DRV8876: Motor selection

Hi,

The device has a couple of fault conditions, we can start from there and find out the issue:

Now, we can start with the easy ones: can you check your VM voltage and the charge pump voltage when the nfault assets low and make sure they are above 4.35 V and VM + 2.25 V, respectively?

If those two voltages are above those fault thresholds, then we can check OCP, OCP trip point is 5.5 A with a deglitch time of 3 us. If you have a current probe to probe the output current when nfault asserts low, check if the current reaches 5.5 A longer than 3 us.

Aternatively, you can probe the IPROPI resistor's voltage to make sure it is not greater than VREF (3.3 V). Note that VIPROPI (V) = IPROPI (A) x RIPROPI (Ω), IPROPI (μA) = (ILS1 + ILS2) (A) x AIPROPI (μA/A). Also note that AIPROPI (μA/A) = 1000 and RIPROPI (Ω) = 1.5k. The moment VIPROPI > VREF, current chopping occurs and you have nFAULT asserting low in cycle by cycle approach. 

Hi Hector

Thank you for your reply.

Regarding to VM and Vcp, there were no problems.
(VM=24V, Vcp = 29V)

I attaches waveform.e2e_ck.xlsx

Regarding to OCP, I faced strange behavior.
When turned EN to clockwise and kept position, motor current increased gradually.

I wonder why the motor current increases regardless I have not proceeded to turn EN to clockwise.

Could it be possible to use DRV8876 for any DC motor?

Is there any DC motor that DRV8876 can not work with?

BestRegards

Hi Na Na78,

I will look into the scope captures.

Can you share the part number or at least the wattage and resistance of the motor you are utilizing?

Hi Na Na78,

There is a possibility you are seeing current chopping reflected on nFAULT (Itrip indicator, the first row in the fault table). Your current chopping threshold is to 2.2 A and I see VIPROPI reach 3.3V (VREF = 3.3V). The next question is to verify this observation.

If you can, probe nFAULT instead of VM and OUT1 instead of Motor Current. Also, you can do a second probe with OUT2 instead of VM and OUT1 instead of motor current. The first scope capture will assist in proving whether or not the nFAULT is tied it to the VIPROPI value. The second scope capture will help to know if low side slow decay (active H-bridge) is occurring.

One final way to check, which would be more time consuming and the scope captures can prove this verification so you don't have to do it, is to turn off current chopping nFAULT notification by setting IMODE back to GND instead of 20k.

Hi Na Na78,

Is the customer still facing the fault condition? Has it been debugged to verify from where the fault comes from as per suggestions above?

Both modes require different logic levels on IN1/EN and IN2/PH if you want to go Forward, Reverse, or Brake the motor. Can the customer very their logical setup is correct on those pins?

Hi Hector

I'm sorry that the file I attached was not updated.

I attach new file again.  Please see the tab of 191010_up and 191011_up.

e2e_ck2_v2.xlsx

PH/EN mode does not work while PWM mode can with the customer's motor.

nFAULT is asserted in PH/EN mode.  I don't know why nFAULT is asserted only PH/EN mode.

BestRegards

Hi Na Na78,

It appears that nFAULT is low for about 30us (current regulation off time), IMODE has 20k to GND (current chopping asserts nFAULT low) and the VREF is set to 0.5V (Itrip = 0.33-A). What the customer might be seeing is the nFAULT going low the moment the commutation starts on the output as the current reaches 0.33-A). 

Something that confuses me is on the 191007 data the current reaches over 2 A but on these two new date-based scope captures VREF is not 3.3-V (it is 0.5-V), so the current regulation is at 0.33-A. Can you confirm with the customer if they want current regulation at 2.2-A or 0.33-A?

I also see some of the captures already start with nFAULT low. If possible, try to get the captures when the nFAULT asserts low for the first time.

Hi Na Na78,

Any update here? Thank you.

Hi

Sorry for late update.

>Something that confuses me is on the 191007 data the current reaches over
>2 A but on these two new date-based scope captures VREF is not 3.3-V (it is 0.5-V),
>so the current regulation is at 0.33-A. Can you confirm with the customer if they want current regulation at 2.2-A or 0.33-A?

Sorry for confusing you.

As I mentioned, the motor worked in PWM mode.

And in PWM mode, I found the motor could start in even 0.33A current limit settings.
Actually, the customer informed motor specification different from the spec mentioned in first question. 

Old information -> New information

VM=24V -> VM=24V
Iout = 0.34A -> 0.03A
Istartup = 0.9A ->0.3A

>I also see some of the captures already start with nFAULT low. If possible,
>try to get the captures when the nFAULT asserts low for the first time.

I will try to capture it and update later. (But I have been in business trip so it would be late)

BestRegards

Hi Na Na78,

Thanks. Please, keep me posted.

Hi Hector

>I also see some of the captures already start with nFAULT low.
>If possible, try to get the captures when the nFAULT asserts low for the first time.

I triggered nFAULT.
I attached waveform when nFAULT was low for the first time.
For investigation, I captured forward mode and reverse mode in EN/PH operation.

e2e_ck2_v3.xlsx

Can you see any issues causing fail to start the motor?

BestRegards

BestRegards

Hi Na Na78,

Can you tell the customer to try and raise their VREF value (increase the current regulation threshold)?

My theory is that the current is limited to a point where the motor does not have enough torque(current) to overcome the initial torque required to spin the motor.

Hi Na Na,

Is there any update from customer? Thank you.

Hi Hector

Sorry for late reply.

Let me continues discussion here.
I added waveform with Vref=3.3V.  And it shows the timing when nFault was low for the first time.

e2e_ck2_v4.xlsx

Not similar to PWM mode, nFault was asserted while in motor staring operation.

Have you ever been faced the situation like a motor could only start in PWM mode.

BestRegards

Hi Na Na,

This is a unique issue. Can you detail if the fault appears when motor is set to spin in same direction on both Operation Modes (e.g. if motor is spun in forward or rever direction in both PH/EN and PWM Modes)?

Hi Hector

As I attached the behavior of PWM mode in 191010_up and 191011_up, nFault is never asserted in PWM mode. 

In PWM, the motor can start spin without errors.  This good operation of motor in PWM mode does not depend on Vref.

What kind of waveform should I take?

A strange behavior is observed in EN/PH mode.

I explain as follows. (I want to give you movie but I can't do due to size limitation.)

CH1 is VPROPI.  CH2 is nFault.

At 0:00, VM=24V is already applied. Of cause CH2 is high and not changed.

At 0:02, when I turn EN to fixed position in clockwise, CH1 and CH2 start to change its states.
At this time, CH1 is increased gradually.

As I described, I just turn EN to fixed position once, so I never turn EN after 0:02.

VPROPI will be relative to motor current. 

I think this situation means motor current is increasing gradually even if input is not changed.

Can I make you understand?  If possible, please give me your view?

BestRegards

Hi Na Na,

I understand. My question is, is the motor spun clockwise in both PWM and PH/EN Modes? Has counter-clockwise direction been tested as well?

Hi Hector

Thank you for explanation.

In PWM mode, the motor can be spun either clockwise or counterclockwise regardless of Vref.

In PH/EN mode, the motor can be spun neither clockwise nor counterclockwise regardless of Vref.

BestRegards

Hi Na Na,

Yes, this is interesting. Questions:

Has another motor been tested?

Has the customer debugged in PH/EN the fault table to determine which fault is triggered while in PH/EN?

Please, share design files privately and I will review in the upcoming days.

Hi Hector

Thank you for your reply.

I update file with additional waveform (191105_up, It is EN/PH mode).

e2e_ck2_v5.xlsx

>Has another motor been tested?

Yes. With a different motor,  DRV8876 could work well both in PWM mode and EN/PH mode.

>Has the customer debugged in PH/EN the fault table to determine which fault is triggered while in PH/EN?

This problem occurs with DRV8876EVM and customer's motor.  So there is no customer's design file.

I could not obtain current waveform due to lack of current probe now, but I think the fault would be caused of OCP, not current regulation.  According to 191105_up sheet above, VPROPI is clamped over 3.3V and nFAULT is triggered while VPROPI is clamping.

BestRegards

Hi Na Na,

It appears that, when utilizing the motor that is generating the fault, it is requiring more current in PH/EN Mode to start the motor than when the motor is driven in PWM Mode.

Can identical screen capture to what was taken in v5 be done but with IN1 instead of EN? I want to see if VIPROPI reaches the maximum voltage to get an idea of the current behavior between both Operation Modes.

Hi Na Na,

Does the customer need to operate in PH/EN Mode or can PWM Mode be acceptable?