Debugging GRBLDuino issue

Use, set up, assembly help, and any other GRBLDuino related questions or comments.
dan.taylor
Posts: 7
Joined: Mon Oct 05, 2020 9:46 am

Debugging GRBLDuino issue

#1

Post by dan.taylor »

Hello,
I'm a programmer and woodworker with almost zero hardware experience, trying to navigate as needed to build a CNC rig. I will not be shocked to learn I've done something wrong, but could really use some help in figuring out what that might be.

Problem: The stepper motors won't move. Any commands to move any axis results in no motor response.

Here's my setup
- GRBLDuino v2 shield
- 24v Meanwell power supply
- Elegoo Uno
- 4 Stepperonline stepper motors
- Pololu TB67S249FTG stepper drivers (https://www.pololu.com/product/3097)

What I've tried:
0. I've flashed the arduino with GRBL 1.1 using the wiki instructions. I'm using CNCJS on an RPi, but bypassing most of it to use terminal to try to rule out any CNCJS specific causes.
1. I'm able to send GRBL commands to the arduino and see results in the terminal.
2. I've verified that 24v is actually being provided at the GRBLDuino motor power terminals, and they are at the correct polarity.
3. I've verified that the drivers are oriented the proper direction in their sockets
4. I only have the motors hooked up, no limit switches or anything else
5. i've tried testing voltage on X (a1,a2,b1,b2) but am not seeing any. I'm not confident I'm coordinating this properly with the multimeter and commands and such, though
6. as far as I can tell the shield can only fit one way onto the arduino. As far as I can tell it is seated properly.

I've attached a pic of my board.


Are there some easy steps I can do to verify the board is ok, the drivers are functioning properly, etc? If this was software I'd know how to narrow it down and debug, but I lack the knowledge of how to do that in hardware.

Thanks for any help!!!
Dan
Attachments
CNCShield.jpg
CNCShield.jpg (280.29 KiB) Viewed 6105 times
CNCShield.jpg
CNCShield.jpg (280.29 KiB) Viewed 6105 times
Last edited by dan.taylor on Tue Oct 06, 2020 6:58 pm, edited 1 time in total.
User avatar
EccentricWorkshop
Admin
Posts: 38
Joined: Thu Jan 17, 2019 9:33 pm
Location: Olympia, WA
Contact:

Re: Debugging GRBLDuino issue

#2

Post by EccentricWorkshop »

The first question is have you set the current limit on the drivers? There is a section on how to do it on the Pololu page (first link on the Resources tab under Recommended links). If the steppers aren’t making noise or jittering, it leads me to think it’s a power issue. Do the steppers lock up on start up? Can the power supply provide enough current to drive the system?

Can your steppers work with 24v? Not likely the issue but it doesn’t hurt to check.

It is possible the drivers are dead. If you unplugged either the driver or the stepper while powered or had them inserted upside down, they will be dead.

The steppers should lock up and move without limit switches connected.

The shield does fit only one way and it will seat all the way down into the sockets. Unfortunately, your image isn’t loading so I can’t see it (you may need to attach it to the post so it is hosted here; this also ensures the images live with the forum and persist despite external changes).

Let’s work through what we can to see where you get.
--------------------
Andy
Eccentric Workshop
Store - Tindie - eBay
dan.taylor
Posts: 7
Joined: Mon Oct 05, 2020 9:46 am

Re: Debugging GRBLDuino issue

#3

Post by dan.taylor »

Hi Andy!
thanks for the ideas, answers in order:


- current limit is set to 2.06 for all 4. The stepper motors are 2.8A, and that's (2.06) what I got running the VREF numbers for the TB67S249FTG.
- you are correct, no noise, no jittering. I see no movement or locking behavior on startup, though this is my first project of this type and I might not know what to look for.
- power is 24v, which is allegedly a good middle ground on what the motors can handle. 12v is considered too low, 40 is on the high end from what I can find.
- I was careful to put them in the right way and while UNpowered. Is there a way to test if they are dead? I'm able to sense VMOT and VREF with multimeter, and those values seem right. I have a spare driver that I've carefully subbed in at intervals as a control, but its behavior is the same. Once again, I'm hammering at the wall of my ignorance on this one.
-glad to hear on limit switches. I want ahead and tested with the jumper in the Low configuration to the 3-pin limit configuring part of the board, no effect. Sounds like that wasnt necessary
- I appended the image, you should be able to see it now

Also, I've got the microstepping set to low, low, high(M2) for each of them, which appears correct for this driver. I had them all low for a bit, which is standby for this driver, whoops.
I have a set of DRV8825s coming tomorrow as another control test, as those seem a lot more common.

I'm also going to be near a Fry's tomorrow, if there's a tool that'd help narrow this down that isn't too pricey, I can pick something up.
Thanks!!!!
Dan
dan.taylor
Posts: 7
Joined: Mon Oct 05, 2020 9:46 am

Re: Debugging GRBLDuino issue

#4

Post by dan.taylor »

Minor update, some coworkers noticed a nearly imperceptible motor twitch when I turned things on today to show them. I think that means there's at least a path for the current.
dan.taylor
Posts: 7
Joined: Mon Oct 05, 2020 9:46 am

Re: Debugging GRBLDuino issue

#5

Post by dan.taylor »

More testing info:
- Voltage on Fault, Enable seem fine and indicate no problem
- I subbed in a set of DRV8825s (power off, orientation confirmed), no change in behavior
- Instead of using the CNCJS terminal, I used putty to directly connect and attempt movement. No movement observed.

I have a logic probe and a few other tools coming by mail tomorrow, hopefully can get some better insight with those.
User avatar
EccentricWorkshop
Admin
Posts: 38
Joined: Thu Jan 17, 2019 9:33 pm
Location: Olympia, WA
Contact:

Re: Debugging GRBLDuino issue

#6

Post by EccentricWorkshop »

Sounds like a bit of progress. It is starting to sound like a power issue. Have you tried turning the current adjustment potentiometer while trying a movement? Certainly not the best way to set things, but it is a good way to diagnose. It may also be the microstep settings are requiring more current than can be provided.

From what I've experienced, failed drivers have sent the full stepper motor voltage into the 5v rail and USB port so you would know instantly if one failed like that. Otherwise, they could certainly fail with no sign but that would not give you the slight twitch on startup.

What were the voltage readings you got on enable and reset? Enable should be near 0v and reset near 5v. Fault won't be of much use since the board is designed for A4988 compatible drivers which has Vdd where the TB67S249FTG has fault so it will always show nearly 5v (Pololu says this isn't an issue because there is a protection resistor in place to allow for compatibility). One more thing is the A4988 has sleep and reset connected together to get reset high and enable the driver; on the TB67S249FTG, sleep is replaced by AGC or active gain control. I'm not familar enough with these to know if that could be interfering. It may be worthwhile switching to your DRV8825s until everything is confirmed to be working.

Are you able to go direct to the Arduino without the Pi and try something like Universal G Code Sender or Candle? It shouldn't make a difference but reducing variables and ruling things out is helpful.
--------------------
Andy
Eccentric Workshop
Store - Tindie - eBay
dan.taylor
Posts: 7
Joined: Mon Oct 05, 2020 9:46 am

Re: Debugging GRBLDuino issue

#7

Post by dan.taylor »

Diagnostic progress!!!

after many hours of testing, I happened to push a little harder with the multimeter probe when again testing the shield incoming 24v DC voltage via the top of the screw terminals. Suddenly, the motors started moving, shuddering, etc. Just wiggles, but it was constant. Here's the positive terminal/corner I was pushing on
ShieldPowerTerminal.jpg
ShieldPowerTerminal.jpg (45.31 KiB) Viewed 6059 times
ShieldPowerTerminal.jpg
ShieldPowerTerminal.jpg (45.31 KiB) Viewed 6059 times

I then figured out I could squeeze that corner and get responses from the motors. There was occasional arcing/popping and a little smoke at the positive connection point, but motors kept twitching, so.....progress. Repeated attempts at re-seating the wire in the screw terminal didn't change the behavior.


I pulled the shield and looked underneath, and it seemed like the pin from the screw terminal was floating in place without solder? I found some solder and a solder iron and did a terrible job (I'm very new at this) of smearing a blob around that pin.
Here's a pic before I carefully vandalized it:
ShieldUnderside.jpg
ShieldUnderside.jpg (1.15 MiB) Viewed 6059 times
ShieldUnderside.jpg
ShieldUnderside.jpg (1.15 MiB) Viewed 6059 times

After that, behavior was (unexpectedly) still the same. No motor response to anything until I squeezed that corner, then arcing/popping and motor twitching while I held it. Could this be an issue with the factory soldering, or something broken inside that part of the screw terminal component? It seems like there's a small gap somewhere that I'm closing when I squeeze. Is there some way I can rule that out, if not?
Thanks!
Dan
User avatar
EccentricWorkshop
Admin
Posts: 38
Joined: Thu Jan 17, 2019 9:33 pm
Location: Olympia, WA
Contact:

Re: Debugging GRBLDuino issue

#8

Post by EccentricWorkshop »

dan.taylor wrote: Sat Oct 10, 2020 1:51 pm After that, behavior was (unexpectedly) still the same. No motor response to anything until I squeezed that corner, then arcing/popping and motor twitching while I held it. Could this be an issue with the factory soldering, or something broken inside that part of the screw terminal component? It seems like there's a small gap somewhere that I'm closing when I squeeze. Is there some way I can rule that out, if not?
Well, there's no factory to blame here - I assemble all of the boards myself so I'm not sure what happened with yours. It's unlikely there's something broken inside the terminal but I'll get a new one shipped out to you tomorrow. Let's just rule out the board itself completely and start from there.
--------------------
Andy
Eccentric Workshop
Store - Tindie - eBay
dan.taylor
Posts: 7
Joined: Mon Oct 05, 2020 9:46 am

Re: Debugging GRBLDuino issue

#9

Post by dan.taylor »

That sounds great. Thanks so much!
If the replacement works out, I'll mail your first one back for an autopsy, if you'd like? I'd love to understand if I somehow did something, or if there's a better way for me to diagnose this type of issue in the future!
dan.taylor
Posts: 7
Joined: Mon Oct 05, 2020 9:46 am

Re: Debugging GRBLDuino issue

#10

Post by dan.taylor »

I'm considering soldering the positive wire to the underside of the relevant shield terminal, just so I can hopefully continue debugging my setup the next few days while the new board in in transit.
Might this approach work, or would you advise against it?
Post Reply