seemed to hold up well i.e. nothing broke but was grumbling a lot so I hit Feedhold and saw that the edges in the pocket were clean, but it was obvious the cutter was trying to erode its way through rather than cut as there was a good "weld" of aluminium on the cutter and a "lip" around the edge of the pocket.

My guess is my feed rate is too low (possibly rpm to high) and I'm re-cutting chips?

What was cut didn't have visible chatter marks on the walls of the pocket, so I'm tentatively confident the rigidity is OK (2 ballscrews does indeed help :) ).

Will have a play later with a less aggressive DOC and high feed rate, but if anyone has their own feeds/speeds I can try then it may save me some time in trial and error.

Chris

Jonathan
30-09-2013, 11:48 AM
Assuming that's a 6mm cutter, then the feedrate and speed you posted should be fine. With only 30% stepover you could actually increase the feedrate slightly due to chip thinning, but I'd leave it for now. I tend to agree that rigidity shouldn't be a problem.

You're problem is almost certainly the lack of coolant. Air will help to cool the cutter, but not much compared to coolant (i.e. mainly water). Also, with no lubricant the co-efficient of friction is greater, so more heat is generated. You'll probably find that even a small amount of coolant will do the trick, or if you've not got any then WD40 works well on aluminium. I certainly wouldn't advise using it regularly though as it wont be good for the bearings.

If you think the rigidity of the machine is a problem, as a general rule don't change the feed or speed, instead lower the depth or width of cut. The reason is if you've chosen the 'correct' feedrate and rpm, then the chipload will be a particular value which you want to maintain. If you just change the feedrate or the spindle speed the chipload will change.

JAZZCNC
30-09-2013, 04:36 PM
If you think the rigidity of the machine is a problem, as a general rule don't change the feed or speed, instead lower the depth or width of cut.

Agree with Jonathan keep playing with DOC and you'll find the sweet spot for your machine. Be aware it will also change depending if your Full slot cutting or Side cutting so often you'll need to find a balance depending on the job and how much it goes in or out of full/side cutting.

A good clue to cutting correct is to check heat of material and chips.? When your cutting correct the Chips will be scorching hot but the material just slightly more than warm.
When your more experienced you'll here the sound and know if your close to cutting correct but at the moment these sounds will probably be
making you nervous and weary.!! . . .Put some ear plugs in and listen to it you'll not be quite so on edge and it dampens mechanical noises and lets you hear cutting sounds better.!!

Also Set your VFD to display Amp's rather than RPM this will give clue to how well cutter is coping, also shows when cutters getting worn while jobs running. If Amps are low while cutting then this gives a clue that you can probably take deeper cuts.

Regards sticking then Mix some paraffin and old engine oil in a squirt bottle and give the cutter a blast every now and again to help with sticking but when you have found the correct chip load blown air is enough to clear chips and stop re-cut/sticking. I never get sticking and for most Jobs only use blown air with few squirts of coolant, for jobs where finish matters then I run a mister system. (just can't bloody breath when it's at full chat.!!)

Now test it properly for accuracy and cut the Aztec calendar.!! . . . That sorts the Baggy fanny's from the tight chuff's .. Lol

Washout
30-09-2013, 04:41 PM
Thanks Jonathan as ever.

I'll try with some liquid coolant next and see how that goes - reason I thought I was being a numpty in the F&S dept. is that most calculators are throwing up higher feeds generally e.g. 18-1900 mm/min, which I can try but that sounds aggressive, although maybe its not aggressive enough?

Must try and remember where I got my last batch of mills from, I'm likely to be placing some more orders ;-)

Washout
30-09-2013, 04:45 PM
Lol - I'll have a crack at the Aztec calendar soon I promise :-)

You're right the sound is a little disconcerting - was even louder than my bloody noisy compressor ;-)

Jonathan
30-09-2013, 04:47 PM
If the coolant doesn't help then take a picture / video and that should make things more obvious...

Washout
03-10-2013, 10:55 PM
Evening all:

A big milestone tonight, thanks again and as ever to Dean and Jonathan (I would be really struggling without you guys).

Firstly, the talk of slack Alices and tight chuffs had me checking over the machine for loose stuff and I found a couple of bolts that hold the Y Axis ballnut housing to the Z-Axis that needed tightening, so that helped as I think there was some play in the Z-Axis "rolling" about the Z Axis if that makes sense.

I also hooked up a bottle of PC coolant to the airbrush "mister" (will look at something less expensive for the future) and although its a bit vicious compared with a true mist i.e. almost flood I figured too much was better than nothing.

So I ran a roughing job for an aluminium catch in the normal 6082 using 12000rpm and 1200 mm/min and a DoC of 1mm. This ran, but I need to refine my toolpaths or how I'm generating them in Vectric, as the machine romped through the initial slot, but I had neglected to notice the toolpath was using alternate climb and conventional along the Y-Axis.

Whilst there was no mishap like last time, the difference between a conventional path compared to a climb was night and day i.e. the climb milling threw the chips out (and this time the chips looked like nice shiny chips rather than welds :) ). The conventional paths didn't throw the chips as well but the machine grumbled through the material, in fact the "grumbling" was vibrating the whole workbench

I also have some strangeness in the toolpath where some long cuts were stepping along rather than smoothly running and resulted in not so much chatter marks but waves.....weird.

Here's some pics of my first completed run in aluminium - I would have videoed, but I was hovering over the feedhold and e-stop :-) Also bear in mind this is a roughing pass and the "chatters" you can see are where the conventional milling occurred. The gouge you can see at one end of the oval was the "weld" and crap from the previous attempt.

Anyway, I am going to play with the strategies in Vectric to see which one does the most climb milling, as that was just eating through the material with nearly no "noise" at all.

I also have a finishing pass to run on the job shown as well as the other side - I hope the tabs hold up.....

More to come



Chris

m.marino
06-10-2013, 05:43 PM
Chris you can set climb as an option in CUT3D. I use Vectric (VCarvePro, CUT3D, PhotoVCarve and also CamBam [Not Vectric]). Why I use the different ones, is a horses for courses item and I could not afford Aspire at that time (the ones I got hurt the budget enough though they are help increase earning a good bit). Drop me a message and will gladly help as able on Vectric. CamBam has some very useful ability if you are doing isolated 2.5/3D type work (I use it to help make chromatic harmonica combs and they are complex planes on the interior). I am very sound sensitive and Dean is correct in that once you get used to it you will know what sounds right.

It is looking really nice. If you are looking for vacuum tables I would look at Sorotec out of Germany as they are selling a really nice model from a US firm and they are well made. I am finishing building a wooden one that will do the job I need and be fit to my work area. If looking for a vacuum set up let me know. -Micheal

Washout
07-10-2013, 02:55 AM
Hi Michael,

I thought I had tried everything to get CUT3D to do climb only. Can you let me know where the setting is?

I use CUT2D for standard work like signs etc. and know that has a setting.

I've also just downloaded a trial of BOBCAD, but that is crashing on my machine and likely something in my models it doesn't like.

I don't think I'll be looking at a vacuum table just yet, but if you're build logging that I'll definitely keep an eye on it. I'm happy enough with toe clamps and t-slot nuts and bolts so far and I also have a decent machining vice from ArcEuroTrade, which I've yet to try out.

Cheers


Chris

Washout
02-11-2013, 06:42 PM
Hi guys,

Thought I'd post an update, although this is becoming more of a "learning machining" thread than a build log as such.....

I've still not found the climb vs. conventional setting in Cut3D, so have been persevering with the alternating raster strategy that get's output. This has been OK and in fact after running over the whole machine and torqueing everything down, it seems to be making less of a noise about it.

Roughing passes seem to be OK, but I'm not too impressed with the Vectric Cut3D finishing passes, which when using a ball nose cutter are not coming out very smooth (i.e. I have ridge lines), so I probably have a setting wrong somewhere. In the meantime I think I'll be using Jonathan's advise and leaving 0.1mm+ clearance and then run another pass to finish (both using the roughing strategy).

I've also tried my first 2 sided part, which cut fine apart from my bolts must have moved a few tenths of a mm, as I had an nice offset step in it.....doh. Anyway I also then tried the same part using the precision vice from ArcEuro, which is a great bit of kit. I also had some "fun" here, when my z axis zero to touch plate worked great and then I got cocky and tried the zero wizard (I'm using the 2010 screenset). This resulted in the tool sailing through my touch plate and leaving half of it lying on the bed when it hit the workpiece and continued on....at least I can use the shank as a new touch tool I guess (need to watch the tutorial video again to see what I did wrong there).

I think it was one of Dean's posts where he reckons more time is spent working out work holding than actual milling and I think I'm finding that to be very true.

Also I have been using APT for my 2 flute cutters, but I am need of some 3mm end mills which have a 3mm shank rather than APT's normal fatter shanked variant. Any suggestions for suppliers are more than welcome (also need to try 1 flute cutters).

Still on my to do list are wire up my limit switches, wire up and get the MACH3 spindle control plugin working and also look into a better misting solution than my airbrush, as that's not very elegant (anyone know where you can get misting nozzles and the rest of the stuff required).

I did however, manage to use some spare sheet Perspex/plexi-glass with some case making extrusions and build a bit of a "fish tank" affair for the machine a. to keep some of the swarf at bay and b. retain the excess "misting coolant".

I'll try and get some video up again soon, but I will likely be busy with a new lathe this week if all goes well.

Later


Chris

Washout
20-11-2013, 10:45 PM
Another update on progress and again machining rather than building the machine, even though I have upgrades queued up waiting for some time.

Anyway, I have a paying customer now and even though my machine is still in the later stages of build he's an understanding sort, so I'm doing my best to oblige. So the pic attached are the result of a few 2D operations -
1. was a 0.2mm facing pass to get rid of the "texture" on the aluminium plate (40mm thick 6082 T651)
2. was a pocketing operation to do the hole in the centre
3. was the pocketing/profiling operation (on one side only at this stage)

All were programmed in Vectric Cut2D and this is where some "fun" started - my machine for some reason likes to mill when running in a clockwise direction in a pocket, which I think is conventional milling i.e. the spindle is running clockwise as you look down on it from the top. My understanding however is that climb milling should be better?

Also another observation is that compared to climb milling around the edges of a pocket my machine actually seems to like doing slotting better?

All 3 operations above were using a conservative 1mm DOC @ 10000rpm @ 800mm/min using an 8mm 2 Flute HSS End Mill-

I think I can go much more aggressive in DOC but I do have an issue with getting the chip out of the pockets at deeper than about 10mm - I probably need a more powerful compressor or wing at on blown air alone rather than the heavy mist I'm currently using, which tends to pool cutting fluid/water/chips mix in the corners and bottom of my pockets. Shallower cutting is great as the chip are getting thrown out nicely.

The results of the main pass and facing is a little rough, but I guess that's why its called a roughing pass. The centre hole pocket is much better.

I will eventually get some more video once I can trust the machine to run on its own on an operation without me brushing out chips and generally hovering about the thing watching.

Cheers


Chris

EddyCurrent
20-11-2013, 11:07 PM
my machine for some reason likes to mill when running in a clockwise direction in a pocket, which I think is conventional milling i.e. the spindle is running clockwise as you look down on it from the top. My understanding however is that climb milling should be better?

I plan to cut hardwood but currently I use an overhead pin router with templates, as you say with the cutter running clockwise from above I would be moving the workpiece anticlockwise as this is the only safe way to cut hand held stock. If it's moved in the same direction as the cutter then the cutter tends to acts as a power feed and just pulls it out of your hands but if you can hold onto it the finish tends to be better. So maybe the best plan is to use conventional milling for the most part and leave a finishing pass in the opposite direction ? at least with it being a cnc and everything clamped it should not 'run away' as a hand held would.

m_c
20-11-2013, 11:29 PM
Climb milling will produce a better finish, but only on a sturdy machine with near zero backlash and flex.
Any flex or backlash can cause the cutter to grab the workpeice leading to an inconsistent cut and poor finish.

EddyCurrent
20-11-2013, 11:31 PM
Climb milling will produce a better finish, but only on a sturdy machine with near zero backlash and flex.
Any flex or backlash can cause the cutter to grab the workpeice leading to an inconsistent cut and poor finish.

Exactly what I experience with hand held routing.

Edit: following Wobblycogs' post below I'll have to clarify that by saying hand held routing I'm not meaning with a cnc machine but just with a standard woodworking router or my overhead pin router. In both cases hand holding the workpiece or at least using some hand held push-stick is the correct method.

Wobblycogs
21-11-2013, 07:29 AM
Climb cutting and hand held routing is asking for trouble. You'll get away with it in something like MDF where there's no grain but I wouldn't do it with real wood. There's too much risk of stresses in the wood or an awkward bit of grain suddenly causing the router to take a bigger bite than you are expecting and getting away from you.

Washout
21-11-2013, 10:45 AM
Thanks guys - that makes sense - so conventional gives the machine an easier time but doesn't always produce a good finish, whilst climb beats the machine up a bit but finishes better, provided its sturdy enough.

I'll try a conventional version of the toolpaths and see how that goes on the other side of that block (hence the L Shaped locating bracket seen in one of the pics).

Whilst I remember I also have a MACH3 query if anyone knows the answer: when I Feedhold (to clear chips) and I hit Cycle Start to resume cutting, the operation I interrupted resumes, but at a much slower speed than before I hit Feedhold. Is there a way of stopping this behaviour as changing my feed rate is a bad idea......

Cheers


Chris

m_c
21-11-2013, 11:51 PM
Is the feedrate, or feedrate override changing after you resume?

It could be a macro is getting called by some stage during the feedhold, and changing something.

EddyCurrent
22-11-2013, 12:21 AM
Could it be anything to do with one of these ?

10.7.28 Set Feed Rate Mode - G93, G94 and G95
Three feed rate modes are recognized: inverse time, units per minute and units per
revolution of spindle. Program G93 to start the inverse time mode (this is very infrequently
employed). Program G94 to start the units per minute mode. Program G95 to start the units
per rev mode.
In inverse time feed rate mode, an F word means the move should be completed in [one
divided by the F number] minutes. For example, if the F number is 2.0, the move should be
completed in half a minute.
In units per minute feed rate mode, an F word on the line is interpreted to mean the
controlled point should move at a certain number of inches per minute, millimetres per
minute, or degrees per minute, depending upon what length units are being used and which
axis or axes are moving.
G and M-code Reference

Using Mach3Mill

Rev 1.84-A2
10-32
In units per rev feed rate mode, an F word on the line is interpreted to mean the controlled
point should move at a certain number of inches per spindle revolution, millimetres per
spindle revolution, or degrees per spindle revolution, depending upon what length units are
being used and which axis or axes are moving.
When the inverse time feed rate mode is active, an F word must appear on every line which
has a G1, G2, or G3 motion, and an F word on a line that does not have G1, G2, or G3 is
ignored. Being in inverse time feed rate mode does not affect G0 (rapid traverse) motions. It
is an error if:
♦ inverse time feed rate mode is active and a line with G1, G2, or G3 (explicitly or
implicitly) does not have an F word.

Washout
22-11-2013, 11:08 AM
Thanks guys,

Its the federate that getting changed on clicking cycle start after a feedhold (I'll run it again today to see if there are any other indicators) - oddly enough it only does the slower federate for one segment of operation (I think more than one line of code is run through) i.e. if my tool path is say doing a profiling cut first and then 4 pocketing cuts in the corners of the part and I feedhold during say one of the corner pocketing operations, it will do only that one at a slow feed rate and then the next segment and onwards run normally (after a rapid usually). Its like MACH has determined there was a feedhold and resume and is being cautious on the resumed operation before going at the job normally.

I certainly have a G94 line in, as you can see from the header in the job here:

( Roughing Out 8mmEnd )
( File created: Thursday, November 21, 2013 - 02:40 PM)
( for Mach2/3 from Vectric )
( Material Size)
( X= 202.800, Y= 153.500, Z= 40.000)
()
(Toolpaths used in this file:)
(Roughing Out 8mmEnd)
(Tools used in this file: )
(1 = End Mill {8 mm} TC 2Fl)
N100G00G21G17G90G40G49G80
N110G71G91.1
N120T1M06
N130 (End Mill {8 mm} TC 2Fl)
N140G00G43Z5.000H1
N150S10000M03
N160(Toolpath:- Roughing Out 8mmEnd)
N170()
N180G94
N190X0.000Y0.000F800.0
N200G00X-40.947Y-54.860Z5.000
N210G1Z0.000F100.0

I'm figuring its something in the way MACH is setup, but can't find it.

Cheers


Chris

Washout
10-12-2013, 09:18 PM
Hi All,
.
An update, a couple of pics and once Youtube has done its thing a video, which I'll edit into here with the link later.
.
Firstly the pics below are a sub-frame mounting bush, which on the normal road car is an ali and rubber affair and in order to increase stiffness on the track are being replaced by these. This is just a roughing pass mostly - the finished bits being the counter-bored hole in the centre, the bottom half which is all 2D stuff and the M8 tapped mounting holes. So this is a bit of a milestone, as its the first proper part to come off of the machine apart from the L-shaped jig so the part can be flipped for 2 sided machining.
.
The warts as I see them are a few ledges here and there where the machine lost steps, due to a bent ball-screw, which I am replacing, general chatter marks, which I should normally clean up with a finishing pass, but as this is going to be buried far from sight won't do unless my customer insists on paying extra for a run over with a ball nose mill ;-).
.
Other bits I am intrigued about is the pretty pattern on the top which is very regular and I'm guessing is feed or speed related?
.
Also once you guys can see the video, I am running what was to be a kind of pocketing pass to cut the part out, but it was such a faff I stopped that and just went for a slotting cut out path. Also the path you can see there hopefully shows the machine behaving nicely when slotting and chattering away when cutting on one edge (using climb is worse) and for info this is 6082 ali with an 8mm HSS end mill @ 10000rpm, 1mm DoC, 800mm/min and ramping over 10mm combined with a 600mm/min plunge (seems to be the sweet spot for my machine so far).
.
The path was also being run after I had done the bore and two pockets, so there is a fair amount of air cutting going on, but hopefully shows a variety of behaviour.
.
Once I have replaced the bent ball-screw and hopefully resolved the stalling/missing steps issue, I'm going to have to play about with the machine as the ease at, which it slots seems to indicate I should be increasing my step over when pocketing/profiling (currently 20% or 1.6mm) or increasing DOC, as the machine actually seems happier when going a bit more aggressive.
.
I'm also exploring other CAD/CAM options, as whilst Vectric is OK, there are anomalies like the path in the video, which ramped on most, but not all plunges and driving the mill in the Z axis @ 600mm/min beats the snot out of the machine/tool, as does a too shallow or slow ramp for that matter. Having a teenage daughter could prove useful for a change, if I/she can get educational licences for stuff like Solidworks (I don't think SolidCAM do educational, as I would love to try their iMachining stuff) - its about time she started paying her way anyway ;-)
.
Before I forget I have also got suspicious about the ZPA5-INT board, so for a few quid more have ordered a controller recommended by Dean in another thread (can't remember the name of it), so will report on that once I have it fitted and it looks like an almost drop in replacement for the current BoB.
.
More when I get the time (Christmas vacation is almost upon me so hopefully not too long) and I have at least 3 more of these parts to make, so will try and get a longer video up next time.
.
Later
.
Chris


http://www.youtube.com/watch?v=aJfs9SsB_Fk

EddyCurrent
10-12-2013, 10:13 PM
Are you using MIDI to control it and the mixer for analogue operations such as spindle speed ? :glee:

Washout
10-12-2013, 11:02 PM
Now don't give me more ideas - I have enough bloody wiring stacked up to do on the machine as it is - hmmm thinks faders wired to feeds and speed control and hey its all XLR connectors isn't it? ;-)

EddyCurrent
10-12-2013, 11:22 PM
I couldn't tell from the video if you had coolant/lube onto the work piece.

Washout
10-12-2013, 11:33 PM
Yeah using cutting fluid at ~6% dilution through the airbrush - it seems to do the job cooling the tool, but makes the chips stick like the proverbial to a blanket and sods law mostly in the pockets/slots I'm cutting. I'm getting a new compressor this month with more PSI so hopefully that stuff will get blown out of the cutting path and probably onto the ballscrews/rails knowing my luck ;-)

JAZZCNC
10-12-2013, 11:54 PM
Strange tool paths those it seem to cut more air than Chips.??

Your correct about step over being small and could possibly be why your getting chatter has not enough tool engagement. At those DOC you could easily do 50% thou I prefer 45%.
I would do some playing around in scrap with different DOC and STO etc and think you'll be pleasantly surprised just how much more aggressive you can be.!

Washout
11-12-2013, 03:23 PM
Yes it is a bit strange as its a pocketing/cut out path which I was running after having done the two rectangle pockets. The idea being that the wider channel/pocket being cut would allow for chips to get thrown out.

Had my nerve held with the camera/phone, you would have seen it ploughing through more material once it got down to the level of the previous pockets, but I eventually ran out of patience and just went for a slotting cut-out and frantic brushing of sticky chips caught in the narrower slot.

I'll definitely being playing with some scrap to try some more aggressive strategies, as I agree the conservative approach seems not to be optimal.

More as I find time to play/work on it........

Washout
04-01-2014, 09:42 PM
Hi All,
.
A quick text update and a shout out for a supplier (no I am not on commission, but credit where due):
.
I have today rigged up a new compressor to my airbrush/mister solution and its a 50L Sealey unit SAC5030VE, which was posted in another thread - this thing is a breath of fresh air (pun intended) and with a blow gun will help considerably with chip clearance, which was one of my previous bugbears. Props go out to Sprayguns Direct, who I dealt with directly off of their website rather than their ebay shop and they delivered quickly and gave good comms/phone calls/emails throughout and threw in a more expensive blow gun in place of the one I ordered which they didn't have in stock...so credit where its due and we'll see how it handles long term usage.
.
I have also got suspicious of the ZPA5-INT controller being a cause of some missed steps and other idiosyncrasies, so took a punt on an Optoport V3, which I'm hoping will be more reliable and I've literally just finished wiring this into my control box. Couple of differences on that board I hopefully haven't fallen foul of are that it needs a 15-70V power source, so have taken a feed off of my main transformer (68V) and also it only has one Enable connector which is wired to all the drivers in parallel and also doesn't have any pins associated with it. I have therefore disabled all the enable pins in MACH3 and redone all the pin mappings for the Step and Dir Outputs. I'll test this tomorrow and feedback on results and see if it is any better.
.
Oh in my other post I alluded to ballscrew misalignment and finally tracked down an issue on my original X Axis ballscrew, which has been there since I originally put the machine together i.e. a piece of extrusion in the hole on the gantry sides, which whilst not preventing that ballscrew turning, was causing some binding - it now turns freely and smoothly by hand so should be much better generally.
.
I am also about to rectify any inaccuracies in my manually marked, cut and drilled Z Axis, by machining two new plates, whilst at the same time replacing the old and probably "weak" spindle mount plates with one of those linear bearing blocks, which have the 80mm bore once the bearing unit is removed (can't believe how cheap that was compared to some components).
.
Anyway more posts, pics and vids inbound once the machine is running again and cutting stuff.....
.
.
Chris

Washout
10-01-2014, 09:22 PM
Hi All,
.
Quick update and a pic:
.
Over the last week I have been retro-fitting my machine with the new spindle mount using the 80mm bore bearing block I have had for a while now. As I also had some stalling issues on the Z-Axis I wanted to replace the manually marked and drilled axis plates with new ones and used the machine prior to dismantling to machine these out - making CNC is a lot easier when you have CNC obviously ;-)
.
So here is a pic of the partially assembled new z-axis with its new spindle mount. I have to fettle the rails/bearings a little before locking them down, but given the two halves slid together easily, the alignment of the drilled holes etc. on the machine is a lot more accurate than the MKI eyeball I used previously.
.
I also discovered when disassembling the axis that two of the SBR16 bearing blocks were loose and may have accounted for some of the chatter I was getting and may also be evident in the pic on the edges of the plates.
.
I will hopefully have some video once the axis is remounted.
.
Cheers
.
.
Chris

EddyCurrent
10-01-2014, 09:43 PM
Looks like a substantial mount, I hope you didn't nick the slides off those drawers for the X axis :joyous:

Washout
10-01-2014, 10:01 PM
Lol - the mount has to be better than the plates I was using that were just butted up against the axis at 90degs.

The drawers are part of a refurb project as there were two units rusting in the workshop when I moved in. They will do as a hiding place for new CNC parts purchases ("honest darling I bought these [insert expensive kit] years ago and just hadn't fit them to the machine and they have been lying in this old drawer") that and/or a place for storing scrap aluminium (a smelter for recycling is on the cards also) ;-)

Washout
10-01-2014, 10:18 PM
Oh forgot to mention - the new compressor (something I couldn't hide from the Mrs in the drawers ;-)) is bl**dy awesome compared to my previous unit and I was almost carefree when slotting out the new Axis plates, as a blast here and there with the blow gun cleared any stuck chips whilst the airbrush "mister" was a lot more efficient at cooling/chip clearing in general. I do have to get something different as a "mister", as the airbrush fogs the workshop up something chronic and wearing my spraying mask is a pain just for doing CNC work. I have seen units like the fog buster and I think I can replicate one myself if I understand the working of them correctly i.e. no atomiser head, but a drip feed into airflow as suggested a few posts back.

Washout
12-01-2014, 06:32 PM
Some pics of the changes to the machine over the weekend, including the new spindle mount, the additional Z clearance and touch probe.
.
Also I have the mixing block for the mister mounted to the right of the spindle and am now awaiting a 1/4" BSP tap so I can hook it up to fluid and air.
.
The axis is chocked up as a safety measure as the more freely move (read better aligned) Z-Axis means there is very little resistance to prevent the axis "falling" when the machine is off, other than the stepper motor "steps" themselves.
.
Hopefully I should be cutting again this week.

Washout
18-01-2014, 11:54 PM
Evening all,
.
A quick pic showing some more progress since my last post:
.
This one shows:
a new and longer air line,
a remount of the "mixer" block so I don't lose any Y axis travel
the coolant/cutting fluid line attached and also a reservoir for the fluid on the top of the gantry at the back
.
The only thing I am missing now is a 1/4" male to 1/8" male adapter for the air line (its current position is "fake" as a "dry run" and the connector there is a mock one from a PC cooling system. Hopefully the adapter arrives soon or I'll have to work out how to setup my lathe for threading (if it can cut BSP - it does metric and imperial).
.
Later
.
.
Chris

mekanik
19-01-2014, 10:24 AM
Screwcutting Programs (http://www.lathes.co.uk/latheparts/page14.html)

Washout
22-01-2014, 10:15 PM
Quick update (sorry no vids yet again)
.
I decided to have another go at GWizard to try some feeds and speeds as there is now a new set of parameters you can plug into the tool for gantry style machines such as mine. I did that for a 6mm Carbide end mill using 2mm DoC and 40% step-over and it suggested 17304 rpm at 793mm/min, which I thought was a bit wacky. However, I have just run a circular pocket climb milling in my usual 6082 Ali which was 10mm deep and 63mm in diameter and it ran like a champ, with just blown air (see note 1. below) and the odd squirt of diluted cutting fluid.
.
That was the good news but I have a couple of frustrations to sort out:
.
1. My misting solution is not working, as air is being blown back up the fluid tube rather than sucking it out - might need to redo or rethink that either by making the injection port for fluid very small (its pretty small as it is) or just dripping fluid out of a separate line into the air flow.
.
2. I have some chatter when cutting in the y axis (x axis passes are beautifully stable) and I have some flex when pushing sideways against the z axis (if that makes sense) i.e. looking at the machine from the end on its like the z axis is trying to roll sideways - I'll have to go over all the bolts around that area and see if anything is loose as something may have loosened or I didn't tighten properly when re-assembling the new z-axis (that chatter wasn't there before I cut the replacement).
.
Oh thanks to mekanik and the tool above I cut some 1/4" BSP threads on my lathe, but even though the SC4 comes with a 127 toothed gear it still took some fiddling on a spreadsheet to work out a close match, as 19TPI doesn't use that gear and the closest I got was 19.05, which was a little tight but a but of brute force "finished" the threads up nicely ;-) Thread cutting was actually quite easy compared to what I remember from doing it at school, as the Seig has a good reverse and the half nut means I didn't have to faff about with a thread dial etc.
.
I have a queue of parts to cut so will make an effort at a video later this week.
.
Later
.
Chris

Washout
25-01-2014, 01:04 AM
Hi All,
.
A couple of vids, as the machine is cutting again. They're not very long as the one was just a facing pass to get the material block to the right thickness and the second one I had to start clearing chips with the air gun (more to be safe now I have decent air).
Sharp eyed amongst you may notice I abandoned the mixing block effort from before and just ran two lines - one brass tube for air and the other PVC for cutting fluid, which just drips into the air flow.
.
The second one was a 2mm DoC using F&S from GWizard, which I was quite happy with and I think I could go deeper. A good sign was scorching hot chips, which I noticed when one fell down the front of my t-shirt, but no discernible heat on the part (I wasn't going to touch the cutter, but trust in my coolant) ;-)
.
I'll try and get some more video once I work out a way of mounting my phone, so I can fuss over the machine whilst recording.
.
Later
.
.
Chris
.

http://youtu.be/FhH2jf8zigk
.

http://youtu.be/0frvDIhey-U

Washout
28-01-2014, 01:51 PM
Gents,
.
I have a problem I could do with some guidance on:
.
Yesterday I cut one half of the job in the video above and all went well.
.
Then I flipped the part and as I am now using a centre hole/bore to zero the part off, I stuck my probe into the collet and jogged over to the hole to zero off of. AT this point my VFD was off, so the only thing running was the machine and my PC. As I have both a probe and touch plate wired from the probe pin I flicked the switch which isolates the touch plate and went into MACH to change the probe to "enable low", which the probe needs to operate.
.
I then started the probing wizard screen in MACH 3 and MACH crashed with an error. OK I thought I've had this before, so restarted the PC loaded MACH and at which point there was a loud hum/fizz and a loud pop. Figuring something electrical had gone I turned everything off and went to diagnose and immediately found that the Optoboard V3 I recently installed looks to have a blown capacitor on it (marked C1) - at least there is a light scorch radiating out of the corner there on the PCB. I also checked all fuses etc. and my wiring on the drivers, just in case something is loose which shouldn't be. Fuses on the toroidal transformer appear to be OK also.
.
Last night therefore, I chucked my old ZP5A-INT board into the control box and this morning started it up and whilst the driver LEDs are green, I have power to the stepper motors, the board doesn't appear to be active and MACH (which I reconfigured for the replacement board) is reporting an External E-stop and try as I might I can't seem to get it to cancel.
.
Things I have tried are:
.
Checked all e-stop/probe wiring, XLR connector (my probe and e-stop share the same 4 core CY
Removed the e-stop xlr connection and switch to take the hardware out of the equation
Tried to disable the e-stop pin in MACH3's input screen - weirdly this saves/applies but then doing a MACH reset trips the e-stop and the pin has been mysteriously re-enabled
.
I'm now stuck, as I can't get MACH3 out of "external e-stop".
.
Any ideas gents? (fully expecting it to be a MACH3 thing I have missed).
.
Cheers
.
.
Chris

Washout
28-01-2014, 02:59 PM
Update:
.
Sometimes I am a plank ;-)
.
I had missed the Active Low setting for the estop - so that's one thing sorted.
.
I also ran over the control box with a decent torch in case I couldn't see previously and had missed a 1Amp fuse on the PSU that had blown - I'll need to order a replacement (sod's law I have every other rating than this one).
.
Of course that doesn't explain why it went pop in the first place, but I have a theory - I noticed that the stepper motors when in holding torque mode normally make a humming noise, which was altering tone when the probe circuit was completed/broken and this was after:
.
a. installing the Optoboard v3 and running a 68V feed from the PSU to power it (my only other voltage feeds in the box without resorting to resistors etc. are 12V and 5V and the Optoboard runs on 15-70V)
b. installing my probe, which should not be shielded at the probe, but as I assumed that and earthed it at the star point in my case
.
I am thinking that somehow either 1. via an earth loop involving the probe shielding (if it is connected to the probe body then my entire machine would also be earthed at that point as well as through the spindle to VFD earth, which is normally the only earthing point for my machine remembering that I have good conductivity from spindle to the bed as everything is metal).
or 2. the 68V feed for the Optoboard from the PSU on top of the 3x68V stepper circuit overloaded it (I'm hoping that the inputs voltage on that board wasn't 68V or I have likely had a lucky escape using the touch probe and plate on that voltage).
.
Of course it could have been something failing on the PSU and that is worrying as it may go again when I replace the fuse.
.
Hopefully I can get this sorted soon as the race car the parts are going into needs to be running by the end of Feb.
.
Later
.
.
Chris

Washout
30-01-2014, 12:01 AM
Quick update and some more video later once Youtube has finished processing.
.
Got my machine back up and running using my old board and luckily the transformer wasn't fried but had strangely blown the 1Amp fuse protecting the 5V output, which wasn't actually connected at the time. Replaced the fuse, clenched sphincter and turned on the power and all was well....phew.
.
I had a funny with MACH3 where the same error/crash occurred when going into the probing wizard and strangely this was repeatable whilst the control box/steppers were powered, but when I tried it unpowered I was able to go into the wizard and then powered up the machine and no further strangeness occurred.
.
The video shows some pockets being cut with ramp in, slot and then pocketing at 2mm DoC. I do have some chatter in the machine still somewhere, which I will try and null out. I have run some cut out paths which in some directions are almost noiseless, but in others are not, so I'll use that as a clue to which axis I have some flex in, but generally the new feeds and speeds, air/misting etc. seem to be working really well.
.
Later
.
.
Chris

Washout
30-01-2014, 12:39 PM
The vid I promised in my last post.
.
I reckon I can go deeper as the machine seems to be coping very well with 6mm Carbide cutters using these feeds and speeds, but as this was a customer part I was playing safe.
.

http://youtu.be/4xanQMtLL-A
.
Oh whilst I remember when I put my old board back in I upped the kernel speed one notch to the manufacturer's recommended setting and the machine seems to be better (less resonance and stutters) than the lowest setting, which I had set it at initially for safety.
.
I think I start looking at a 4th axis now as I have a job I want to run on some thick walled ali piping. Just trying to decide whether to buy a complete unit from Arc euro or just the rotary and make my own stepper mount/coupling - if anyone has done this I'd be happy for any advice/pit falls etc.
.
Cheers
.
.
Chris
.
PS. If anyone knows how to get paragraph spacing into posts on this forum without resorting to periods I'd be very happy.

Wobblycogs
30-01-2014, 01:37 PM
I find the enter key to be quite reliable at spacing paragraphs :biggrin:

11426

Sorry, being serious for a second what happens if you try to leave a blank line between paragraphs? I'm guessing that everything is running together onto one line. Are you by any chance using Linux or some other Unix like operating system. They don't use a carriage return in the new line characters which could be misinterpreted by the editor I suppose.

Washout
30-01-2014, 02:56 PM
Yeah I've tried that (see this post), but for some reason it shrinks out my carriage returns.

Cheers


Chris

JAZZCNC
30-01-2014, 03:17 PM
Yeah I've tried that (see this post), but for some reason it shrinks out my carriage returns.

Cheers


Chris

I had same problem when got new computer that only had Explorer on it then M_C mentioned it was something to do with Explorer and this forum not liking each other, or Lee not sorting it out.!!. . . I changed back to Firefox which I used to use before and every thing worked fine.!!

Wobblycogs
30-01-2014, 05:30 PM
Ah, it's probably this bug (http://www.vbulletin.com/forum/forum/vbulletin-4/vbulletin-4-questions-problems-and-troubleshooting/3963341-ie-10-and-carriage-return-in-editor-4-2-1) then. vBulletin uses CK Editor which apparently doesn't play nicely with IE10. Odd that it's not working correctly as the forum has the appropriate header that should make it work correctly even if you are using IE10. Perhaps try hitting "F5" when you are next on the site just to make sure everything is up to date. Failing that use Chrome ;-)

njhussey
30-01-2014, 06:49 PM
Does this in IE11 too...on other forums when typing a reply it misses letters out unless you turn the HTML editor off.

Washout
30-01-2014, 08:40 PM
Thanks guys,
.
I figured it would be an IE problem - not sure I'm up for a switch to another browser and so long as you chaps don't mind the periods, I'll live with it until/and if Lee fixes it (if possible).
.
BTW - I am trying to find feeds and speeds for a 6 and 4mm Ball nose 2 flute carbide cutters to do finishing passes on some other parts and GWizard is throwing up 24,000rpm as too slow (regardless of the feed speed). I'm thinking that using misting/air should be OK for the tool, but wonder what other's experiences are like when doing ballnose 3D work in aluminium?
.
My Youtube vids have my end mill f/s in the descriptions if that helps and I'm pretty happy with those.
.
Later
.
.
Chris

Washout
17-04-2014, 10:37 AM
Not posted anything on the machine for a while, so here's a video of some woodwork in pine (fence panel uprights):


http://youtu.be/D20CjbAnNVs

Washout
28-05-2015, 06:58 PM
Hi All,
.
Seems like ages since I last posted and probably is.
.
Anyway the machine has not been idle (as you'll see from my channel) and I have recently done a video showing a new addition to its functionality on a fun project, which I can show publically (I figured milling a modular RIS conversion for my HW97K air rifle might raise a few eyebrows ;) ). I'm going to try and convert the rotary for CNC, as I have everything apart from the motor mount and pulley/belt if it needs it and I can't get away with direct drive:
.

https://www.youtube.com/watch?v=H5t3Vv0bmQg

routercnc
28-05-2015, 08:15 PM
Hi Washout,

Good to see some machining and an interesting way to get a fourth axis into the table.
Noticed that your X axis syncronising belt looked a bit loose by the way . . .

Washout
28-05-2015, 08:58 PM
Hi Washout,

Good to see some machining and an interesting way to get a fourth axis into the table.
Noticed that your X axis syncronising belt looked a bit loose by the way . . .
.
Thanks - it seemed like a good way of getting the thing mounted without faffing about each time. Especially as the mounting holes in the chuck of course didn't line up with the table slots (sod's law at work as usual). I could have put the additional piece of extrusion on top of the existing bed, but I'd have lost a lot of Z clearance that way.
.
Re: the belt - thanks for that as well - it needs a bit of a tighten and its normally a little slack but the tensioning pulley has obviously moved a bit through use - probably the higher speeds I've been using on the acrylic in the other videos.
.
I'm also cutting the logo in the acrylic into the other side of the PC case, which is steel and will be the first time I've tried that (deliberately at least anyway ;) ).

Jonathan
28-05-2015, 11:27 PM
Nice start. There are some posts somewhere on this forum where I motorized mine and others rotary tables. Have to take some care to get low backlash, but it's do-able.

I made something yesterday, then noticed it looks a bit like a light-saber:

http://www.mycncuk.com/attachment.php?attachmentid=15432&stc=1

Can anyone guess what it actually is ...

Washout
28-05-2015, 11:43 PM
Hi Jonathan,

No idea what that is - are you selling them to a less than desirable government's internal affairs ministry? ;)

Clive S
28-05-2015, 11:59 PM
It's a wimwam to wind the sun up:stupid:

Jonathan
29-05-2015, 12:10 AM
No idea what that is - are you selling them to a less than desirable government's internal affairs ministry? ;)

Not quite! Clue - I made it for free and it will definitely be used next week.

JAZZCNC
29-05-2015, 01:54 AM
Not quite! Clue - I made it for free and it will definitely be used next week.

Bet it's going on a Boat for a blast around some very very scary roads.???

Jonathan
29-05-2015, 05:03 AM
Bet it's going on a Boat for a blast around some very very scary roads.???

Something like that ... I just finished working on the 'boat' for tonight, so 2 hours sleep before recommencing tomorrow!

Washout
07-06-2015, 11:03 AM
A first for my machine a few days ago gents.
.
Got it to cut steel sheet (very slowly) - props to GWizard for the feeds and speeds:
.

https://www.youtube.com/watch?v=32GZe6e7-SY