Centroid Community CNC Support Forum

I'm having issues the Y axis of my router.

It's using 1.8 kw AC servo's and drives off Amazon.

The machine had been running great for the last 11 months. About a month ago I noticed that if it sat on for awhile, the Y axis would occasionally be slightly off from where the DRO was reading.

Today I was setting my work offset. A couple seconds after zeroing the Y axis, the machine moved in the Y direction by about .2". I jogged that axis back a forth and every couple of times, the axis would move slightly a few seconds after the command was executed.

I replaced the encoder. No change.

I swapped the wiring from my acorn to the drivers. Jogging my X axis is now moving my Y axis. I figured the driver would be the next place to check. I'm getting the same result with the Y axis using a different driver.

I'm now guessing that there is a problem with the pins since both drivers are experiencing the same issue when wired to the Y axis pins on the acorn board. Any other insight would be welcome.




EDIT:
I had another Acorn board I was planning to use for a different project. Plugged it in and the problem went away. Seems my other board was bad...I'd been using it for one year an two weeks.

Do you have access to an oscilloscope? A motor moves when it receives electrical signals telling it to move. I'm guessing that you have significant noise on the step signal. A 'scope will show that noise, then you can turn off / move other signals until you've found the source of the noise.

Richards wrote: ↑Thu Jan 07, 2021 1:49 pm Do you have access to an oscilloscope? A motor moves when it receives electrical signals telling it to move. I'm guessing that you have significant noise on the step signal. A 'scope will show that noise, then you can turn off / move other signals until you've found the source of the noise.

I don't have one. The machine worked great for almost a year and recently started having this issue. It wasn't driver specific and the problem seems to be cured once I swapped out the original board.

Your report will help those who know how to read them help you find the source of the problem.

Can you attach a photo with clear detail on how things are wired? Someone might see something that might help. There is always a possibility that a board is faulty, but I haven't read about Acorn boards sending out extra step signals on the forum. When I look at the step pulse timing requirements, I'm inclined to think that something inside your system is generating noise that your current Acorn is picking up or noise that is being induced into your wiring. When I received my first Acorn board, I hooked an oscilloscope to the board and was very impressed by the clean signals that the board produced. I haven't done that on the other two boards, but they have been rock-solid.

Back in the early 1980s, a client hired me to fly to Tennessee to help him troubleshoot a machine that he had modified. When I got there, even before we took the cover off the electrical controller, I could hear the problem. He hadn't told me that he was using relays on the telephone call, but I could hear them clicking. He had build a controller using relays that were being driven by 74xx chips, chips that worked using TTL signal speed and TTL signal levels. The mechanical relays had 5V coils, but no suppressors. So, every time a relay changed state, it produced electrical noise that the fast-acting 74xx chips picked up as a signal. I taught him about electrical fields caused by the electro-magnets in relays and about 74xx logic and why the two didn't work together. When he understood the source of the problem, he replaced the relays with solid-state relays and everything worked.

So, when you wrote that your machine was moving on its own, I immediately thought of noise being induced into your wiring.

Mike, apologies for the dumb question but how exactly do you connect (which terminals) an oscilloscope when tracing interference ? (either end of each circuit etc?) I don't have any problem on either of my machines, just interested

Nigelo, one lead on an oscilloscope is connected to GRN,; the other lead is touched to a terminal or connection point of the signal being tested. The oscilloscope must be configured so that the voltage range being checked fits on the 'scope's display screen. The frequency range must be configured to show the normal wave form being tested, then you can see any noise on that signal. I usually set the frequency so that I see at least five steps from the stepper motor when the motor is at full speed. One thing to keep in mind, every oscilloscope that I've owned needed a "dummy" AC adaptor connected to its AC Line plug. The adaptor bypasses the GND connection. I've blown a few fuses when I connected the oscilloscope without using a dummy plug. There was a time when it was considered bad practice to connect DC GNDs to AC GNDs. In fact, I was surprised when Geckodrive told its users to always connect all GNDs in a system to a common point. I've been connecting all GNDs together since then with excellent results.

That's excellent Mike, very many thanks
Nigel