DRV8825 current regulation problem

Hi Mikolaj,

Can you provide additional info, such as the capacitance on VMA/VMB and the bulk capacitance for the VM line?

Does the current available from the power supply decrease with the voltage increase? I have seen some power supplies act this way.

Do you have in schematics, layout, or scope captures that you can provide?

Thanks.

Hi,

I've recorded a example movie (motor is enabled): http://youtu.be/x8S-rfuDhho

Driver additionally contains 3V3 and 5V regulators, some LEDs that current current probably is <50 mA .

I've also tried with impulse power supply of 15 V and have the same issue.

There is a part of schematic:

http://imageshack.us/a/img703/5162/drv1.jpg

We can see there is a capacitors: 10u, 1u, 100n, 10n, 1n for VMA and VMB. Additionally there is 470u LOW-ESR capacitor (on the schematic it looks like it's nearer VMA, but on the board there is similar distance). Before D1 diode there are also a lot of 47u, 10u, 1u and 100n capacitors and one 2200u.

and there is a part of layout:

http://imageshack.us/a/img254/4854/drv2.jpg

The 470u capacitor is more on the left side of the board. The GND is highlighted.

Best regards,

Mikolaj

Hi Mikolaj

The VM current behavior in your video should be normal that if the raise the VM voltage then the average VM current will be lower.

The reason is that the current you set according to Iphase=Vref/(5xRsense) just determine the current through motor phase, and usually it is regulated by the automatic current chopping. It means the current in motor phase is totally different with the current out of the power supply.

For example, the phase current is set to 1A and the motor will takes nearly the same input power for example 5W from power supply at both 10V VM and 20V VM. Ignore the loss on the device itself, then, in 10V power supply, the VM current will be 0.5A, while in 20V pover supply, the VM current will be 0.25A.

Another thing in your video shows that the real phase current may be less than your setting level when VM is below 10V. If the motor resistor is too big, this could happen because the motor cannot draw more current at this VM.

A recommendation to your schematic is that you can add two 0.1u caps to both sense resistors just as our EVM does. 

Hi Rick, please add your comments if I miss something.

Thanks,

Wilson

Motor Application Team

Wilson Zuo said:

For example, the phase current is set to 1A and the motor will takes nearly the same input power for example 5W from power supply at both 10V VM and 20V VM. Ignore the loss on the device itself, then, in 10V power supply, the VM current will be 0.5A, while in 20V pover supply, the VM current will be 0.25A.

You're right. Now it's clear, but why motor doesn't work with higher voltage? For example for 15 V it steps only for a couple of seconds and with no stepping motor is hold only to 22 V. For 13 V it seems work, but motor loses steps, so I can't be sure the system is reliable for 10 or 11 V.

Wilson Zuo said:

A recommendation to your schematic is that you can add two 0.1u caps to both sense resistors just as our EVM does. 

I will try.

Thank you for your reply.

Hello!

Wilson Zuo said:

Could you check the nfault signal?

nFault signal seems be high.

Wilson Zuo said:

And the big VM cap should placed near VM and GND.

And here is a problem. I soldered two additional standard 100u caps near VMA and VMB - no result. I've added 47u LOW-ESR in half way from used 470u LOW-ESR and it works great to 20 V! Above 22 V motor isn't hold. It can be caused by thin motor wires. I've soldered additional 10u LOW-ESR capacitor next this 47u and it didn't change anything.

I don't want to use voltages above 20 V so I think the problem is solved. Anyway, I am interested in reasons of problems for higher voltages.

In pololu site we can find the drv8825 module (http://www.pololu.com/catalog/product/2132) with no electrolytic capacitor on the board but with recommendation of use one 100u.

Thank you very much for your help.

Best regards,

Mikolaj