X axis moving when not being used?

Hi All,

Let me clarify....

I have just rebuilt my Computer following a hard-drive failure. I had been sensible in that I had all my settings and xml file backed up to another drive so it didn't take long to get things back to normal. I am now running Mach software and have a strange problem.

The machine is running on a HobbyCNC 4 Axis board with a 40v Power supply and a 600mm x 320mm table. Y axis is length of table and X axis is width.

Following the rebuid, I started to test all the axes one by one - just to make sure that all was working as before. I started with the Y axis by using a random Gcode generation program in Excel that produces the code, and literally sends the Y axis up and down the table in irregular patterns from your defined zero for a period of about 15 minutes - eventually stopping at the defined zero spot.

All went well with the 'up and down the table' bit, but I noticed that after the program had been run and ended, the X axis had moved across the table almost 0.25" from its original position!

Now, I have to admit that I dont know if it was doing the same before the hard-drive problem or not - I had never thought to test it this way previously. I have checked the code output from the generator and there are no X moves anywhere!

I find it strange that it is even possible for the X Axis to move with there being no commands in the code to move it - but somehow it has managed to do just that.

Would be grateful for any suggestions on how this might have happened - along with any thoughts on how the problem can be overcome. My initial thoughts are that maybe the currents operating the Y axis are causing the X axis to move via induction into the X axis cable?

I hope someone can help me with this....

willower

It would be unlikely for the other axis to induce movement in your x-axis.
When you are running the tests, is the X-axis 'locked' (you can't move it by hand)? It should be locked.
The inputs are pulled-up with 10k resistors, so it might be possible for some input noise to move the axis (but also unlikely).
I assume no change to your 40V power supply, but you didn't mention the current capability - how is this voltage holding during your test?
Tell me about the parallel cable - same, changed?
What about how your setup is grounded - potential induced noise or ground loop causing noise to the x-axis input?
Is the parallel cable physically close to any of the stepper motor wires?
Are all your power supply connections firm/solid?
BrianV

Hi Brian,
As far as i know, yes the X is locked (powered). I have never checked the 40v supply whist operating - it is a 10A supply so should be sufficient, I will check tomorrow to see what voltage I have whilst running the rendom code on the Y.
The parallel cable is the same as before and it does not run near the motor cables.

Re grounding, the control box is plastic and all grounds are common with mains ground and all power supply connections are solid.

One final thought on the parallel cable, it is really too long for the job as my control box sits on top of my PC. The excess cable length I have wound into a coil and that sits inside the control box on the base - but that coil does come quite close to the HobbyCNC Driver board maybe less than 2" away from the 'Inputs' terminal block i.e. the side furthest away from the motor connection blocks.

Will update tomorrow after checking PS voltage to verify that it is OK.

Richard

How long is your parallel cable? I provide a 6 foot one with my combo kits, and 15 feet would be considered "long" for a parallel cable. The parallel standard is a single-ended, un-balanced circuit that is suitable for short distances only (like from your big-old-desk side computer to your dot-matrix printer. There are longer cables available, but I'd be afraid to go much past 6 feet and not ever past 15. I'm sure others have done it without issue, but I couldn't recommend it.
Can you try your tests with a short-as-possible parallel cable too?
BrianV

Brian,re: your previous comments...

My parallel cable is only 2 metres long and as said, I have the 'excess' length coiled inside the driver housing plastic box - which has never caused a problem previously so I don't think this has any bearing on the problem I have outlined.

I have now checked voltage while under power and all is as expected - 40v give or take by less than half a volt. However, I have now run further tests (using the same gCode file as before) and found no real problems at all - comes back to zero every time. The only thing I can now suspect is that the mains power supply possibly suffered a glitch whist running the program initially after my rebuild of the hard drive. Being summer here in the UK we were having a pretty severe thunder and lightning storm at the time of the initial rebuild test and I now think that this may well have been the source of the problem.

I have now added a surge protector to the input of the whole system (Computer/Power Supply/HobbyCNC board) and run no less than 6 further tests on both the X and Y axes without showing any further discrepancy, so I hope that the problem is now well and truly sorted!

On the other hand, it may have been all those bad vibes we were getting from the EU member States over our Referendum as to whether or not we should leave? Having now 'bitten the bullet' as they say, I think it is abundantly clear they really didn't want us in there with them in the first place (just look at our history with the French and Germans!) - and I for one am happy that the decision has now been made - it's just that we now have a whole new set of problems to sort out (which we will of course), it might just take us a little while to do it.....

Many thanks for your help Brian