Stepper Heating / Board Fail Potential

[3dGriff]

Over the last few weeks I have been working with a user in NC who is doing a retrofit using a SainSmart 2in1 control board.

Last week (March 24, Friday) he reported a failure shown in the attached image. The failure was a total loss of the board and appears to have impacted the FET regulator used to regulate power to the stepper drivers.

note failure on U8 FET

Yesterday (March 27, Monday) it was reported internally that elevated heat levels here being observed on the stepper heatsinks. The design specs for the DRV8825 stepper drivers do not provide any exact numbers for max temps, however, it does state that "this product can get hot enough to burn you long before the chip overheats".

In the past, heat has not been a concern on the stepper drivers, it is possible that the latest firmware build is causing issues.

To ensure this does not happen to our users, I think it wise to develop a study on this issue and try to identify impact and root cause.

@BotleHewer any thoughts much appreciated.

I Propose the following

Control Info

• Current ambient temp in room reported
• Current print speed reported
• Current vRef for each driver reported
• Current Movement and Acceleration settings reported

Testing Process

-Read driver temps prior to printing (5 readings and average)

• Print the CaliCat with the following specs: -- .2mm height -- 35mm/s speed

• Read driver temps at 10min intervals during print (5 readings and average)

• Report results

Finding the source

• I will be making a few modifications to the current firmware and doing the same tests, the goal here is to see if we can ID a causal link.

Any input on this would be great.

[3dGriff]

Test 1 Results For Ref

Conditions

• Abmbient Temp 69F (21C)
• Bed Temp: 60
• Extruder: 200
• Material: PLA
• Speed: 35mm/s

Main Power Input: 12.09v

Vref

• E - 0.62v
• X - 0.50v
• Y - 0,59v
• Z - 1.20v

Print Time: 55:32

Raw Temp Readings:

Temp Plot:

Results:

As expected the Z axis is the hottest of the drivers, looking at the data I can rule out the increase in Z movement due to ABL Z correct as the Z Temps do not decrease in any meaningful way after the z fade concludes. After 10mm layer height the z corrections are no longer impacting the driver, however, we see no change in the temps at this point.

These temperatures are not alarming in any way. Although this test was done at a typical high quality print speed I don't see anything significant to suggest major spikes at higher speeds. I will follow up on this as well.

What's more important is that the U8 FET in question maintained a steady temperature between 38-43c. This suggests that while operating within the specified voltage ranges we are not causing undue stress to the regulator.

[BotleHewer]

Here is what Angi reported during a spot check on her bot several hours into a print.

Room Temp 69-70* F Print speed 84% 60mm/s support 45mm/s perimeters 60mm/s infill

X axis 63C Y axis 63c Z axis 60c Extruder 44C

[Angi-B]

second reading 2:45 same print X axis 57C Y axis 76c Z axis 64c Extruder 49 C

[3dGriff]

@BotleHewer @Angi-B thanks for this, good so see it seems to be holding at higher speeds. I think for now we are looking great as far as tolerances. I wasn't able to find anything explicitly saying what the failure temp is on the DRV8825's but based on other systems I might estimate it in the mid 100c's... For now we can move forward without any active cooling on the board. The Sainsmart was designed with high heat tolerances in mind.

Consideration and further reading.

The new latest firmware defines a minimum step pulse of 4us this was done to accommodate the mininimum pulse width of 1.9us on the DRV8825 whereas the standard A4988 drivers utilize a minimum pulse width of 1us.

I will move forward in testing the effects of reducing this to 2us in order to more closely match the driver's specs.

@BotleHewer I am interested to see your results on this test as your machine is still utilizing the default step width of 1us.

References

If you wish to know more about this check these links out...

https://www.pololu.com/product/2133

https://en.wikipedia.org/wiki/Pulse-width_modulation

[BotleHewer]

@3dGriff I don't own a laser therm. I'm due to get one from my father when he sells off his shop but that could be months away. Will see if I can borrow Jeff's for a bit to confirm, but I'm definitely not seeing danger temps at this time

[3dGriff]

Standing solution seems to be adding space above the drivers on the control board lid. Carry on in issue #46

[3dGriff]

Leaving issue open for now, please report any issues that may be related to this one.

[3dGriff]

Closing due to inactivity