my 6040 experience (yeah, and problems) so far

Being in a hurry and not wanting to drop a big wad of cash yet, I decided on an eBay 6040 to get me started. I went for 800W spindle with the standard 1500W VFD and the blue controller box from hell, but I made sure it was a recent version with 3 separate M542 drivers rather than an all in one pcb.

The long axis (labelled Y and that's probably what I'll call it) has one 1605 ballscrew driving the gantry, with SBR16 supported round rails. Cross axis and Z are unsupported 20mm and 12mm round rails.

First worry was the low quality of the mounting of the 1605 screws, just a washer (or not) and a badly made (wonky) nyloc nut. I tightened it all up just a bit more than finger tight. The X drive is a bit graunchy at times, in one direction only. Any ideas on what that might be? corrosion of the bearings? badly made ballnut? I intend to take it apart and fiddle with the mechanical bits before I run it in anger.

What sort of lubrication does a ballscrew need?

Next problem was that on and off (mostly on) and for a prolonged period when it happens, x+ and z+ wouldn't jog after doing y+. but doing y- would make them work again. I swapped the X and Y leads at the control box, and the problem switched axes. Therefore the problem is inside the control box, either the BOB or the M542s. I will swap the M542s next and test the grounds. See if that helps or tells me more.

Final thing fo rthe moment, how do I swap axis directions in Mach3 ? currently Z+ is down on my machine?

I've checked the VFD setting and run the spindle up to speed. Not plumbed the pump yet. How long can I run the spindle without cooling?

Answer to some of your points.

Not saying this is your problem but I had ball trouble too, see here for how I solved it; http://www.mycncuk.com/threads/6565-...6350#post56350
If you run the ball nut up and down by hand you will see if the problem lies with that, it could be the end bearings or the screw is bent, http://www.mycncuk.com/threads/6565-...6172#post56172 or it's mounted out of line. I also had problems with the BK type bearings blocks but not sure if yours uses that type. http://www.mycncuk.com/threads/6565-...6128#post56128

This is the right grease for linear bearings as well, some people use oil for both instead of grease.
http://www.toolstation.com/shop/Auto...0/sd180/p89691

On the Ports and Pins / Motor Outputs screen, if you want to reverse the directions of any of the axes, toggle the "Dir Low Active" checkbox for that motor.

Good thread here about spindles cooling etc.
http://www.mycncuk.com/threads/5814-...nd-Advice-More

Thanks Eddy, will have a read now.

I'm pretty sure I've narrowed the controller problem down to the BOB ( I knew having separate drivers would pay off :) ). Each axis works OK by itself when other are unplugged. And swapping the BOB outputs to different M542s moves the problem axes around.

I should take the long axis drive apart now and inspect, so that I can email the seller with all the bits I want replaced in one go.

Oh yes, the M5642 drivers were set at 16 microsteps. I changed this to 8 but it made no difference. Is it worth trying other micro-step settings?

Would I be right to assume that smaller (more) steps is harder for the electronics to do / more likely to go wrong ?

Quote:

---

Originally Posted by jimbowley

Would I be right to assume that smaller (more) steps is harder for the electronics to do / more likely to go wrong ?

---

No it's the Parallel port that struggles not the electronics. The higher the steps the more pulses required for the same distance hence why the higher the step count the slower the machine can be tuned with the set Kernal speed in Mach3.

Going higher on the steps helps smooth the motors but works the PP harder and this can cause missed steps if it's struggling.

8 is a good setting and I would leave it there.

Regards the drive problems check the STep & Dir connections at the drive giving the trouble this as been a cause of trouble in the past on these machines along with shity wire quality.

Spindle will run for ages 1 hour + without cooling if not under load. I've often cut small jobs without turning my pump on.

Not sure if your aware or not but if you change the MS you will need to re-tune the motors Step per setting.?

Also don't run higher than 45Khz kernal speed in Mach3 other wise your opening your self up to potential problems. Best to keep at 25Khz or if must then 35Khz but don't change unless you really can't get the speeds you would like.

Quote:

---

Originally Posted by JAZZCNC

Regards the drive problems check the STep & Dir connections at the drive giving the trouble this as been a cause of trouble in the past on these machines along with shity wire quality.

---

I'm not familiar with what the diffferent wires are, but the BOB has a 4pin connector for each axis. 3 of the 4 pins are used and the 4th pin has a loose wire hanging from it. The BOB looks like it has also the option of output via screw terminals, so I could cut the terminal blocks off and use those. But I'm wary of doing that before I hear from the seller, and if I get a replacement BOB Id rather just plug it in.

The wiring all looks pretty nice to me, but I'm not an electrics person. All the sockets into the controller feel like plastic, but they are conductive, I've tested the continuity of the casing.

Just grabbed this pic of my BOB from ebayAttachment 13217

Ok well I'm not a fan of those silly plugs and I'd wire to the terminals every time but your correct to wait untill seller responds.

That said if the problem moved then it won't be the BOB and chances are it's a drive or wiring to drive.? . . Again I would check the connections at the drives.

I moved the outputs of the BOB to different drivers. So the problem changing axis shows that the driver is not the problem as the problem follows the BOB output independent of which driver it is wired to.

Here is a boring video of the 2 problems.
https://www.youtube.com/watch?v=IoAGHdbffnQ

Using the red LED's on the drivers as diagnostics, what happens when everthing is connected apart from all the motors ? i.e. just unplug all the motrs, does the fault still happen ? DO NOT unplug them with power on, it will probably blow the drivers.

The LEDs behave the same with no motors connected (there was just 1 connected in the video)

Did you try a different computer ? Is it currently a desktop or laptop ?

new findings: forget what I thought before about the drivers all being equal.

When I disconnect the inputs and outputs from the non-used drivers (easy with plug in modules :) ) I find that I only get the malfunctions when using driver2, drivers 1 and 3 can't be made to fault by pressing the jogs on the other axes.

I will go run this simple test on the BOB X and Z outputs in a bit (if driver2 doesn't fault on the X or Z axis then the BOB is also involved in the problem).

Obvious but it's got to be said, make sure all the drivers have the same settings/links re. this manual (which I guess will be the right one ?) http://www.leadshine.com/UploadFile/Down/M542d.pdf
Also make sure any wiring is well clear of the Enable pins.

This looks very much like Keyboard trouble so Try another keyboard.!!

Also one thing to try is another parallel port cable. Did this one come with the machine or is it one you've sourced.? I've seen many weird things like this happen and it be the parallel port cable.

Do all the axis move ok with G-code commands.

Also where did you get Mach3 from did it come with the machine.? Wgat version number is it.? You'll find it in the help under about.!

Yes that's why I asked about another computer, also is the BOB powered by USB only or does it have a 12/24v supply ?

dip switches are all the same (I adjusted them from 16micro steps to 8 yesterday).

Mach3 is latest version from Mach3 website.

Will try a different cable (I'd like to dismiss this, but my given one failure of deduction already I'm more open to testing).


I got the graunchy sound (which I thought was ballnut, but now think it is motor caused) also when using Gcode. I haven't cut anything yet so don't know about circles.

I just ran my tests using the X and Z outputs from the BOB (alone with no other connection) and got the same result, ie the fault only occurs when using driver2. different output from BOB also means different keyboard key and different pin on parallel port, so tends to rule those out as cause.

Defect appears to be in driver2 and maybe also BOB? I can't work out how pressing eg Y axis (with no driver connected) can cause problem on X axis, other than the BOB also being at fault.

I'm about to put the gantry back on and try driving X and Y using drivers 1 and 3. Which hopefully should work.

I almost don't want to mention this, using driver1 (a supposed good one) I did an extended test for the graunchy sound, after about 5mins of up and down (still just the long axis, no gantry) it started to make a bit of a noise.

This leads me to think the ballnut is fine and the noise is related to the drive problem. Driver2 starts to breakdown(?) quickly, Driver1 breaks down after longer time. I don't know what I mean by break down, maybe a heat thing? (I'll put my hand on the heatsink next time).

So I've established that D2 faults even when by itself. D1 and D3 do not fault.

When I use D1 and D3 together, D1 faults (whether it is X or Y axis). Go to about 2:40 to see the problem demonstrated.
http://youtu.be/raNVEsXro5c

I did a prolonged test with the strongest driver D3 on the long axis, almost no graunching noise after 10mins (, except at 9:21 !) video: http://youtu.be/3EQlJWFV8XY

I've now tested some GCODE, same fault. eg zigzg lines don't work on one diagonal.


My 1flute cutters arrived so I had to make my first cut in delrin, 11000rpm DOC 1mm feed 1100mm/min. It was quite loud, the delrin was in a vice clamped to the table. It's all a bit like an aluminium drum at the moment, I will have to add a layer of something once I've thought about how I'm going to clamp my work etc

Hi iv just picked up a 60/40 just the same as yours but it has the all in one board. Iv had problems with my motors as well. if i set the tuning to what they say in the set up sheet, Then run them at 100 jog they jam when i running them up and down. they run for about 50mm then jam. I have set the motor tuning to half what they say and they seem to run ok. so that would be 400 1000 100 . Iy runs slower but no horrible crunching. This is my first cnc so i dont know enuf to know if this is ok but its running.

Must be age related but I've been getting mixed up with this thread http://www.mycncuk.com/threads/7816-...0874#post60874
As amxen says get the velocity and acceleration down on Mach3 to something that won't (shouldn't ?)cause any issues.
Once that's eliminated my stab is that the breakout board is faulty and it's either a bad soldered joint on the inputs pull-up resistor pack or the pack itself. (sometimes I like to live dangerous :bull_head: :hysterical:)

The good driver is happy running the long axis at 4000mm/min & 400mm/s2 (twice the standard) on its own. The fault occurs even with all axes at 1/2 standard speed.

What's a half decent PSU/BOB/3x drives gonna cost me? If the seller can't replace quickly or the next ione is also faulty I'll be after a discount towards my replacement controller box.

I may be wrong but I think these are the main UK suppliers of stuff used by DIY'ers, plus obviously there's ebay.

http://www.diycnc.co.uk/
http://www.zappautomation.co.uk/
http://cnc4you.co.uk/

If I was in your position I woud change the breakout board first and would most likely choose a similar board from ebay, they are less than a tenner from abroad, in fact I have just realised I have a board like yours that I bought to get extra inputs to my system, it's working fine in that role. It connects into port2 of my SmoothStepper

Attachment 13225

I would go for the part refund and buy decent electronics with a mind to transplant to your next machine. . . . No point replacing crap with crap.!!

To be honest I'd want more than just the refund of the Drive and BOB there's your fiddling time and Down time to take into acount.
These things need to be accounted for when your thinking about electronics and machine because what's the point having a machine if your spending more time chasing faults or fixing the bloody thing or than using it.?

just found out what my problem is. made my own post

http://www.mycncuk.com/threads/7839-...1165#post61165

dont know if it will help any one

Bit more info. summary: Mistakes corrected, but no happy ending.

When adjusting dip switches, refer to the instructions on the side of the driver, not to an instruction for a different driver that you found on the internet! Yes folks, I don't have leadshine M542s, I have Tongka M542s and the dip switch setting are different.

Funny thing is that when I looked at the dips I found a setting that indicated 16 microsteps, unfortunately that was just chance. So when I thought I was changing from 16steps to 8steps, I was actually changing from 3A to 2A, and from 0% decay to 50% (whatever that is).

I've put everything to how it was set originally. And I've tried 8steps instead of 16. In both cases all the faults are just the same as before.


I did spot that dip7 is set at 1/2 current. The other option is full current. Looks like I'm running at half throttle ?

That setting is to reduce current in the motor at idle to reduce heating effect, it has no impact on running torque.

Quote:

---

Originally Posted by jimbowley

I did spot that dip7 is set at 1/2 current. The other option is full current. Looks like I'm running at half throttle ?

---

This is to drop the current when motor is at stand still to reduce over heating keep it set.

Edit: Opp's sorry irving didn't see you'd replyed already.!

Seller responded just in time (I was openeing a case as the message came in). He says no controller boxes in UK but will send me 2x driver controllers. I said I also want interface board, even if it ships from china.

I could do some cutting using manual control for Z, as that wouldn't trigger the problem. but I will wait and get on with other stuff until spares arrive. Signing out for a few days :(

A pair of 2flute 6mm cutters arrived. Very cheap and came in 4 days from UK not HongKong. http://www.ebay.co.uk/itm/190984562714 price was only £1.69 last week ! I've got some more expensive ones as well coming from china, comparisons will come in due course.

So I disconnected Z, and that allows me to drive around in X and Y.

Feed 1200mm/min. 6000RPM.

First I did full width passes along the top at about 1mm doc > all fine.

Then I sidemilled the endface with doc 3mm stepover about 1.4mm. Fine conventional but when I tried to climb mill on the way back I got a loud vibration.

That made me panic :unconscious: and I hit a key (aiming for reverse!) and instead took a cut at full width 9mm doc. That too made a noise but nothing broke.

Attachment 13229

I'm pretty pleased. Acetal is the easiest plastic to work with, and I'm all for making life easy. These feeds and speeds are plenty for the small pieces I have planned.

I'll probably go back to 16 microsteps giving a max feed around 2300 as I don't think I'll be needing 4700 mm/min.


Hoping my replacement drivers arrive soon so I can do some CNC.

Quote:

---

Originally Posted by jimbowley

I'll probably go back to 16 microsteps giving a max feed around 2300 as I don't think I'll be needing 4700 mm/min.

---

Be aware that this will put more stress on the parallel port as your needing twice the pulses to travel same distance no matter what the speed so your open to missed steps if the PP struggles at higher feeds.!
Don't think of MS in terms of increasing resolution because in reality at this level it doesn't really it has bigger affect on smoothing the motors. But it does come at a price by putting the system under more stress. With machines running external motion control cards this isn't an issue but it can be trouble some for PP driven machines. Esp if using cheap drives and Bob's etc.!

I'll leave it at 8 then, and enjoy high speed rapids over my 60x60mm workpiece :)

2 new drivers arrived today. Everything now appears to be working. I took one of the dodgy ones apart and the heatsinking was poor - blob of paste covering about 25% of the surface area. A nicely ground bit of aluminium plate between chip and heatink, but it had been chopped leaving a burr which lifts the end!

Before I do any sustained work, I think I will take my good ones apart and ensure a good heatsink connection.

Also spotted one very poor solder connection, so there is a posiibility of fixing the bad ones.


I did take the goods ones apart, and made sure the heatsinking was good. The 2 replacements, and the one original that worked had smaller (but flatter) heat sink pads than the faulty ones. Indicating that the design is at least evolving! The pad edges still needed a clean up though, and the paste was still only 20% coverage.
Attachment 13333

So what was causing the faults / resets when jogging from the keyboard ? did you change the breakout board ?

2x faulty Tongka M542 drivers. It would appear that the original BOB is OK.

YAY! first cnc cuts done.

Another Mach3 mistake though. I set Z high and ran the gcode in free air, double checking the x and y extents weren't going to hit the vice. But the way I stopped mach3, set Z zero and then started again meant that it re-started mid code. Instead if doc 1.5mm with my 3.2mm cutter, it plunged to 9mm and started pocketing. The noise wasn't too bad, I only stopped it because the pocket was filling up with 15mm long strands of swarf all standing up and filling the pocket.

So I went back to line 0 this time and started again, much quieter.

At least I know that cutter is good for more than 1.5mm doc !

feed was 800mm/min, RPM 11500 (need to adjust my vfd as that's the max I can get)

Somewhat risky for a newb like me to have metal screws in the work area, but I took a chance. All went well until I hit the z-axis limit switches (piss poor planning) in the middle of my job. In the confusion I did the same as before, started the job mid program but I'd done a goto zero. I should have noticed it was heading off inthe wrong direction but I wasn't ready ready to pounce onthe stop button.

The tool scored a direct hit on a holding down screw.

Hopefull no damage to my spindle.Attachment 13240Attachment 13241


All seemed to be going well but I've got one concern. The third pocket, is undersized slightly eg 13.85mm instead of 14.0mm in the long axis direction. Does that indicate missed steps?

I ran it again at 70% feed and another sliver got taken off one face, so now 0.10mm undersize rather than 0.15mm.

I think I understand the undersize problem on long axis, I think it comes from slack on the long axis bearings. I can feel slack when I rock the gantry back and forward (on top of the flexibilty).

The setup is SBR20 round supported rails with matching bearings. I don't suppose they are adjustable? or maybe I can buy adjustable repalcements?

Hi,

Quote:

---

Originally Posted by jimbowley

Somewhat risky for a newb like me to have metal screws in the work area.

---

whenever I'm not sure if everything will be OK I put some styrofoam on the machine (with doublesided tape) and add a mechanical stop to the z-axis just before the tool touches the machine bed. This way you don't need to worry about breaking anything even if things go completely wrong.

Christian

Quote:

---

Originally Posted by jimbowley

The setup is SBR20 round supported rails with matching bearings. I don't suppose they are adjustable? or maybe I can buy adjustable repalcements?

---

Some of the SBR20 bearings have a little grub screw in the bearing housing which will let you take some play out but still not good ideal.

Unfortunately this is what you get with these low spec machines so you either accept and work around or start going down the upgrade route which soon gets you to the point of diminishing returns. Problem is you quickly get to the point where the machine structure becomes the weak link and then you may as well just start a new build as it will be cheaper and probably easier.

Wrecking cutters and material is just part of the learning curve and after a few cutters and sheets you soon start paying more attention to what the machines doing and reacting quicker. To be honest even now after running millions of lines of G-code if I break cutters it's nearly always from clipping something like a screw or clamp on way to home or some fixture offset.
One of the common Gotcha's for new users is when the machine goes homes or Rapid moves to some point on table it doesn't always take the route you expect. This is true when moving two or more axis at same time to a point on table. The motion planner will interpolate the path and move at feed rate so all axis arrive at same time.
So what happens is that while you think it will clear clamps because at the commanded feed rate for that axis it would be no where near the clamp the planner actually slows or speeds up that axis to ensure all axis arrive at same time. This catches many out, esp with Z axis because when retracting the other axis are moving fast because they have a long distance to travel but the Z axis gets slowed right down to ensure they arrive at same time. So the cutter doesn't clear clamps or material like you'd expect..!! . . . . . . This is why you should have Safe Z enabled in Mach3 when homing but in G-code it doesn't apply so can Catch you out if code is produced this way.!

Well inspection of the linear bearings revealed that one grub screw was loose, and one was missing. It may have shaken loose but I couldn't see it on the floor.

Quick tighten up (no idea how tight to do these so I just pinched it up) with M6 screw inserted for the missing grub screw. And now the slack has gone :tennis:

The flexibilty remains, but that can be managed.