[Despontene]

I'm in the process of putting together a kit from Avid CNC (CNC Router Parts). I got everything setup and working, but one thing is vexing me. The machine runs super smooth as long as it's jogging in one direction, but every time a motor reverses, it jerks in the other direction instead of moving smoothly.

I uninstalled a motor to test on my workbench and got the same behavior. Super smooth rotating in one direction, but reversing direction causes the motor to jump. This happens both when I reverse direction immediately and when I wait a bit between direction changes.

Anyone have an idea as to what the problem might be?

My setup:

Linux CNC
Nema 34 stepper motors
Gecko GR214V drivers
Mesa 6I25/7I76 I/O cards

[Sterob]

You may need to try and isolate the problem.
Is it on all axis's or just one?

Move the signal wires for one stepper driver over to a different Output on your Motion board and see if anything changes.
Are you using a ESS? What BOB do you have?

[Despontene]

This happens with all of the motors and each one is on a different output.

I’m using a Mesa 7i76 breakout board.

[Jim Dawson]

Is any backlash compensation input? Maybe in the settings somewhere?

[MGHood]

I have an AVID CNCRP machine I have installed Clearpath Nema 34 servos and have found that if your motor tuning is off you will get violent jerking in the gantry. Try to adjust the acceleration down and see it that helps. Another thing you might try is to release the tension on your motor mounting plates and move your axis and see if you can feel anything that is abnormal. I'm sure that the rocket scientist at CNCRP can assist. You might see if you can speek to Aaron.

MGHood

[MGHood]

I forgot to mention that taking a stepper or servo an a bench and testing it doesn't do that much good. They will both violently jerk and need to be clamped to the workbench. Being your machine is new and unproven to you you should check for mechaanical issues first. Disenguaging the gears on your rack and pinion will more than likely indicate which direction you need to investigate.
MGHood

[Despontene]

Is any backlash compensation input? Maybe in the settings somewhere?

I don't think there's any backslash compensation. There are settings for it in pncconf, but they're all turned off.

[Despontene]

I have an AVID CNCRP machine I have installed Clearpath Nema 34 servos and have found that if your motor tuning is off you will get violent jerking in the gantry. Try to adjust the acceleration down and see it that helps. Another thing you might try is to release the tension on your motor mounting plates and move your axis and see if you can feel anything that is abnormal.

I did try decreasing the acceleration. It didn't seem to do anything, but maybe I should take it super low and see if it helps. I'll try messing with the tension next while I wait for a response from CNCRP, thanks.

I forgot to mention that taking a stepper or servo an a bench and testing it doesn't do that much good. They will both violently jerk and need to be clamped to the workbench. Being your machine is new and unproven to you you should check for mechaanical issues first. Disenguaging the gears on your rack and pinion will more than likely indicate which direction you need to investigate.
MGHood

Interesting. Thanks for the reply.. Why do the motors jerk on direction change when they're not attached to something? From what I know about how stepper motors work, I would have expected a step in either direction to be pretty similar mechanically, so clearly there's something I don't understand about what happens to the motor when the direction changes.

[PCW_MESA]

I'm in the process of putting together a kit from Avid CNC (CNC Router Parts). I got everything setup and working, but one thing is vexing me. The machine runs super smooth as long as it's jogging in one direction, but every time a motor reverses, it jerks in the other direction instead of moving smoothly.

I uninstalled a motor to test on my workbench and got the same behavior. Super smooth rotating in one direction, but reversing direction causes the motor to jump. This happens both when I reverse direction immediately and when I wait a bit between direction changes.

Anyone have an idea as to what the problem might be?

My setup:

Linux CNC
Nema 34 stepper motors
Gecko GR214V drivers
Mesa 6I25/7I76 I/O cards

Sounds a bit like an electrical or perhaps setup issue of some kind...

How are the drives wired to the 7I76s step/dir + and- outputs?

Does is just make a jerk when reversing or run poorly in one direction?

What kind of drives are you using? EDIT sorry I missed that you are using Gecko GRV214s

What step timings are you using?

[GME]

It sounds like an acceleration setting issue. You didn't mention what setting you are using. Try taking it down to 20 or 25 and see what happens. If the acceleration is set to high, it can also affect cut quality. Keep in mind that direction changes are going to be abrupt, and especially noticeable when cutting something like a detailed 3d model. I am in the process of cutting some custom designed Victorian bullseye casing rosettes. The direction changes on the roughing passes, and stepover position changes on the finishing passes are quite noticeable. Like you, I am using NEMA 34s and Gecko 214Vs. However my control hardware and software are different. I'm running a PMDX 126, PMDX 107, ESS Smoothstepper with Mach4. Doubtful that the hardware/software are having anything to do with your issue.

Please report out on what the folks at Avid have to say.

Oh, one other thing. This thread probably doesn't belong in the Gecko forum. Very unlikely this has anything to do with your Gecko stepper drives.

Gary

[Despontene]

How are the drives wired to the 7I76s step/dir + and- outputs?

P1 pin 2 (stp) -> 7I76s TB2 pin 3 (step0+)
P1 pin 3 (dir) -> 7I76s TB2 pin 5 (dir0+)
P1 pin 4 (+5v) -> 7I76s TB2 pin 6 (+5vp)

Does is just make a jerk when reversing or run poorly in one direction?

It just makes a jerk when reversing. It runs smooth in either direction as long as the direction stays the same.

What step timings are you using?

I've tried several different settings and none of them have made any noticeable difference. Here are two sets that I tried:

DIRSETUP = 100
DIRHOLD = 600
STEPLEN = 2000
STEPSPACE = 1000

DIRSETUP = 1000
DIRHOLD = 1000
STEPLEN = 4000
STEPSPACE = 3000

[Despontene]

It sounds like an acceleration setting issue. You didn't mention what setting you are using. Try taking it down to 20 or 25 and see what happens. If the acceleration is set to high, it can also affect cut quality. Keep in mind that direction changes are going to be abrupt, and especially noticeable when cutting something like a detailed 3d model. I am in the process of cutting some custom designed Victorian bullseye casing rosettes. The direction changes on the roughing passes, and stepover position changes on the finishing passes are quite noticeable. Like you, I am using NEMA 34s and Gecko 214Vs. However my control hardware and software are different. I'm running a PMDX 126, PMDX 107, ESS Smoothstepper with Mach4. Doubtful that the hardware/software are having anything to do with your issue.

I tried a bunch of different acceleration settings from 1 up to 60 and none of them really changed the reversing behavior. It's possible that what I'm seeing is normal, but I haven't see this behavior in videos of Avid CNC machines I've watched. Here are a few videos where you can hear the CHUNK sound it makes when the direction changes:

Oh, one other thing. This thread probably doesn't belong in the Gecko forum. Very unlikely this has anything to do with your Gecko stepper drives.

Sorry about that. I looked for a more general help forum, but probably missed it. Which one should I post in? I decided on this one because I thought maybe I had the drives misconfigured or something.

[GME]

. I looked for a more general help forum, but probably missed it. Which one should I post in? I decided on this one because I thought maybe I had the drives misconfigured or something.

A couple come to mind. Either the "DIY CNC Router Table Machines" or the CNC Router Parts subforum. Posting in the Gecko forum isn't likely to get much attention from other Zone members. The two I mentioned gets vastly more traffic, and more likely to have someone stumble across it who can help.

I watched your videos. No, the direction changes definitely do not sound anything like they should. I wouldn't run any more than absolutely necessary for diagnostic/testing purposes. It sounds like something is trying to break, or is already broken.

It sounds mechanical to me. In fact, it sounds like your pinion gear is trying to jump out of the rack (tension screw too loose). It also sounds like something is sticking. Odds are pretty low that you would get the same sounds from all the axes though - your Z axis is on a ball screw after all. Rack & pinion jumping doesn't explain no noise when stopping and starting while going in the same direction.

Is everything lubed, i.e., rack & pinion, linear rail bearing blocks, Z axis ball screw?

If you haven't done so yet, I suggest you send the videos to Avid for their evaluation and advice. However, if they believe it's something to do with the electronics (I cannot imagine what could be going on with the electronics), they may not help you. They have great customer service for the products they sell. Maybe not so good with the products they don't sell???

Sorry, but I'm tapped out on this one.

Gary

[MGHood]

You are welcome. My CNCRP clunks less with the servos than the steppers that I originally received with the machine. It seems odd that your issue is with Z and Y+A. I am wondering if your drive belts are loose until I remembered that the Z axis is direct drive. Just FYI I don't remember the instructions saying to grease the shoulder bolts (Axels) on the X and y+A drive sprockets and should put some wheel bearing grease on them. By the time I figured this out I had to replace the sholder bolts because they wore out. I'm curious as to what this jerking might be and let all of the responders to your message know what you find.

MGHood

[PCW_MESA]

P1 pin 2 (stp) -> 7I76s TB2 pin 3 (step0+)
P1 pin 3 (dir) -> 7I76s TB2 pin 5 (dir0+)
P1 pin 4 (+5v) -> 7I76s TB2 pin 6 (+5vp)



It just makes a jerk when reversing. It runs smooth in either direction as long as the direction stays the same.



I've tried several different settings and none of them have made any noticeable difference. Here are two sets that I tried:

DIRSETUP = 100
DIRHOLD = 600
STEPLEN = 2000
STEPSPACE = 1000

DIRSETUP = 1000
DIRHOLD = 1000
STEPLEN = 4000
STEPSPACE = 3000

Not sure if if its the cause of your issue but the step/dir wiring is not correct, it should be

P1 pin 2 (stp) -> 7I76s TB2 pin 2 (step0-)
P1 pin 3 (dir) -> 7I76s TB2 pin 4 (dir0-)
P1 pin 4 (+5v) -> 7I76s TB2 pin 6 (+5vp)

This is because the GR214Vs step/dir inputs are active low
wiring the step pin with the wrong polarity can cause violations
of the step/dir hold times

[Despontene]

Not sure if if its the cause of your issue but the step/dir wiring is not correct, it should be

P1 pin 2 (stp) -> 7I76s TB2 pin 2 (step0-)
P1 pin 3 (dir) -> 7I76s TB2 pin 4 (dir0-)
P1 pin 4 (+5v) -> 7I76s TB2 pin 6 (+5vp)

This is because the GR214Vs step/dir inputs are active low
wiring the step pin with the wrong polarity can cause violations
of the step/dir hold times

Thanks. I switched your pinout. I'm glad to have it configured correctly, but unfortunately it didn't fix the problem.

[CoAMarcus]

Despontene,

How old are your GR214Vs? We have seen some scenarios with a current spike on direction change with the GR214V with certain motor and power supply matchups. This is with the newer version of firmware only, though, and it only affects the initial DIR switch. We downgraded the firmware until we track down what is causing the 5% current spike that makes that noise, and I wonder if you have that firmware. There is no damage that can happen but it can be irritating to hear; can you check the date of manufacture sticker on the PCB and let me know what it is? If this is indeed the cause we can send you three replacement GR214Vs with downgraded firmware.

[GME]

My sincere apologies. I didn't believe the GR214Vs could be implicated in symptoms presented. Obviously, I was wrong. Equally obvious, the OP opened the thread in the right forum and would probably have missed getting the right answer had he posted in a different one.

[Despontene]

@GME No problem. I didn't know for certain it was a driver issue either. Posting here was a bit of a shot in the dark.

Some new drives are on the way, hopefully that turns out to be the issue : )

[CoAMarcus]

Some new drives are on the way, hopefully that turns out to be the issue : )

Hopefully that will get you squared away. I always prefer to err on the side of caution, and this should get you back up and running quickly if the drive is indeed causing the problem.