Regarding your issues with the reliability of the NEMA 23 motors: What are the specs on your NEMA 23 motors? It may be that the gShield can't supply the amperage they need. My suspicion is that that isn't the problem, but the specs would help others on the forum home-in on the problem.

Note: I have almost no idea what I'm talking about. At this moment, my OX build consists of an Arduino running GRBL hooked up to a laptop. But I have opted to use GRBL, without a shield, using external TB6600 drivers.

 

I noticed that the gshield only has three axis control which means that you probably connected 2 of the motors to the same axis. Even though the driver is supplying the 2.5A each motor only sees 1.25A each. This is a problem if you are trying to move too large of a load which I assume is what is happening. Also, the driver datasheet says the following.

If the current through any FET exceeds the preset overcurrent threshold, all FETs in the H-bridge will be disabled
for a period of approximately 800 µs, or until the ENABLEn pin has been brought inactive high and then back
low, or power is removed and re-applied. Overcurrent conditions are sensed in both directions; i.e., a short to
ground, supply, or across the motor winding will all result in an overcurrent shutdown.


Basically what this is saying, is that if you are trying to drive your steppers with too much current, the driver will shut it self down and start back up after 800 us. This could cause your sputtering too. I've never used a Gshield but check to see if you can set your maximum current on the driver. I would set it to 2.0-2.4 amps. As long as you stay away from the max 2.5A the over current protection shouldn't kick in and you should have much better consistency.

To solve this, you will need a fourth axis to run your two steppers on the x axis as well.

Are you single stepping right now? If you don't have enough torque to each motor, you may try microstepping the stepper. Although it will slow down your CNC, it will provide slightly more overall torque to move your gantry.

I don't know if this is the problem, but it seems like this could be the issue. Please let me know how everything works out and I hope this was at least some help!

 

And there you go. Problem solved and I learned something.

I thought the same thing about using external drivers. But it doesn't seem that bad to me now. Just have to clean up the power (which I don't yet understand at all) and use a resister depending on your supply voltage. If I somehow manage to pull it off, it will be a sign that nearly anybody can.

 

I wouldn't use a resistor. THe Gshield is rated for a voltage input from 12V to 30 V Your power supply is at 24V which is just fine. The resistor would just add unnecessary power dissipation in your system.

 

Remember that I'm using external drivers. Sure, if you're using a GShield, you have no worries.

The need for the resistor came from the EBay page on the TB6600 drivers I bought. Specifically it says:

Input interface circuit (common cathode connection)
- Note: the VCC is 5V, R short；
- VCC value of 12V, R 1K, more than 1/8W resistor;
- VCC value of 24V, R 2K, more than 1/8W resistor;

I will definitely be looking closely at more complete specifications before wiring the drivers up, but the above is where my (mis)understanding is coming from.

 

Ahh I see what your saying. Its for the Vref into the TB6600. I was under the impresssion you had a resistor in series with the power supply. This makes more sense. Sorry for the confusion.