Hi James,

acording to datasheet it seems you are are right - I will add this capacitor and report back. I hope I can keep thread open for this for a while.

regards

I just soldered 0.015uF as it is described in documentation for DRV8302 (8.2 Typical Application page 21) but that didn't helped - the same situation - I get same voltages as described above.

"When you fix the buck" ... yes - the question is how to fix it... Do you know how I can troubleshoot?

When I turn on 12V power supply BST_BK goes slowly (in 5second timeframe) to 5V, but PH stays all time at 0V.

Also Vsense stays 0V all the time, but I see corect ss_tr and en lines as per TPS54160 datasheet:

I got this values copy paste from another project ( page 7) - now I recalculated as you suggest - and I get:

R2=10kOhm

Hi James,

Yes I can remove D4, and I did remove it.

I changed capacitor between PH and BST to 100nF as per datasheet, restart plugging in 12V and no change - still nothing.

Hi James,

yes I use osciloscope all the time, the problem is knowing what is signals shout output what voltages this is why I documented voltage level in one of my first posts.

But now I am posting all signals at the time of starting up drv8302 chip. Maybe you coud diagnose better (attached to word file)osciloscope_traces.docx

Yes it could be the part is damaged - but I already resoldered another chip. Now I ordered 3 of them and I think this is cheaper than reference design that costs on farnell aprox 340€.

Regarding bridges yes I use flux and i am making sure there sre no bridges (I resoldered part one time), annd you can check scope traces - if there were briidges - two neighbouring pin would have same trace - but from picture you can see they dont have it.

regarding pads - yes I know this should be fixed in another pcb iteration - some of the caps are not available in 0805 and I soldered 1206 part to 0805 pad - but there is no shorts I double checked that (also traces from scope I think confirm this - no?)

Hoping some expert from your team with good knowledge of this part could find out why the part is not working (maybe scope traces in word doc I posted could help)?

Is there some watchdog pin to know if this part is alive or not?

I ordered 3 new chips from Farnell, but since they're out of stock, I am still waiting.

I am almost 100% sure ground pad was soldered, but I plan to put big via beneath gnd pad so I could solder it from bottom side.

Now I had small vias beneath and it was not possible to solder it with iron.

I guess bigger via could make me 100% gnd pad is soldered.

Why on Earth Farnell doesn't keep stock of this chips?

Hi Zhivko,

I'm not sure what is the situation with Farnell, but you can order parts from us directly using the TI Store.

For soldering the PowerPad on the chip, you may want to look into using reflow solder techniques. Soldering from the bottom through a large via may not be reliable. Here are a couple application notes about soldering our devices:

Mounting of Surface Mount Components

Wave Solder Exposure of SMT Packages

I just got new DRV8302 (3 pieces from farnell), resoldered it, just to found out I left BIAS pin floating - yes no 1MOhm resistor there.

Could that be reason why part is not working?

Obviously this and shorted pin as pointed out on bottom picture solved the issue.

Buck is hapilly producing 3.3V now. Thanks for generous support!

I wonder - if I do not want buck operation could I let BIAS pin floating?

More precisely - I have another DRV8302 on pcb, on which I do not want buck to be operational - for this reason I left some of pins floating - is this OK or not? I marked those pins with red rectangle - now I am getting nOCTW pin turned on on this DRV8302 with pins connected as below:

It was low - because I mislooked I need to pull it high - now I did that in SW (in stm32f3) and now it is not low - so it seems OK. This is same for fault pins - I needed to pull them high so they became NOT FAULT.

" If the IC is getting hot" ... I still havent tried to turn on motor so they couldnt be hot for that reason for sure :) since I wrongly read fault and OCn pins.

I am setting OC_ADJ pin with DAC on stm32f3 chip - I need to software control max current - I hope that part would work.

I will try to turn some brush two phase motors tomorrow - I am looking forward to do that! :) So far so good :)

I am attaching also top and bottom PCB. Mosfets I use are IRFS7530TRL7PP.

Schematics: 2350.shema12 - Schematic.pdf

shema11 - PCB-bottom.pdf

shema11 - PCB-top.pdf

I made motors turn on 12V power supply, than I tried on 24V, and driver became hotter - and went into FAULT state. the current consumption (shown on power supply) got to 187mA.

Now it reports FAULT also on 12V and I couldn't make motors turn anymore - what could be wrong ?

When I turn on power supply current consumption is 14mA, after I change PWM on INH_A and INH_B, it starts consuming 193mA and gets into fault state and stays there (could not be resetted with togling EN_GATE)... I believe 24V is well bellow operating range of this driver chip.
Drivers becomes hot so I could not touch it for more than 3sec only.

Hi Zhivko,

If the DRV8302 is getting hot and not responding to the EN_GATE signal, it could be damaged or the FETs could be damaged. Review section 7.3.4 in the datasheet and probe your board to determine why the device reports a fault.

Are you using 3-PWM or 6-PWM mode? If you are using 3-PWM mode, I would still ground the unused inputs (INLx) rather than leave them floating.

I am using 3PWM mode and I have question how should I put pwm signal on PWM1 and PWM2 (marked on bottom picture) to make 2 phase brush motor turn left and how to put pwm1 and pwm2 to turn it right:

OK I will tie INLx to GND but on second PCB version - is it ok to assume if it worked on 12V without INLx tied to gnd it should work also on 24V ?
Can you check my pcb layout (pdfs I attached) to see if GNDs - PGND, AGND and GND are connected corectly? I see in datasheet that it mentions AGND, PGND and GND, but this 3 grounds are not clearly visible on layout picture in datasheet - can you clearify or anotate that picture to make that clear?

Is it possible that damage happened to gate driver because of missing capacitor on power line?

Another thing - with new chip soldered  - I see this trace becomes VERY HOT:

Designing board I followed datasheet suggestion to connect grounds (power ground and drv8102 gnd) starwise - should I just connect those grounds as much as possible?

And fyi - I disconnected SS_TR pin - and buck works without problem.

Here is the truth table for PWM driving with the A and B phases in 3-PWM mode:

Regarding grounding and PCB layout, you want to make traces which carry large currents as wide as possible. This includes VM, GND, and motor phase traces. If you download the hardware design files from the following links, they provide good layout examples.

Motor Drive BoosterPack featuring DRV8301 and NexFET™ MOSFETs

Three-Phase Brushless/PMSM (Permanent Magnet Synchronous Motor) High Current Motor Control Solution

Three Phase BLDC & PMSM Motor Kit with DRV8302 and Piccolo MCU

Now I am geting nFault because of GVDD undervoltage when I make PWM signal on INHA and INHB. See scope trace below. What could be the reason or why could that happen? Do you think mosfets gates draw too much current?

I am using: IRFS7530TRL7PP

Gate Drive RMS Current = MOSFET Qg × Number of Switching MOSFETs × Switching Frequency

Gate Drive RMS Current = 236 * 2 * 23 = 10.856mA

This should be within drv8302 specification - yes?

I was able to move motor, but mosfet are getting really hot. I can see this scope pictures:

on PWM2 (PWM1 is 0V):

on GL_B:

Trying to get to the source of problem I see INH_B set to PWM freq of 24 kHz, and INH_A set to 0kHz, I get GL_B and GH_B turning on simultaneously - how is that possible?

The only thing that gets on my mind is current limiting mode - is this current limiting I am seeing? From datasheet I read "If M_OC = LOW, the gate driver will operate in a cycle-by-cycle current limiting mode". In my case I have M_OC set to LOW.

What is purpose of BIAS pin? For upper scope traces - I have left it floating - but will connect it to 1MOHM today to see if it makes any difference. Also on some schematics in datasheet it is capacitor connected on other schematics it is resistor connected to GND - what is the reason? What is correct way to connect BIAS?
Is it possible that BIAS influences or is anyhow connected to current limiting?

Yes I solved GVDD undervoltage events with soldering large cap to power rail (as I wrote in my previous post).

This are pictures taken after I soldered BIAS pin to 1MOhm resistor - however I cannot explain that whole thing is eating 4A - I have set 4A current limit and it is taking it all just for switching mosfets with no motor connected... I need to check this more in depth - to answer where is all that energy dissipating...

I think this undervoltage events are happening for some yet unknown reason for my pcb.

Regarding OC_ADJ - I have DAC connected to DAC from my uC and disabled overcurrent limiting with DAC outputing 3.3V to OC_ADJ (just for now when I am still struggling with proper motor operations using DRV8302).

HEreare pictures of motor phase1 (blue is GHA yellow GLA)

motor phase2 (blue is GHB yellow GLB)

Now this is unexplainable - why it is eating 4A on 12V - see below picture

Ok I think there is some flaw in my thinking - I am not sure why GL_A and GH_A are inverted while PWM1 is present and PWM2 is 0V.

I think that correct situation should be like (one of gate signal either GL_A or GL_B should be 0V) like this (0V denoted with horizontal green line):

But actually the situation is like this:

So as this last situation is happening - i have both mosfets short circuited - I am sure I am missing something... Why one of either gates of GH_A or GL_A is not 0 V but GH_A is inverted GL_A ?

Here are traces:

PWM1

PWM2

Gate signals - blue GH_A, yellow GL_A:

I wanted to try operating DRV8302 without MOSFETs beeing connected - right now I see if I remove gate resistors - chip stops outputing GL_A and GL_B - so is it possible to have GL_A and GL_B operating without mosfets beeing connected?

Another thing - is it possible that dead time between GH_A and GH_B is so small that it happens that ground mosfet is on phase A is turned on and 12V mosfet on phase A is turned on? That is the only explanation of short circuit situation that I am seeing now - would increasing DTC help (increasing dead time)?