PDA

View Full Version : BUILD LOG: Tiddy CNC.



TonyD
05-11-2012, 12:03 AM
Hello.
Naming the machine seems the thing to do so Iíll call this one Tiddy after my sons favourite teddy :smile:

Iím creating the basics of my electrical control system while aspects of the mechanicals of my machine are still in the design stage. Iíve looked at several build logs and the required wiring diagrams for my components. I would really appreciate it if some folks would help me validate my assumptions and help with a couple of questions.

Background.
The hardware I have purchased should see a machine with a 1.5M X Axis, 1.1M Y Axis and a 0.2M Z Axis. I will be using a total of 4 Stepper Motors. My basic electrical bill of materials purchased so far is:
Power:
1 x PS806-5 PSU (3x68V-6A, 1x5V-1A).
1 x PS408-12 PSE (3x36V-3A, 1x12V-1A)
Steppers:
4x SY60STH88 Ė Nema 23 Ė Anticipating wiring in parallel (2.73V-4.2A)
Drivers:
3 x AM882
1 x PM752 (Z Axis)
BOB:
ZPA5A-INT 5 Axis breakout board.
Plus PLCM-E3 Mach 3 Ethernet Interface

730773087309

Component rationale:
Iím not an electrician and although I could see from the forum comments it was possible to build my own PSU I didnít want to and so bought the higher output unit for the X and Y Axis and then the smaller (and cheaper) PS408 for the Z axis, BOB feed and spare capacity. I did not want to stretch my budget to 4xAM882ís so settled for 3 and the PM752 for the smallest Axis. I liked the ZPA5A-INT 5 and purchased the PLCM-E3 because I wanted to run the machine from a lower power and spec PC. Iím not convinced of the PLCM-E3 yet.. So these are the choices I have made and from this basis and looking at the photos can I check these assumptions please.

1) I am bringing mains into the panel and switching the Live through a 10A 600V toggle switch, in order to show the mains in live I have wired a Neon light with appropriate resistor in parallel with the switch. The circuit then continues on to a 10A 20mm fuse in a fuse holder.
Assumption, have I wired the Neon correctly? Is 10A the right choice of fuse for this stage of power distribution? Is a toggle switch rated for 10A sufficient?

2) From the panel the mains will continue to the main board via a 30A Ďchocolate blockí connector so I can remove the panel from the enclosure for maintenance/installation etc. Mains power is then sent in parallel to both PSUís.
Assumption: Iím assuming this is ok.

3. From the PSUís Iím using 22AWG wire to distribute the live to each stepper driver via a 6.3A fuse for each driver. Iím using Ďspadeí type connectors rather than soldering these connections.
Assumption: Is this size of Fuse OK, the steppers should draw no more than 6A so I thought this would be a good size to choose. I believe the wire size is good for 11A 600v so Iím assuming this is OK for the HV (but not mains side) of the PSUís.

From this point on, assuming I fix any identified issues, I plan to:
A: Connect the HV output side of the stepper drivers to the enclosure panel sockets with my CY cable (detail below)
B: Connect the BOB to the stepper drivers
C: Generally tidy the wiring up.

Iím looking at buying shrouded (but not shielded) cable for the signal connections between the BOB and Stepper derivers.. but I do have some CAT5 Ethernet cable around.

4. Can I use CAT 5 cable within the enclosure for the signal side of the steppers to the BOB?

So far I havenít attached a component to the board that requires an earth connection (other than the mains-in) but I have a connector block set up as my central earth point.

5. Have I missed anything that requires connection to the Earth connector within the case so far? (based on what you can see in the photos)

6. I am assuming that the Earth connector on my board will go back to the Earth connector on the mains-in socket?

My panel has connectors for the steppers, limit switches and some potential expansion. My plan with these is to make up small junction boxes on each stepper so that I can unplug the motors from the cable, hopefully this will make the connections tidy and maintenance/installation easier. I have a small reel of CY cable at 4x1.5MM2 cores for the connection between the motors and the enclosure.

7. I am assuming that I should somehow Earth the CY cable shield at the enclosure panel. I am confused how to do this, should I just connect an Earth line to the panel that houses the connector sockets (Same one the mains in and on/off switch is on)?

Out of scope for now.. I have the ubiquitous Chinese water cooled spindle and VFD, for now Iím happy for these to remain out of scope and on their own mains lead supply.

8. I am assuming I can do this and effectively use the VFD ĎManuallyí without any need for electrical control?

Finally, I realise there is no E-Stop in the plan right now nor Relays used to switch the mains. I have not been able to find good material or circuit diagrams on how to wire this up. Right now I am willing to use the on/off switch I have to get the unit running and Ďupgradeí to an Estop later.

9. If the enclosure must have an Estop and/or relay switching the inbound mains, can anyone point me towards an easy to follow circuit diagram?

Thanks for taking the time to read and for any advice.

audioandy
05-11-2012, 09:53 AM
Hi Tony

If I were you I would implement a E-Stop circuit from the word go, when you are new to CNC which I assume you are, you will inevitably make mistakes if you have a nice BIG red e-stop button you can hit in the blink of an eye it will pay dividends in the long run.

Do not under estimate the power of these machines your cutting tool will cut through skin and bone with ease.

I have used a Pilz safety relay on my machine which kills everything safely these are fairly easy to wire up but not cheap!

Regards

Andy

This would do the trick (eBay No:- 281014635581) I could give you a diagram to help fit it.

TonyD
05-11-2012, 02:26 PM
Thanks Andy, I'll PM you for the diagram.

Tenson
06-11-2012, 05:25 PM
I read it as 'Tidy CNC'.

TonyD
06-11-2012, 08:04 PM
Tidy it aint currently, working on tidying up the wiring tonight. Not that I have confidence it wont create a fireworks display yet.

TonyD
16-02-2013, 01:52 PM
OK, I've just about finished the control system version 2.0. As Andy and others said what I had may have worked in so far as it drove motors but it was not safe, had no estop chain and had other material weaknesses. Thanks to AudioAndy, Jazz, Adil and others who helped me on the Electrical. For the sake of anyone who reads the above with the same issues, I'm going to answer my own questions based on what I've learnt so far, then to the next phase, the mechanicals... :)

1. No the neon should be in parallel with the live and neutral of the mains inbound, the way it was configured to start with the neon would only light upon a general closed circuit in the rest of the control box. The purpose of the neon is to show a live mains so in parallel to the socket pins works best for this.

2. Didnít need to be this big, 15A will suffice. Also I discovered these in various sizes and current ratings and they helped move power and signal around the control enclosure. CamdenBoss Terminal Strip (Male+Female) 8mm 12 Way 6A (http://www.rapidonline.com/Cables-Connectors/CamdenBoss-Terminal-Strip-Male-Female-8mm-12-Way-6A-21-2496)3. Ditched fuses in the end and went for DIN Rail based circuit breakers on the advice of forum members, less fiddly, more robust. The wire was OK but having terminated using the spades I ended up stripping the ends and creating a neat soldered end due to several instances where wires worked loose.

4. Didnít need to, I just used normal 22AWG wire but as the runs were so short, (4 inches) there was no need for the shroud. Also there are only actually 3 signal wires needed for each stepper driver.

5. Did that but few components needed it.

6. Did that, made sure I earthed the shielding from the motor power wires to it.

6a. Used epoxy to create small junction boxes on each motor to tidy the wiring and create a robust connector to make assembly and maintenance easier.

7. Earth configured as described. However used the pilz and safety contactor to also switch power to the VFD via the control box, added a separate mains in on the control box to facilitate this.

8. Seems to be the case but have not go there yet.

9. Jazz and AudioAndy stepped in a helped with a pilz relay and very detailed instructions on creating the safety circuitry as well as phone conversations to assist - thanks guys.

Other lessons learnt so far....

A) I wasnít happy with the 752 and AM882mix so I swapped the 752 for another AM882 so that all four motors are matched with the same drivers and for the same reason swapped the PS408 out for another PS806 so that there are now two of those. Thanks to Gary at Zapp. PS806-5 Linear power supply. (http://www.zappautomation.co.uk/en/power-supplies/438-ps806-5-linear-power-supply.html)

B) Enclosure. I found them expensive and bought my first one based upon the design of the controls before the significant change to DIN rail, circuit breakers and the safety circuitry. As a result the enclosure was way too small to contain the components neatly, rather than waste the money spent on the original box I just used the lid to support the stepper drivers, BOB and smooth stepper, it actually made the wiring runs for these components small and relatively neat.

C) Heat. Slightly concerned the enclosure may get hot so I've added some small fans, they are only 5v DC though and in use not surprisingly producing a small airflow. If I do detect heat build-up as I progress, I'll swap them out for mains powered fans.

D) Smooth stepper. From reading various comments about the use of a smooth stepper I realise that itís not a generally popular choice. For me though it brought one large advantage in that it allows Mach3 to run on my Windows 7 64bit OS. I didnít want to install a separate 32 but partition or run something like Mach from within a virtual machine so it solved a constraint for me. In time I may see what other benefits the board (Purelogic PLCM E3) can bring.

I'll attach some pictures of the electrical journey so far later.

JAZZCNC
16-02-2013, 03:18 PM
D) Smooth stepper. From reading various comments about the use of a smooth stepper I realise that itís not a generally popular choice. For me though it brought one large advantage in that it allows Mach3 to run on my Windows 7 64bit OS. I didnít want to install a separate 32 but partition or run something like Mach from within a virtual machine so it solved a constraint for me. In time I may see what other benefits the board (Purelogic PLCM E3) can bring.

Tony SS is mostly not popular because folks are tight arse's.!! Those that have them generally love them and see better or more stable performance.

If yours is the USB version then power from separate 5v supply and you won't have any troubles. If you use USB power it can and will lockup at times.!

Regards the PLCM E3 then I have one and other than ethernet then there is very little difference, it allows ever so slightly higher tuning but nothing i'd swap the USB SS for. Ence why SS still runs my main machine.!

TonyD
21-02-2013, 11:25 PM
Different question. Into the much loved world of mechanical and forces. :)

I'm planning on using a Y axis gantry thats made out of a 4 pieces of 15mm Ali plate, the front part has a slot as per previous discussions to allow the moving Z Axis to connect to the Y Axis screw.

Key question on my mind right now is the length of the block which will support the gantry onto the rails relative to the center point of the spindle.
As shown below, it feels right to me that the gantry support block should at least extend to the spindle center point, no maths to back this up but without this dimension it feels to me like the gantry would be out of balance relative to the forces acting on the spindle. Any comments? Only concern is that the block is 260mm in length right now and thats a fair amount of X Axis travel I'm eating up. Naturally the klunky design of my Z Axis, which was made with a press drill only is not very compact and thus adds to the problem. I'm comfortable with the design as is as job #1 when Im competent with the machine will be to make a new better Z Axis but thats for the future. Comments welcome. Tony.
PS. I know there's going to be a side plate etc, I'm just looking at this one component now, happy to take feedback an anything else though.
8257

Jonathan
22-02-2013, 12:29 AM
Only concern is that the block is 260mm in length right now and thats a fair amount of X Axis travel I'm eating up.

The deflection due to the bearings on each axis is inversely proportional to the linear bearing spacing squared (which is obvious from the formulas in the Hiwin linear guide datasheet), so increasing the spacing by a relatively small amount can make a big difference. That's why it's important to space the bearings out as much as you can afford to, assuming the part they are mounted on is strong - which currently isn't the case as it's just a plate which would benefit a lot from a triangular/trapezoidal piece on the side. 260mm is not bad...probably not much point going above 300mm as other parts of the design will have a greater effect on the overall stiffness.

TonyD
22-02-2013, 11:02 PM
Thanks Jonathan. I thought 260 was a lot as it was but happy to leave it there, if as I suspect this is the basic design I go for.

TonyD
22-02-2013, 11:18 PM
Folks. I've been wondering about some more design concepts and I thought I'd ask for some advice. I should start out by saying that my goal is to orient the machine at a relatively steep angle but as I've played around with frame concepts I have no idea how to approach that yet. I've found that one step at a time suits this pursuit in my mind…So I'm hoping for some vertical but may go for regular horizontal. The basic parameters are a 1500mm X, 1100mm Y and currently a Z with a conservative 150mm, plenty scope for changing the Z once I can make some parts.

So. I have considered three basic options and presented them from a front view and my basic question is how to place the rails and screw so that the Y Axis gantry is best supported with minimal adverse forces being introduced?

Option A is the most conventional I think? Option B has the virtue that the screws are under the machine, need to take more care over a frame that allows this. Option C seems to put a lot of force on one side of the linear bearing block. Any suggestions, comments welcome. Sorry it’s not in CAD but I'll flesh out the winner into a more detailed model later. Thanks. Tony.
825882598260

TonyD
24-09-2013, 01:10 AM
10215102161021710218Next Steps.

Hello all. I've made some progress lately. The control system is nearly finished (bar some limit & home switch mounting), and tested on The Z Axis which is just about there. The Y is nearly there, just need to mount the stepper which leads me to the frame. I have looked at many frame's and tried to come up with my own. There were three features I was looking for

i) I wanted the frame to be robust enough to support cutting aluminium
ii) I wanted to be extend the rails over one end for over sized jobs (thanks jazz)
iii) Thinking of the over size issue, I wanted the frame to be tall enough that I could mill out the lock on a door.

The machine will be inside so I'm looking for welded frame sides with bolted cross members. I also decided I didnt want the hassle of a moveable bed, frankly I want to get the thing built (this year), so left 110mm of Z travel from the tip of the spindle, I think thats enough for general use? I also decided at this stage to have two rails (in orange) welded to the sides as support for the drop-in welded bed frame (purple). With the exception of the top rails, everything else is 50x50x3. I havent costed it yet, will probabily have to revise on that basis alone but wondered if there was any feedback on the design so far. Thanks.

Boyan Silyavski
24-09-2013, 11:06 PM
Just a note there about the Estop button: Buy a brand one. The cheap chinese ones have weak springs and stick sometimes right when you need it. I have el cheapo and it betrayed me a couple of times. For my new build i bought Japanese and the difference is huge. believe me, you will need an Estop button

TonyD
25-09-2013, 07:59 PM
10227I spent some time today calculating the cost of the steel. Not sure its a useful metric (for discussion) but there are just over 30M in steel box in the design. I was quite dismayed to start checking on the web for pricing only to discover several sites quoting around £350, which was more than I had in mind. Then I came across a company called ISG Steel who routed my call to the nearest depot (I live in Berkshire). Thier quote on a like for like quantity of 50x50x3 steel box section was around half the websites I'd found up to that point, also they were friendly and when I mentioned I needed more than just the one length offered me a discount. From then on in I looked at how to cut the required lengths from 7.5M stock and came out at 4.2 lengths, which was dissapointing. I modified the design to use small 5cm lengths of steel to support the bed on each of the 5 side member points, instead of a continuous rail (Orange in the picture above) and that nicely got me to all the steel in 4x7.5M lengths. Once I asked for a quote including the top rail material of 90x50x3 the whole thing with shipping came to £220. It seems a fair price and much better than many of the websites I found.

JAZZCNC
25-09-2013, 08:11 PM
Tony I pay around £26-30 depending on steel prices at that time per 7.5mtr of 50x50x3. This is from local steel supplier so I'd checkout any local steel stock holders because that 90x50x3 seems to be costing quite a bit. My local supplier delivers free because I'm just up road but he only charges £15 for 20mile radius and i'm sure other steel stockist's will be similar.!

Boyan Silyavski
25-09-2013, 10:27 PM
10215102161021710218Next Steps.

Hello all. I've made some progress lately. The control system is nearly finished (bar some limit & home switch mounting), and tested on The Z Axis which is just about there. The Y is nearly there, just need to mount the stepper which leads me to the frame. I have looked at many frame's and tried to come up with my own. There were three features I was looking for

i) I wanted the frame to be robust enough to support cutting aluminium
ii) I wanted to be extend the rails over one end for over sized jobs (thanks jazz)
iii) Thinking of the over size issue, I wanted the frame to be tall enough that I could mill out the lock on a door.

The machine will be inside so I'm looking for welded frame sides with bolted cross members. I also decided I didnt want the hassle of a moveable bed, frankly I want to get the thing built (this year), so left 110mm of Z travel from the tip of the spindle, I think thats enough for general use? I also decided at this stage to have two rails (in orange) welded to the sides as support for the drop-in welded bed frame (purple). With the exception of the top rails, everything else is 50x50x3. I havent costed it yet, will probabily have to revise on that basis alone but wondered if there was any feedback on the design so far. Thanks.

Hi Tony,

I am currently doing my first real build so take my word with a grain of salt but i spend a lot of time considering, redesigning, calculating and so on...

So looking at your idea i could not help but share my thoughts, even if they are quite influenced of my current design. I hope it helps though. Sorry, i dont mean to hijack your thread

What i don't like at your idea is that the diagonals actually don't reinforce the profile where the rails are.

I went on my design from 60x60x4 to 100x100x3 as i was faced by the same problem as you, great lengths and a lot of cutting and soldering.

So here i just twisted a bit my design to accommodate it to your measures and needs. Much less rails, much stronger, much simpler to execute and cheaper, as much less weight. I would like to hear what others say but i am quite sure is strong enough for routing aluminum, not scratching.

All in all there are 17.8 meters of 100x100x3 x 9kg per meter = 162 kg

10228

10229


My thoughts about bead and gantry:

I made my gantry design to be from 2 pieces of 100x100x3 soldered together, just to have in mind when considering the following.

For me to build a machine and later to upgrade it is not OK, i mean what you say for your Z depth. Please take a look below for my solution to this.
My Z plate is 500mm long. My useful Z travel is 200mm, but i will limit it to 180mm. However the idea is the bead to be deep but over it to have a cheap wooden structure 100mm high to serve as a removable table. So now the working area is raised and there the aluminum or normal jobs will be cut. The table can be made from a couple of MDF sheets glued together or whatever scrap. Even from ribbed inside wooden structure to weight less and easy removal.

So with everything soldered i would have quite versatile machine/ in fact its not for me the machine/ .


Of course you can add 2 x 1600mm rails and make the bed removable, but i don't see a reason for it as you can remove the wood bed instead and keep everything much stronger. I believe on my design there is no need to add anything or change anything, as even it looks simple is thought to withstand forces from all directions. The only thing that comes to my mind that can be bettered is to place everything at asymmetrical distances to avoid frame vibrations further

10230

If you use Sketchup i will be happy to send you the file with the table i draw with your dimensions so you can play with it.

I hope that will give you some ideas .

PS. One key part of that design that i forgot to mention is that there are plates that fix the rails supporting beams. I will lasercut them, but they can easyly be fashioned from something else. They play the role of not permitting to the beams to vibrate left right, looked from front and also seal the profile end.

10234

TonyD
26-09-2013, 04:29 AM
Silyavaski. Thanks for taking the time to modify your model with the dimensions I am looking for. I have to say your frame design does look so much more solid than the one I designed. The one one constraint I have is that I need to move the machine in the house so I cant have it all welded. I can see though that if I made the cross members you have proposed as bolted rather than welded, this might work. I'll have another look at the Z height and see what that means for the rail height. Thanks again. Tony.

CraftyGeek
26-09-2013, 07:44 AM
Tony - it may be worth seeing if you have another local steel yard near you. My local supplier charge £4.1 +VAT per meter for 50x50x3 (£147.60 for 30m) as a base price & charge less than that if you order a reasonable quantity. I also get them to cut some of it to size for me - depending on quantities, they charge 50p per cut or less...this means I can fit it in my car & pick it up myself removing the delivery charge as well as saving me time cutting stock to size.

Boyan Silyavski
26-09-2013, 07:49 PM
You guys are lucky to live in UK. Here in Spain i think in kg. Roughly any profile costs 1kg=1euro at the big warehouses. Unfortunately no near where i live and they don't cut to size. As i don't have a band saw i have to order from the local metal shop. If you befriend them they cut to size and sell it for 1.50 the kg with the work. if you don't know them they sell it for 2euro per kg and you pay the cut extra and that if you are lucky that they pay you any attention...

Hence my designs are influenced by the KG and by the 6 meters length. Guess who i will be bringing some beer tomorrow:beer:

TonyD
26-09-2013, 10:11 PM
1024210243I've been caught by the CNC bug again :)
Silyavski emailed me a model in Sketchup to play with which convinced me to ditch Visio for my drawing and start again in Sketchup. As I tried to explain to my wife "the guys on the forum made it stronger, cheaper and lighter", which is I guess the power of good design. Trying to manage the constraints of cost and size I ended up going for 80x80x3 due to an inherent constraint with my Y Axis (which is built), which then left me trying to get the frame out of only two lengths of 7.5M stock, which led me in the end to cut the legs down to 350MM from 500MM. I phoned around today and a local Steel supplier quoted me £50 ex Vat per 7.5M length of stock. Now that its down to just £138 with VAT and delivery I'm happy. Bonus is the machine can be carried through a doorway so no bolting. So thats my current plan. Thanks for all the input, any other comments welcome.

Boyan Silyavski
26-09-2013, 10:30 PM
Hi Tony,
i forgot to tell you but you went to exactly the same conclusion like me. The 80x80x3 is the size of profile that is at the golden middle point for weight and strength and value, also permitting this type of design. Great that you figured it on your own.

Today i went after work to the lazy guys at the metal shop and urged them for my profile. After helping them to cut i figured one point more. Its not a real necessity though. As i have ordered 1x 6m new profile and used some scrap they had, there was i piece of 100x100x4. So i figured at the last moment that the 80mm pieces could be even stronger if made from that. Ask them if they have by chance some scrap 80x80x4 just for the 80mm pieces. It will not raise the price. Just give it a try.

Hey, now i want small legs for the machine i am building :hysterical:

JAZZCNC
26-09-2013, 11:56 PM
Tony The design would benefit from one more support for bed has the centre distance is on the wide side and could cause bed flex.
It also definitely needs some triangle braces in the corners to help hold frame square.

TonyD
27-09-2013, 08:07 AM
Thanks Jazz. I skimped on the extra rail in the design to keep the Steel costs down to two lengths. What I'll do is find some smaller sized box section for the central bed support, put two or three support rails in there (depending on size) and use the existing single span of 80x80x3 to create the support triangles at one end. I guess at this point I could go for triangles on the 'fixed' end or just move the central bed support rail to the same height as the top rails to create a U shape. I'll draw them up shortly and ask for comments.

TonyD
01-10-2013, 06:48 PM
I have to admit that I'm not planning on cutting or welding the frame for my machine myself. I dont have the kit, or a suitable space to make the frame. So I'm off to see a recommended chap locally who has agreed to cut, weld and paint as per the attached. Not sure when it will be done yet, should confirm things by the end of the week then its just a matter of hauling 136KG's of frame to my study up the stairs.. :)

Boyan Silyavski
01-10-2013, 09:22 PM
Wise decision if you don't have the means. I welded mine with my new welder this weekend. I fiddled a lot, as my first real weld and had to wait for a friend to take it of the welding supports and bring it inside the garage. You will be surprised of the robustness of the design. Dont forget what i suggested you in the email, spend 5-10 pounds more and weld vertically a 5mm or similar whatever flat bar plate at least along the middle profile of the table bed for vertical reinforcement. If it was me, i would weld on the 3 of them. Just for the sake of extreme robustness. Of course, depends whatever you intend to use as table bed. Also these bars can be bolted later if needed. However with them you will have astonishingly minimal overall deflection/less than 0.05mm i believe/

I don't know about your gantry and Z but with this you can do aluminum on the breeze



Whatever, looks very nice!

TonyD
14-10-2013, 07:40 PM
Checked in on the chap who is building the steel frame earlier today and was really pleased that its coming together. Some of the welds right now are just tacked and there's end caps for the box as well as adjustable feet to add yet. Its great to see something come from CAD into reality. I should collect it later this week and then it will be time to mount the other hardware I have got and the Y and Z that I was able to build myself with the trusty drill-press and hand tools. I realise that the mouting of the stepper to the side of the Y Axis side plate is sub optimal but I guess its part of the learning.

TonyD
18-10-2013, 03:29 PM
Step closer today. Got the frame up the stairs with a lot of help. (The floors are concrete - so should be able to take the weight). Travelling a bit with work for a few weeks now but hopeful for a Christmas live-run. Still got the mechanical X movement, limits, echain, wiring etc to complete.

GEOFFREY
18-10-2013, 07:57 PM
That looks very nice Tony, very neat. G.

audioandy
25-10-2013, 12:03 AM
Looking good Tony

Boyan Silyavski
26-10-2013, 05:19 PM
Very nice to see a drawing in real life.

What made you choose this size table? I wonder as i have to start urgently my other build and am contemplating something similar.

TonyD
26-10-2013, 09:15 PM
As a saw the frame take shape I did find it amazing to see CAD tranform into a real (and heavy) object, it felt great to see, thanks for your help with that Silyavski. Originally I wanted to design the machine to fit an alcove in my study but as I explained what I was doing to friends and family, someone said "You could cut the parts for a kitchen [cupboard chassis and doors] I want to build" so that became a goal. If I were starting again I'd design it to cut 8x4 foot plywood as the Wikihouse project facinates me. I dont really have the space for a machine that big really so maybe if we move house. :) I mounted the Z Axis on the Y last week and fitted that to the linear rails, its getting close now. I've got some bearings on the way from Chai and then I'll be able to mount the X Axis screws and start some movement; I will get the limit switches working first though, last time I played with the Z Axis I crushed a limit switch that wasnt wired in yet!!

Skysurfer
28-03-2014, 09:55 AM
Tony, I'm heading into this size machine, any chance you could throw those sketches my way? I haven't even started, yet, so I'm trying to figure out what will suit me best (size wise). I have a load of steel in my shop at the mo, and I'm trying to use it all ('cos I'm tight;-)
Cheers,
Pete

mekanik
28-03-2014, 12:06 PM
Skysurfer
Post 24 has the required info
Mike

TonyD
28-03-2014, 01:11 PM
Skysurfer,If you cant find what you're looking for, let me know.
I have made progress last few months, I'll post some photos this weekend, maybe a video even. I'm close to turning on now. :)

Skysurfer
28-03-2014, 03:45 PM
Cheers, Mike & TonyD, will take a look. I'm still playing with my Zen 7x7 :courage:
Pete

GEOFFREY
29-03-2014, 01:19 PM
Cheers, Mike & TonyD, will take a look. I'm still playing with my Zen 7x7 :courage:
Pete

I wish I had 7" to play with! G.

TonyD
30-03-2014, 02:32 PM
Current Status:
I'm nearly there as far as the initial build goes. I've tested the safety circuit & limits now but not powered on the drives and motors lately, although they were spinning many months ago. Learning a lot along the way, lots to try differently in version 2.0, more on that when its working. Right now am waiting for a PC from Ebay that I can dedicate to running the machine so should be able to run the smooth stepper and BOB up fully in a week or so and once the Home switches work I can start to join all the dots together and get some movement. Enjoy the photos (When I see a way to enter them).

TonyD
30-03-2014, 03:55 PM
1194111940119421194311944

GEOFFREY
30-03-2014, 07:34 PM
A nice neat looking unit Tony. This site really does seem to have members that do some lovely work. Well done. G.

TonyD
18-04-2014, 12:57 PM
Hello.

This is with reference my build http://www.mycncuk.com/forums/gantry-router-build-logs/5302-tiddy-cnc.html

I have slowly configured the components and tested at each step along the way. I have the e/stop circuitry killing the main PSU's (Thanks AudioAndy and Dean), the smoothstepper and BOB working, home switches trigger the Mach3 Diagnostics page and all of the AM882's have driven thier motors. So far so good..

This morning as I was jogging the machine around (thankfully slowly) I noticed it start to rack on the X Axis, only a little, maybe 2-4mm then I killed the power & released the mechanical stress by uncoupling one of the X Axis connecting plate and looked closer.

I couldnt reproduce the fault until a few mins ago when I noticed that one of the 882's went into an error state and stopped the motor with 4 blinks which per the manual is 'phase error protection' (short/break).

So my plan is to replace the cabling between the 882 and the control box socket, I'm going to pay particular attention to the soldering and shrink wrap on the connector. The reason I suspect this the most is that I noticed that as I opened the door of the electrical cabinet a little wider the X Axis motor in question made a little noise like a non continous jog so I assumed the mechanical action of the wires moving are causing the short/break within the cabinet.

My question really is, if this is intermittent how will I know when I've fixed it and what am I missing? oh.. and yes I'm now a fan of Wobblycogs alarm circuit and will look to impliment that sharpish! Thanks for any advice. Tony.

m_c
18-04-2014, 01:31 PM
Intermittent faults are a pain. Being a vehicle master tech, I've spent weeks (and occasionally several recovery trucks :whistle:) trying to find some faults.
Only way to know for sure is to try it.

Personally, I wouldn't solder wiring. Solder and wiring that moves/vibrates has a high risk of breaking causing intermittent problems. Connectiosn should ideally be crimp, or screw with suitable support to avoid movement at the join.

EddyCurrent
18-04-2014, 08:52 PM
Maybe this is worth reading too ? it#s AM882's and SmoothStepper http://www.mycncuk.com/forums/gantry-router-build-logs/6565-ready-steady-eddy-27.html#post56967
although I didn't get an alarm on the AM882 so your problem is likely something different.

TonyD
18-04-2014, 11:10 PM
Thanks folks. I checked the wiring earlier and found a couple of loose terminal connections to the drivers, cant say for sure but one of them was quite suspect. In addition the driver alarm circuit is in and tested, I was quite motivated to put that in! :)

JAZZCNC
19-04-2014, 12:19 AM
I couldnt reproduce the fault until a few mins ago when I noticed that one of the 882's went into an error state and stopped the motor with 4 blinks which per the manual is 'phase error protection' (short/break).

Now arent you pleased you bought good quality drives because lesser drives would have let out the magic smoke when a phase wire falls out of bounces.!!

Now the missed steps concern me because while a loose connection could cause loss of position it wouldn't loose 2-4mm just jogging around, which by the way isn't a small amount when your talking several hundred MS steps per mm.

I would check that the SS frequency is same for Each Axis and it's not set to some silly high setting. 128Khz is more than enough.

Also do you have a Cable to connect to AM882 drives and use the protune software.? If not then get one for the sake of £2 worth every penny.
Check that each drive is set exactly the same with regards microstepping and Amp's. Protune as a Quirk in that it will some times default the Amp settings back to default which is around 2A, so check it hasn't done this on one of the drives.

Makes sure the Motor tuning and steps Per in Mach is exactly the same for both slaved Axis. Don't set high feed or Acceleration until your sure machines mechanicly fine and no binding Etc. Even then be careful not to go daft with Accel which is the biggest cause of lost steps.
What are your Motor tuning settings now.?

TonyD
19-04-2014, 12:46 AM
Hello Dean. That would have been all I needed to see smoke! Its great building the thing but setting it up is nerve racking (no pun intended) :) I watched the youtube videos on Mach3 homing *a lot* before I finally hit the button, thankfully Z and Y homes nicely. X groaned too much so I stopped to investigate.

I do have the RS232 cable, tired 3 pc's before I could get comms to the drives but I did get there, havent run the protune software yet, will add that to the list for tomorrow. I'm assuming that the racking was becuase one motor was running, the other wasnt.

Currently, kernel speed (from ports and pins screen) is 35KHz, Motor Slaving & pins appears fine. 3 home switches fine, estop circuits pilz etc fine. I cant see where to check the SS frequency, in the plug in control screen I just get mach3 buffer time, 1000ms, PLCM buffer 500, accell overshoot 20%, plcm feedrate for DRO 1075.

I'm using ballscrews with 5mm pitch. For X,Y,B motor settings are 320 steps, velocity 4000, acceleration 1000, step pulse 5, dir pulse 5
For z its 320,3000,300,5,5.

I'll definately run protune tomorrow, thanks for the advice. Before the gantry was bolted to the ballscrews it run quite well, there's a lot of mass there so it wasnt really easy to push but could move it about easy enough. That leads me to conclude that the linear rails are perpendicular. Now its bolted to the ball screws when I release the left hand side binding the right motors is really more stiff to rotate with no power to the drives than the left, so I'm looking at the bearing mounting for the ball screw on the left side of the machine, I may be able to shim it a little, I'll take more detailed measurements tomorrow and see if I can determine what the resistance/binding is... All advice, as always, appreciated :) thanks.

JAZZCNC
19-04-2014, 01:44 AM
Yep just a motor wire to fallout while Jogging would fry lots of cheaper drives.!

Are you using the USB SS or ESS.? in either case they don't use Kernal speed but it's best to set it at default 25Khz.

If using ESS then look in Plug in control and choose ESS Config. On left you'll see Max Step Frequency for each axis and Spindle. Set to 128Khz you won't need any higher than that.

TonyD
19-04-2014, 11:05 PM
I will try to consolidate these issues under my build log later but a seperate post allows me to appeal to those who know and love ProTune directly.

So for reference I have an EPLC3 SS and ZP5A-INT Zapp BOB and 4 AM882 drives. Following on from my adventures in racking yesterday I'm trying to tune the motors with ProTune and... it really is trying.

I have 'the cable' (DB9 to Rj12), also a USB to RS232 cable with seperate driver (CH41?) and so far 4 different PC's trying to get stable comms to the 882.
I ditched the USB to RS232 lead completely and cut off the DB9 connector from 'the cable' checking in the pinouts from RJ12 connector (using the data sheet from the Leadshine site) so as to re attach it from new, I didnt find any misaligned pins in the DB9. I've used 4 PC's to connect and generally I'm not getting anywhere.

I went to the Leadshine site and downloaded the correct version of the AM882 Protune (according to them) and found it didnt play nice with my legacy XP (up to date as of the April cut-off) and found myself re registering several ocx components at which point I thought, enough, ask for directions!

For what its worth I do get a read from the drives so I know that serial comms are working but every time I try something interesting in ProTune it says in effect comms dropped. I've tried all the dip switches on the 882 to off, and not (although not all on), the 882 is powered up and motor is connected to it, the BOB is not.

I'll send real chocolate Egg to whoever can help me out here. This is driving me nuts. Thanks. Tony.

Clive S
20-04-2014, 08:40 AM
Tony. I have not used Protune as yet but have you set the com port up on the PC with the correct baud rate, stopbits and type of handshaking ie xon/xoff etc. Just a thought, I am not sure the usb to rs232 don't always work ..Clive

m_c
20-04-2014, 12:57 PM
I've got no expereince of digital stepper drivers, but I've never had any issues with USB adapters setting up other drives. Could it be something like the USB/RS232 interface you're using not implementing the full RS232 standard?

Not all USB adapters connect or implement things like the DTR/DSR/RTS/CTS lines, which could be the problem.
I normally use FTDI based adapters, and any I've had implement the full pin set, however some do exist that don't.

TonyD
20-04-2014, 04:10 PM
Well, I've spent hours rebuilding PC(s) to Windows 7, finding pinouts on the web (leadshine's site) and I'm still no better off. Added to which is the frustration that I feel that whenever I'm able to some quality time on the machine its a public holiday weekend and I cant just drive to a supplier to buy a new cable.

If anyone has used ProTune on a PC can I please confirm the following settings with them.

RJ12 to DB9 Cable (the oh so special one!)

RJ12 Pin # Cable Colour DB9 Pin
1 White none
2 Black none
3 Red 2
4 Green 5
5 Yellow 3
6 Blue none

PC Settings: Windows [7 fully patched. No 3rd party firewall.]

Control Panel>System>Device Manager>Ports>COM1>Properties

Bits per Second: 9600, Data Bits:8, Parity: None, Stop Bits: 1, Flow Control: None
>Advanced> Use FIFO buffers: Checked, Receive and Transmit Buffer: 1

Please let me know what you have used that has worked. Thanks.

m_c
20-04-2014, 04:17 PM
Hmmm...only 3 wires would indicate it's only using RX/TX and GND.
What USB adapter are you using?


When I had a quick scan of the AM882 manual earlier, it said use a baudrate of 38400.
However whatever software is using the COM port should configure the port itself, ignoring any windows settings.

EddyCurrent
20-04-2014, 04:24 PM
Tony,

http://www.leadshine.com/Download.aspx?Type=software

This is my cable, it's a USB to serial converter then a short length of cable cut from an old phone with the correct plug already fitted.
I used default settings for everything else.

Pinout details for cable http://www.leadshine.com/UploadFile/Down/RJ11%20Connector%20for%20RS232%20Communication.pdf


12133


If you plug it all in then on your PC go to, Settings - System - Device Manager, you will see if it's connected and which COM port number it has. In my case it's showing the USB to serial converter on COM10
12134

As you can see I get an error when trying to read from the drive, this has always been the case with my system ( Windows XP ) and I've never followed it up, but I might now.
12135121361213712138

To be honest I would just as soon use the DIP switch 4 method as described in the manual , where you quickly switch it twice and the drive will set itself up.

JAZZCNC
20-04-2014, 05:04 PM
To be honest I would just as soon use the DIP switch 4 method as described in the manual , where you quickly switch it twice and the drive will set itself up.

Nah it's too hit n miss and you never know if you've done it fast enough, software is 100%. Also you can set the amp's for the motors exactly.


For what its worth I do get a read from the drives so I know that serial comms are working but every time I try something interesting in ProTune it says in effect comms dropped.

When you say try something interesting expalin what exactly your doing and what's happening.?

The fact your reading the drive means it's working. There is a quirk in protune Like eddy showed where trying to read or write to memory doesn't work, I'm not exactly usre which it is and I'm not at my setup laptop to check Protune but it gives a COM error.? . . . .I think if you click Write to Ram you get COM error but write to EEprom you don't.?

EddyCurrent
20-04-2014, 05:34 PM
Nah it's too hit n miss and you never know if you've done it fast enough, software is 100%. Also you can set the amp's for the motors exactly.

Every time I use the 'switch 4 method' I can hear the stepper motor 'singing' while it goes through the routine.


I think if you click Write to Ram you get COM error but write to EEprom you don't.?

I always get a pop up message telling me to do a 'Read' before I can do a 'Write'

JAZZCNC
20-04-2014, 07:44 PM
Every time I use the 'switch 4 method' I can hear the stepper motor 'singing' while it goes through the routine.

Yes maybe does energise stepper but doesn't mean it's tuned correctly and still it doesn't set the Amps. Software is much better and takes any doubt away.
Also with a little tweaking you can get motors running much smoother than the autotune does.



I always get a pop up message telling me to do a 'Read' before I can do a 'Write'

Well yes suppose that makes sense really because you would want to read the drive for it's current settings before changing them.
It's been a few weeks since I've used Protune as I've been using EM806 drives lately which have supaceeded the AM882 and the software is differant so can't remember if my Protune does this or not.

firetrappe
20-04-2014, 08:02 PM
Eddy, I've been playing with Protune on my XP machine today. I kept getting the error that you show. After a bit of research it turns out that it is due to the file stored in RAM on the drive being generated in a different language (probably from a Chinese machine).

If you go to Control Panel - Regional and Language Options then click the Languages tab and tick 'Install files for East Asian Languages', once it is applied you won't get the error message anymore.

Si.

TonyD
20-04-2014, 08:32 PM
Here is an example of what I am seeing. The four error screens should be read right to left.

1.Start ProTune, Select Com1 all ok so far.
2.Select 2nd icon from left, Read/Write drive parameter. OK So far.
3.Immediately I get error message 1, please check if the communication cable or drive is OK.
4.I click OK and I get the message Read finish successfully.
5.I see Screen 2. The parameters Configuration screen.
6.I leave this open or closed, dosent matter, I want to do something interesting and Select 4th icon from left, System Config.
7.Immediately I see Com Error! message, Please check connection and restart!
8.I click OK and immediately I see screen 3, the SystemConfig screen, clicking Start does nothing.I cant close this screen top right.
9.So I decide to close the whole application, restart and go straight to System Config, I see error message 4, Com Error, Fail in reveiving monitor data.
Thats about it, I have to click OK 4-5 times to close the dialogue and the protune app essentially does nothing for me.

TonyD
20-04-2014, 08:33 PM
I'm still not convinved I have the cable wired correctly, can anyone confirm the pinouts above please? They are as per the leashine data sheet but I just want to be sure.

firetrappe
20-04-2014, 08:47 PM
Tony, it looks like a problem with the cable to me. I bought mine from Gary @ Zapp so i'm not sure of the wiring. There is a little thread on Zapp's forum which talks about the cables : Zapp Automation Ltd (http://www.zappautomation.co.uk/forum.php?gp_do=s&id=1)

Not sure it it'll help, but maybe worth a quick read.

Si.

EddyCurrent
20-04-2014, 08:51 PM
I'm almost certain it's not going to be COM1, did you look at the ports in device manager ? while you are on that page, keep plugging/unplugging your interface cable a few times so you can identify which component is changing then you can see which COM port it's using.

TonyD
20-04-2014, 08:52 PM
Thanks Si, yes I suspect the same, I found that thread on the Zapp site but wasnt sure what he meant by pull the plug. In any event I think I'll order another cable wait and see.

EddyCurrent
20-04-2014, 08:59 PM
Eddy, I've been playing with Protune on my XP machine today. I kept getting the error that you show. After a bit of research it turns out that it is due to the file stored in RAM on the drive being generated in a different language (probably from a Chinese machine).

If you go to Control Panel - Regional and Language Options then click the Languages tab and tick 'Install files for East Asian Languages', once it is applied you won't get the error message anymore.

Si.

Si, that's an excellent find I can't wait to try it !

TonyD
20-04-2014, 10:11 PM
Eddy, thanks. It does seem to be COM 1 but thats becuase I only have one physcial port on the machine and I do get someting out of it when it find a port that works. I plan to go and get a RJ12 crimp too tomorrow from Maplins and have a go at building another cable, I'll post as to my progress.. :)

TonyD
20-04-2014, 10:12 PM
Thanks Lee for moving these threads under the build log.

TonyD
21-04-2014, 10:05 PM
It was the Cable! Thanks all for the advice and support. :)

firetrappe
21-04-2014, 10:44 PM
Glad you got it sorted out in the end Tony.

Si.

TonyD
22-04-2014, 12:46 AM
So. frustration with the electronics over and frankly the "is that it?" feeling from ProTune in the rear view mirror (apart from the fault codes which were interesting and somewhat revealing) I took a look at the mechanicals again as there was some resistance/binding on the right hand X Axis. I attached the fixed and floating bearings on the frame using the link to the gantry to support the alignment and did the best I could do at the time. The problem is there is no adjustment in the design, mainly becuase it was made using hand tools/small press drill and one or two trips to people with milling machines (for the slot in the front Y Plate, and a small pocket on the inside of the Z Axis).

So I'll need to come up with some plates that allow the fixed and floating bearings on both sides to connect to the frame with some (mainly horizontal) adjustment, its about 1mm out end to end currently. More on that when I get some time..

JAZZCNC
22-04-2014, 12:58 AM
So I'll need to come up with some plates that allow the fixed and floating bearings on both sides to connect to the frame with some (mainly horizontal) adjustment, its about 1mm out end to end currently. More on that when I get some time..

If it's out that much then Be very careful not to run the gantry right upto the ends as you can easily bend the screws.

TonyD
22-04-2014, 01:07 AM
Thanks Jazz, I've halted all planned movement now until I sort the adjustment. I was worried I had damaged the screws so I stripped them down and thankfully they roll nice and even.

TonyD
24-05-2014, 10:38 PM
<Do not pass go, do not collect 200 pounds>
So. Adjustment plates produced by laser and returned nicely, tapped OK in the laser hole but had to try 3 M4 Taps to find one that worked well. Adjustment plates assembled and fitted on one side, was getting ready to fit the other side when the movement of the gantry got my attention.

Several months ago I applied some light machine oil (3-in-1 I think) to the hiwin rails and ran the gantry on the X Axis back and forth a few times, was satisfied with that and thought nothing of it. Fast forward a couple of weeks and the oil had dried into an almost varnish like residue that was possible to remove with a scriber but left a glue like substance behind. Bottom line the oil had gummed up the rails and the linear bearings were not so much cutting through it as riding roughly over it. The gantry had become still to move and it didn’t feel smooth at all. So took one rail off today, checked the bearing block and it seemed OK. I spent about some time with light sanding paper on the frame where the rail has been bolted as there was a rim of hardened oil residue all around the rail, that done I took a wire brush to the back of the hiwin rail to remove the frame residue that was stuck there and then thoroughly cleaned the rest of the rail with degreasant and elbow grease. The rail came out really clean, the bearing I had removed ran rock solid and smooth over the rail and once bolted back in the gantry is noticeably smoother. I'll do the other side next day or so. I have already got the right type of grease for future lubrication but thought I'd post this as a warning to others of what not to do. Once that's done I'll work on the alignment of the second X Axis screw and bearings and then go back into Mach, check the connections, limits & homing and finish setting the soft limits which is as far as I got before I took several steps backwards.

Jonathan
24-05-2014, 11:59 PM
[...] thoroughly cleaned the rest of the rail with degreasant and elbow grease. The rail came out really clean, the bearing I had removed ran rock solid and smooth over the rail and once bolted back in the gantry is noticeably smoother. I'll do the other side next day or so. I have already got the right type of grease for future lubrication

Just to be sure, did you put any grease or oil in the bearing before testing again and finding it to be 'smooth'? If not, then (at risk of stating the obvious) do put some in before going further as running without any lubricant (which is presumably the case if you've cleaned it thoroughly) even for a short time will reduce the life of the bearing. Hard to say by how much, but there's no need to risk it.

Edit: You might want to check the ballnuts for debris whilst you're at it.

TonyD
25-05-2014, 11:06 PM
Thanks Jonathan, I havent moved it much at all but now the other side is done too I am much happier with how smoothly it moves. Thankfully I didnt oil the ball screws at the time so the residue isnt on them. I'll grease it all up again soon. Tony.

TonyD
01-06-2014, 12:34 AM
Move to Go. Do not pass Go.
OK, so I'm about back to where I was before, when I was racking the machine and trying to bend the ball screw out of true. I've got the machine moving smoothly and I'm setting up homing prior to figuring out the limits and setting the soft limits and I've hit a snag, done some reading to avoid wasting the time of people and I'm not confident yet.

Problem is dual X axis homing, Z homes OK, Y homes OK, when I select X home, (thankfully at 2% speed) the maching starts to rack as one X axis motor is turning, the other isnt. The weird thing is that under jog movement they both move together, so now I'm into the Mach ports/pins and homing screens. Ports and pins I have X and B axis configured correctly I believe, they move on jog so I'm OK with that, Z,Y,X Home I have values and the diagnostics page in Mach confirms they work and Y and Z home OK anyway. I *do not* have a home input signal for the B Axis as I assumed that as its slaved to my X axis that would be OK, clearly its not so I suspect I need to add input signal values for the B Axis and copy the X Axis home port and pin setting to fool mach into seeing the same switch? Is that the answer? as always advice appreciated, thanks.

Clive S
01-06-2014, 12:49 AM
Tony. I am off to bed now and had a small drink. But in the morning if nobody has responded I will check my Mach3 and let you know the settings for it. You need two homing switches as the way it works is that both motors should drive the axis back to the home position then one switch trips and backs off then the other does the same. Its also important to have exactly the same setup in motor tuning for both the motors. When it works it will work fine. ..Clive

TonyD
01-06-2014, 01:02 AM
Thanks Clive, for now I only have one homing switch, can I/do I have to use the inputs for one home switch for both home inputs?

JAZZCNC
01-06-2014, 01:15 AM
Tony you MUST have 2 switches on separate inputs for homing to work with slaved motors. The first switch can be wired in series with Y & Z but the second must be on it's own input.

Also in General Config you must UNCHECK "home slave with master axis"

TonyD
01-06-2014, 01:30 AM
Thanks. I guess I need to figure out a way to mount the switches on both sides with enough adjustment they both connect at the same time.

Clive S
01-06-2014, 09:08 AM
Thanks. I guess I need to figure out a way to mount the switches on both sides with enough adjustment they both connect at the same time.
Tony Jazzcnc hit the nail on the head I could not think straight last night. If you pm me your number I will run through the settings with you.
If you are worried about racking when testing take the belts or motors off both sides then you can see what is happening. ..Clive

JAZZCNC
01-06-2014, 10:01 AM
Tony Jazzcnc hit the nail on the head I could not think straight last night.

Alcholic Pisshead syndrome. . .:very_drunk:


Thanks. I guess I need to figure out a way to mount the switches on both sides with enough adjustment they both connect at the same time.

Tony if you use the bullet type proximity switches they have a threaded body which you can use for adjustment. Also if using proximity switches you can use the sensing distance as kind of fine tuning which is handy for setting slaved motors.

TonyD
03-08-2014, 11:36 PM
Finally, auto home on all axis - thanks for the help. For the help of others, although as Jazz says the dual input was required and I didnt have that to start with, once I wired that up I still didnt have the dual X axis home working. I was about to write a detailed post here with all my Mach config settings when I noticed a discrepancy between the various B or A axis settings, once they were all pointing to the same B axis it worked. Happy days :)

EddyCurrent
03-08-2014, 11:46 PM
It's great when a problem is solved, say "hi" to Peppa Pig for me

TonyD
04-08-2014, 12:21 AM
Oink! Oink!

TonyD
03-09-2014, 11:53 PM
Some teething problems I think, seems to be Z axis...?

I'm using Cambam but only really want to draw on the MDF bed of the machine at this stage. I've defined my stock as 5mm thick but the arcs lines etc are set to profile to only -1mm in the cambam dialogue. The G Code for the test is below, the machine homes just fine. when I Jog the machine I see Z- going in the direction of the bed. So I get a post-it and manually jog the Z until it bites the paper and then back it off say 0.650 mm and then set Z to 0 in mach. The theory being that the -1mm "cut depth" will push the biro into the mdf bed.

However when the program runs the pen seems to move up not the 1mm down I want it to go. I did some googling and changed the Z motor direction in Mach and that didnt feel at all right, homing went backwards (expected really). So my question, should I have selected a + z move (cut depth) in Cambam or is there a clue in the G Code below that the machine is set up wrong? Advice as always welcome. Thanks.

( Made using CamBam - http://www.cambam.co.uk )
( CNC_Template 9/3/2014 10:17:13 PM )
( T6 : 6.0 )
( CUTVIEWER )
( FROM/0,0,5 )
( Select dummy tool to avoid warnings )
( TOOL/MILL,1,0,20.0,0 )
( STOCK/BLOCK,500.0,500.0,5.0,0.0,0.0,5.0 )
G21 G90 G91.1 G64 G40
G0 Z3.0
( T6 : 6.0 )
( Tool Taper coming soon )
( TOOL/MILL,6.0,0.0,0.0,0 )
T6 M6
( Profile1 )
G17
M3 S1000
G0 X207.0 Y120.0
G0 Z0.2
G1 F300.0 Z0.0
G2 F1500.0 X196.5 Y113.9378 Z-0.1667 I-7.0 J0.0
G2 Y126.0622 Z-0.3333 I3.5 J6.0622
G2 X207.0 Y120.0 Z-0.5 I3.5 J-6.0622
G2 X196.5 Y113.9378 Z-0.6667 I-7.0 J0.0
G2 Y126.0622 Z-0.8333 I3.5 J6.0622
G2 X207.0 Y120.0 Z-1.0 I3.5 J-6.0622
G2 X196.5 Y113.9378 I-7.0 J0.0
G2 Y126.0622 I3.5 J6.0622
G2 X207.0 Y120.0 I3.5 J-6.0622

Clive S
04-09-2014, 12:15 AM
It is more common to set the home for Z at the top, that means all cuts will be -Z. So when you zero your Z at the bed a command to cut would be a -Z move.

You can tell Mach to home in any direction. ..Clive

TonyD
04-09-2014, 12:22 AM
Dont know if this will double post? Thats how its set up Clive. the code looks ok to me too I think, start of Z3 as a positioning move, moves down to Z0.2 still just above bed and then gets to Z-1.0 4th line up from the bottom. so why is the spindle not actually going down but going up?

Boyan Silyavski
04-09-2014, 12:24 AM
The Z as Clive said should be 0 at the Top and go down with negative sign.

Do you recalculate the toolpath? When set working Z0 i mean.

Ok Zero at the middle of the air. Then in the mdi window write Z-1 and hit enter. It should go down from there 1mm/ becarefull if in inch cause will go 1 inch down :loyal:/

TonyD
04-09-2014, 12:35 AM
Thanks chaps. I did the MDI -1 and it does come down 1 mm.
Maybe I'll start again with cambam and see if that helps.

EddyCurrent
04-09-2014, 11:31 AM
Try it using the Mach3-CutViewer post processor as you are, then try it with just the Mach3 post processor, see if there's a difference.

TonyD
05-09-2014, 02:19 AM
Thanks Eddy, I did change the post processor and reset Cambam to start from scratch. I think its the stock to cut relative to the Z that had me confused.

I created two tests. For the first I created a simple shape in Cambam than had a depth of 5mm relative to 0 and marked the Z axis depth on a post-it note. I stopped the program when it was at depth and although I didnt measure it accurately I'm OK that it ran a clearance height plus Z and ended up around -5. The camera work on the first video is pretty awful, makes you realise that those who do youtube videos well put some work into it :) should be at https://www.youtube.com/watch?v=ngDW4964BR4

For the second test I tried manually adjusting the Z and ran the program, It using a hacked Biro and the ink tube slips a little as well as bing not the smoothest line. I found that around 1.05 into the video the pen caught for a few secs, I think its down to replacing the cut down biro with a felt tip now. This video is still not great but has the virtue that I'm not really trying to do too many close-ups. http://youtu.be/iah4JGcsFbI

For those on thier journey to making a machine, I can say it is very satifying when it starts to move and it does kind of sing a bit during arcs. More videos and stuff in a few weeks when I get another chance to play.

TonyD
21-09-2014, 03:01 AM
And for my next problem.. :)
I'm getting comforatble jogging the machine around now, I'm about to use it to drill some holes through the MDF bed and into the Steel base structure prior to tapping the steel and using a Forstner bit on the MDF to counterbore it so I can firmly fix the bed prior to checking the gantry for good geometry, surfacing etc and.. I cant seem to get the VFD (& Spindle) running and have Mach communicate with the PLCM control board. Interesting enough I did have issues a few weeks ago where I would lose comms to the control cabinet whenever I closed the control box, which was most weird and I bypassed the problem by removing the USB chassis mount and just running the USB cable direct from the PC to the control board through a hole in the box. This all leads me to suspect earthing as a culrpit? Any ideas how I would trace the issue, sorry its a bit vauge, I can post pictures and update the electrical schematic if that would help. Any advice, as always, much appreciated. Thanks.

JAZZCNC
21-09-2014, 01:28 PM
Ok well first why are you using USB when the PLCM is capable of using Ethernet.? Ethernet is much much better at handling noise than USB.
Need some idea or pics of how you have wiring/earths etc because you shouldn't be having PC earth interfering with Control box earth and vise versa. You haven't earhed the PC to control box earth have you.? PC should just use the Earth in the Plug and no external earth to control or machine frame.
Also have you got a Earth from machine frame to Control Star earth.? If not then do so.

Can you post your Mach3 XML file (Zip it up) it will be in mach3 folder using same name as the profile your loading from Mach3 loader.

What VFD is it and does it work normally Ie: ~using buttons on front.?
What settings have you changed in the VFD and what did you enter.?
What inputs on the VFD are you connecting to.?

EddyCurrent
21-09-2014, 01:49 PM
And for my next problem.. :)I can post pictures and update the electrical schematic if that would help.

That's the best idea, also I don't know why you suspect earthing as the problem, I would be thinking it was the chassis mounted USB connector from your description, or that the closed door was exerting undue pressure on the cable thus twisting it somehow in the socket.

JAZZCNC
21-09-2014, 03:34 PM
also I don't know why you suspect earthing as the problem

Because it's a common hack with USB that is easily affected by noise. The USB SS suffered from it and Tony's probably read something around these lines.? Just guesssing on that thou.!

EddyCurrent
21-09-2014, 06:59 PM
Because it's a common hack with USB that is easily affected by noise. The USB SS suffered from it and Tony's probably read something around these lines.? Just guesssing on that thou.!

Well I was just going off this comment, I've had this problem in the past when cables are routed to the front panel so that when they are moved upon closing the door they affect other components they touch.


. . . I did have issues a few weeks ago where I would lose comms to the control cabinet whenever I closed the control box,

TonyD
21-09-2014, 07:52 PM
Thanks Chaps, I'll do some documentation tonight. I may have muddied the waters a little, the first issue was the closing door, it was weird and I couldnt see anything, so I removed the chassis connector and it was resolved, I'll post pictures to describe that one better. The second issue (possibly realted) seems to be just when the VFD spins up the spindle as I can have the VFD on spindle stationary and the comms stay up. The VFD is the standard Huanyang variant from China, nothing changed from default, it does make some noise on start-up, the water cooling is active but I guess as its indoors and I'm usually tinkering in the early hours, luckily our neighbour is on holiday :) The reason I'm on USB is the board doesent have the Ethernet module on it, I'll try to source one shortly. It does feel noise/interference currently but frankly the control box was very prototypish anyway with lots of connectors across different faces of the box and the door used becuase I ran out of space, I was planning to move to a larger box and possibly mount the vfd in there also. Anyway, thanks for the ideas, I'll document where I am, look at the ethernet module and see about a bigger box shortly. Three quick questions though..

1). Do folks power the VFD, PSU's and all other CNC stuff (not including PC) from just one 13A wall plug, I'm using the VFD on its own plug right now.
2). For the ethernet SS, do I need to add a router or switch between the SS and PC or will a cross-over cable generally do between them?
3). I would really like to mount the control box on the frame, I've read comments about vibration etc, is it really a bad idea or do some folks find it works ok, it would make my cabling neater and the whole machine would be more compact.

Cheers. Tony.

Clive S
21-09-2014, 08:03 PM
If you have mounted some parts on the door have you in fact grounded the door to the box, or are you just relying on the hinges for the ground? ..Clive

EddyCurrent
21-09-2014, 08:09 PM
1). Do folks power the VFD, PSU's and all other CNC stuff (not including PC) from just one 13A wall plug, I'm using the VFD on its own plug right now.
2). For the ethernet SS, do I need to add a router or switch between the SS and PC or will a cross-over cable generally do between them?
3). I would really like to mount the control box on the frame, I've read comments about vibration etc, is it really a bad idea or do some folks find it works ok, it would make my cabling neater and the whole machine would be more compact.

Cheers. Tony.

1. Mine is hard wired to the consumer unit using armoured cable.
2. The ESS connects directly to the PC with a standard ethernet cable.
3. Mine is wall mounted but maybe you could mount it to the frame using rubber engine mountings.

TonyD
21-09-2014, 09:16 PM
Thanks to Gary @ Zapp who answered a query today, the PLCM E3 card I am using also support's ethernet, I've been assuming that it didnt since the there appeared to be a daughterboard on the SS literature that isnt on the card I have.I guess the Rev 2 of the SS that I have didnt need it. I have comms to the SS now, I had a crossover cable adaptor spare so thats working now :) I can jog so thats working fine. The VFD is soundling like my very own Rols Royce Trent 900 turbo fan & I can jog so no lost comms :) :)

Thanks for the feedback Eddy, until I upgrade the control box I'll stick to a seperate wall plug for now. Time to read the VFD manual now and find out how to not have it spin so fast it wakes the neighbourhood. Thanks again.

JAZZCNC
21-09-2014, 09:33 PM
Thanks to Gary @ Zapp who answered a query today, the PLCM E3 card I am using also support's ethernet, I've been assuming that it didnt since the there appeared to be a daughterboard on the SS literature that isnt on the card I have.I guess the Rev 2 of the SS that I have didnt need it

Ok first let me just clear up something because it becomes confusing for others reading this.! You don't have a SS this relates to a particular Brand name of Motion control card called the Smooth stepper shortened to SS or ESS for the ethernet version.

You have PLCM E3 which is still a motion control card but not to be confused with the SS.

Now what breakout board do you have because the PLCM E3 doesn't have spindle speed control directly so for this you'll need either external spindle speed control board or a breakout board with one built into it.

If you haven't set up the parameters in the VFD then don't go trying to make it start else you'll possibly do it some damage if the parameters are set wrong.
After setting intial parameters like Voltage, amps etc then if you want to control it thru G-code and the control software then you'll need to set more parameters in VFD so can be controlled from an external source. Also along with this you'll need to set outputs and spindle settings etc in Mach3.

Do a search on the forum and you'll find the settings and parmeters to change.

TonyD
21-09-2014, 09:40 PM
Thanks Jazz, you're right, I meant the generic, it is as you say an PLCM E3 from Purelogic, a Russian company. I'm using the Zapp ZP5A-INT breakout board, I'm not planning to try and control the spindle speed using that but I do have a USB dongle that supports RS485 that I've read will work for speed control. We'll see, right now as you say I've got to explore the VFD manully and get that going, with perhaps a little less noise.

Clive S
21-09-2014, 11:26 PM
You could put a pot on the vfd front panel to control the speed if its the standard Chinese one. ..Clive

Clive S
21-09-2014, 11:41 PM
Fitting a pot to a vfd.

Here is the link to the way I did it http://www.mycncuk.com/threads/7388-Fitting-a-pot-to-a-Chinese-Spindle-VFD?highlight=speed ..Clive

TonyD
22-09-2014, 02:09 PM
Great Thread, thanks Clive.

TonyD
19-10-2014, 11:32 PM
I have two NPN NC bullet type proximity switches on my X Axis. Normally the inbuilt LED lights up when the target metal approaches and triggers the switch. As of last week, one of the proximity switches is lit the whole time, I cant change its state. Mach is reporting the error when I try to home the X axis. I've run a new switch to the same connectors as the faulty one and the new switch stays lit also. Note, in case it matters both switches were in parallel during this test. I can't spot any obviously loose wires in the control box. Before I start to tear into the control box, has anyone experienced this type of failure mode from the proximity switch before? Cheers. Tony.

EddyCurrent
19-10-2014, 11:39 PM
Have you tried them on another input ? are they 3 wire or 2 wire type ? 3 wire likely.
Did you change the position of them so the end is more flush with the surrounding bracket ? (what I mean is that's a bad thing)

TonyD
19-10-2014, 11:54 PM
Hi Eddy, thanks. 3 wire, I'll change the input and report back, no - no adjustment, cant see a good reason for the signal stage and the proxy in paralell shouldnt have triggered either it was hanging in mid-air, I should have said that too.

TonyD
20-10-2014, 12:05 AM
OK, turn out that the PX is actually behaving fine, its the other X Axis one thats dead, ran a new one in parallel and that one works as advertised so I'll just swap the dead one out. Sorry for the confusion. Tony.

TonyD
20-10-2014, 12:50 AM
Also, for those who are thinking "I thought those Px switches were reliable", they probabily are, the two I have on my X Axis have little dents in the front, no doubt from an impact with the X Axis! So no surprise that at least at one of them failed :)

Clive S
20-10-2014, 07:57 AM
Also, for those who are thinking "I thought those Px switches were reliable", they probabily are, the two I have on my X Axis have little dents in the front, no doubt from an impact with the X Axis! So no surprise that at least at one of them failed :)
Tony can you remount them so that the target slides over the end instead of crashing into it. ..Clive

TonyD
29-12-2014, 06:14 AM
And for my next problem.. :)

So things are generally going well but very slowly. I was nicely surfacing the MDF bed of the machine but saw one of the ball screws coming away from the coupling so a good spot of tightening up was carried out and I am much more confident that wont happen again. Having created infeasible amounts of dust and also realised that the Cambam feed was set way too low I have paused to consider some options. Firstly the dust is a problem so I've got some parts on order for a cyclonic dust extractor into my Aldi ash bin and I think I'll be OK to machine some acrylic for a dust shoe but I digress...

I thought I'd add tool auto height. I read the threads, set up the inputs and did the wiring which (as its probably) relevant was just two wires from the BOB input to a set of crocodile clips one for the tool for now and one for a little plate. All was well and I tested it by touching the probe to the plate in my hands to make sure the Z retracted and it did, ... the joy :) so then the real test and as soon a I connected the crocodile clip to the end mill in the spindle the Pilz tripped out. Not the RCD's the Pilz. So, I guess its an earthing issue but without me having to draw up a wiring diagram, has anyone experienced this type of thing before? Thanks in advance chaps.

Stupoty
31-12-2014, 06:23 AM
Tony - it may be worth seeing if you have another local steel yard near you. My local supplier charge £4.1 +VAT per meter for 50x50x3 (£147.60 for 30m) as a base price & charge less than that if you order a reasonable quantity. I also get them to cut some of it to size for me - depending on quantities, they charge 50p per cut or less...this means I can fit it in my car & pick it up myself removing the delivery charge as well as saving me time cutting stock to size.


Do you have the name of the stockist, thats a fairly good price :)

i always check metals4u.co.uk for steel and their about £60 + vat for 6m

stuart