I am trying to bring up a MP6602 in a pretty basic configuration and am surprised to find that I can’t increase current (using the IS registers) beyond around 300mA stationary without triggering an over current condition. Even if I set current to the lowest setting of 125mA (IS=0, IL=0), as soon as i try to step the motor it trips an over current fault.
The fault varies, most often it is OCPAH, sometimes OCPBL, OCPAL or OCPBH.
Setup is:
10-12V Vin with 3.3V digital supply.
SPI control.
EN, Sleep, Reset, Fault are all open-circuit (relying on the built-in pull resistors).
Motor is rated for 1.2A
Desired drive current is 1.0A
100uF input capacitance with VG and VCP capacitors connected correctly.
Tested two PCB and multiple motors with same issue.
Previously used DRV8825 with absolutely no issues.
Looking for suggestions on what to check next.
ADDITIONAL INFO:
I tried adding 1 Ohm resistors on all 4 motor connections (Bipolar). This helped only a tiny bit, allowing me to increase current to around 600mA when motor is stationary, but it still trips over-current as soon as I try to take a single step, even at 125mA setting with full step enabled (ms = 0).
Confirmed that current through motor is at or below the setting, no where near the 10 amps it should take to trip the current fault.
Confirmed that with motor removed, all four motor lines toggle between Vin and Ground.
Tried lowering Vin from 10V to as low as 5V. Was able to increase current a bit (or or two steps) at lower voltages.
Added more bulk capacitance, up to 1000uF to Vin, didn’t make a difference.
-Connecting either motor winding ONLY to BOUT pins - works if the 1 ohm resistors are in place, full current, but without resitors I can only get around 500mA.
-Connecting either motor winding ONLY to AOUT pins, with or without resistors - Doesn’t work, trips over-current protection even at lowest settings when you try to step.
so this seems to be isolating the issue to the AOUT (even though both connections are just pure copper right to the chip.
One more addition:
I noticed that after soldering the chip works much better, and as it cools down the trip changes.
Hot chip = working
Room temperature = doesn’t work
Tried replacing the chip, exact same effect.
Fet Drive voltage is 4.85V above Vin (i.e. it is about 15V when Vin is 10V)
Eval should arrive today. To be honest, I am kinda disapointed how much MPS charges for their eval boards which is why I didn’t buy one (it was cheaper to make my own).
I have compared layouts (as mentioned) and other than a few vias and minor copper widths (which might affect operation at 4A, but shouldn’t cause a drop to 0.125A, the layouts are similar.
I see that you must have received the EVB by now. Have you seen similar performance after testing with it when compared to the previous behavior you were seeing?
To fully address everything that you have discussed, I would like to ask and verify the following things about your MP6602 design that could be a root cause as to why overcurrent is tripping:
For the bulk capacitance that you have in your original layout that you may see that is reflected in the EVB (in regard to input capacitance placement), you will need to be sure that those capacitors are placed as close as possible to the IN pin of the MP6602 where the return path to ground is short and direct.
Ensure that this bulk capacitance has a low ESR.
Ensure that VCC is decoupled with at minimum a 100nF ceramic capacitor.
VG and VCP require a 16V rated 1uF ceramic capacitor, not 100uF as specified in the datasheet. Although this is just an inconsistency and perhaps not a large cause for OCP faults for occurring.
The fact that AOUT and BOUT varies to a high degree may suggest that there is with the A channel gate drive. Soldering inconsistencies can lead to a higher leakage, so it could be worth monitoring the AOUT gate voltages during operation to see how well they match BOUT.
Since you are using open circuit control, there is a chance that high frequency noise may be coupling in these pins, so it may be worth externally pulling the EN, Sleep, and Reset pins externally. The Fault pin is an output so this should be ok to not externally connect.
Let us know in this thread if there are any updates in operation. Significant waveforms showing the issue or even aspects to check based off of what I had recommended above would be great information to help narrow this down. Also, if you have a schematic, we may debug this issue further.
The EVB works perfectly. I am suspecting that the chip may be extremely sensitive to layout around the central ground pad as the biggest difference is that I have only a few vias in that area compared to “many” on the EVB.
I have eliminated most of the component related items as possible causes by transferring MY capacitors onto the EVB (and it still works).
VG and VCP were already 1uF on my design.
I will also hard tie the weakly-pulled inputs per your suggestion and update this thread.
Is the EVB made with 2 Oz or 1 Oz copper?
It will likely be a few weeks before I re-spin my board and am able to confirm layout, although it still would seem weird if it’s the middle ground pad that’s the problem.
Lots of times these circuits back off their current to keep the die temperature within limits. So if the thermal layout is poor the performance will be limited by the die temperature
You bought the evaluation board (EVB): Thank You for performing that service!
My project is MP6602 No Power to Motor while No Fault Condition - #4 by tmi11er passing NO current and it is far from anything like the recommendations for heat dissipation.
Proposition: The chip trips its current limit very very early in proportion to the die temperature.
Dialectical: Nowhere in the datasheet do they write the mathematical relationship between the die temperature and the current limit.
Conclusion: There is a large amount of work necessary in calculating a suitable mounting for the MP6022 to operate properly and not enough information to know it will operate without experimentation unless you have the necessary experience in the same circumstances.
I look forward to hearing any updates to your issue on top of the latest feedback I have provided in the issue that you have linked. Also looking forward to an update from you as well @stevec
First board built and at first it didn’t behave properly, in fact, matching the other posters issues - but I think it was a SPI pin that wasn’t quite connecting. Re-flowed the part and now all is well. I am able to get higher current.
The main layout change was to ensure LOTS of Vias on the ground pad (I had 5 originally, now I have 12) for the part and I ensured that each of the corner ground pins had exactly the same amount of copper so all the current flows through every ground was balanced.
Clearly the chip has some unexpected sensitivities, but this seems to have it working now.