Help debugging dual stepper y-axis losing square

Post Reply
priority
Posts: 4
Joined: Sat Feb 22, 2020 8:01 pm

Help debugging dual stepper y-axis losing square

Post by priority » Wed Apr 15, 2020 1:59 pm

I have a v2.60 board connected to 1.5m x 1.5m Workbee with dual y stepper motors. The y-axis is cloned to the a-axis via solder bridges on the bottom of the v2.60 board.

I initially was running with DRV8825 drivers, but would notice that my y-axis was losing squareness to the x-axis. Initially suspected the DRV8825s were having problems. A bit of searching found substandard counterfeit chips can cause problems. So, I moved to external TB6600s.

The y-axis is still losing square relative to the x-axis, but the situation is better. Before, the loss of squareness would sometimes be visible in minutes of cutting. Now the relative shift of the y- and a- stepper motors is about 2-3mm/hour based on my most recent long CNC runs.

I checked all the wiring and made sure all screw terminals were fully tightened down.

Resistance of all the coils in the steppers (measured from the unplugged connector that goes the TB6600's) on all 4 stepper motors was 0.5 to 0.7 ohms.

Micro-stepping has always been set at 1/8.

It has been cool in my garage (40-50F/4.5-10C) the past few days and fans are directly cooling the Raspberry Pi/v2.60 board and TB6600's.

My questions are:
(1) What I am I missing that could be causing my y-axis to slowly lose squareness?
(2) Ideas on how to debug/isolate the cause?

Thanks!

Bertus Kruger
Site Admin
Posts: 1649
Joined: Wed Feb 03, 2016 10:26 pm
Location: Wellington , New Zealand
Contact:

Re: Help debugging dual stepper y-axis losing square

Post by Bertus Kruger » Wed Apr 22, 2020 2:46 am

That is a hard one.

DRV8825 run hot. So it takes only one of them to cut out for you to lose squareness.

My X-Carve had the same setup , but because its belt driven I think it self corrects when the motors are not energized. Not the greatest but seems to work.

On the other hand my one pick and place machine has one Y stepper motor with a shaft from one side to the other side of the machine. From that shaft it then has pulleys setup to drive a belt on each side.

The only solution I can think of is slowing down your machine or increasing the micro stepping. That way you prevent losing steps

priority
Posts: 4
Joined: Sat Feb 22, 2020 8:01 pm

Re: Help debugging dual stepper y-axis losing square

Post by priority » Wed Apr 22, 2020 4:35 am

Thanks for the ideas.

After lots of testing and tweaking since my original post, I am convinced I was having 2 different problems.

The DRV8825s were simply not reliable for me. I am convinced that they were slowly failing, albeit at different rates on the different axes. I noticed that after several hours of running, when the machine was idling without dust collection or the spindle on, I would hear a faint ticking noise from a stepper motor on one axis; there was no consistency in which axis would have the sound. Changing out the associated DRV8825 with a new one would seem to correct the ticking sound.

So, moving to the TB6600's (purchased them from one of your US distributors--Zyltech) got rid of the slowly, but surely failing DRV8825s.

The second problem was the amount of force being placed on the Workbee's v-slot wheels by the eccentric nuts. Simply stated, I had too much force on the wheels. In my zeal to eliminate slop in the machine from tolerances in the v-slot and v-slots wheels, I had tightened the eccentric nuts too much.

Aside the small amount of slop that has returned to the machine from properly setting the force generated by the eccentric nuts, my machine is now running with 1/8 micro-stepping and achieving reliable performance when machining MDF with a 1/4" bit at 90-100ipm with an 1/8" cut depth. The leadscrews are held in tension as documented on youtube by Sean Martin.

I should note for others that read this, I have what the various Workbee kit vendors sell as "standard" NEMA 23s with 1.26 Nm of holding torque. In hindsight, I should have purchased the "High Torque" steppers for a machine of this size and started off with the TB6600 drivers.

Bertus Kruger
Site Admin
Posts: 1649
Joined: Wed Feb 03, 2016 10:26 pm
Location: Wellington , New Zealand
Contact:

Re: Help debugging dual stepper y-axis losing square

Post by Bertus Kruger » Fri Apr 24, 2020 2:58 am

priority wrote:
Wed Apr 22, 2020 4:35 am
Thanks for the ideas.

After lots of testing and tweaking since my original post, I am convinced I was having 2 different problems.

The DRV8825s were simply not reliable for me. I am convinced that they were slowly failing, albeit at different rates on the different axes. I noticed that after several hours of running, when the machine was idling without dust collection or the spindle on, I would hear a faint ticking noise from a stepper motor on one axis; there was no consistency in which axis would have the sound. Changing out the associated DRV8825 with a new one would seem to correct the ticking sound.

So, moving to the TB6600's (purchased them from one of your US distributors--Zyltech) got rid of the slowly, but surely failing DRV8825s.

The second problem was the amount of force being placed on the Workbee's v-slot wheels by the eccentric nuts. Simply stated, I had too much force on the wheels. In my zeal to eliminate slop in the machine from tolerances in the v-slot and v-slots wheels, I had tightened the eccentric nuts too much.

Aside the small amount of slop that has returned to the machine from properly setting the force generated by the eccentric nuts, my machine is now running with 1/8 micro-stepping and achieving reliable performance when machining MDF with a 1/4" bit at 90-100ipm with an 1/8" cut depth. The leadscrews are held in tension as documented on youtube by Sean Martin.

I should note for others that read this, I have what the various Workbee kit vendors sell as "standard" NEMA 23s with 1.26 Nm of holding torque. In hindsight, I should have purchased the "High Torque" steppers for a machine of this size and started off with the TB6600 drivers.
Thanks for sharing your solutions... ;)

Post Reply

Who is online

Users browsing this forum: No registered users and 0 guests