DRV8874-Q1: DRV8874-Q1 thermal dissipate problem

Ziqi Yang

Part Number: DRV8874-Q1
Other Parts Discussed in Thread: DRV8874EVM

Hello

I`ve bought some pre-released DRV8874-Q1 for testing, according to the evaluation board of it, it can bear continuous current of 3.3A as suggested on datasheet,

but on my designed board in PWM mode, as I tested it in 12V/3A, it can only run for about 7 sec;

for 12V/3.3A it can only run for 3 sec then the fault LED started to blink, current will drop,

is this problem of heat dissipate problem of my board or some other issue, can you me some advice??

Thank you

over 5 years ago

0 Hector Hernandez Luque over 5 years ago

TI__Expert 8985 points

Hi Ziqi,

Are you utilizing the DRV8874EVM? If so, is it modified in any way?

Can you go through the faults the DRV handles to check if you go below the VM's UVLO, the VCP's UV, or above the OCP? Finally, is there anyway for you to measure the temperature of the DRV IC?

0 Ziqi Yang over 5 years ago in reply to Hector Hernandez Luque

Prodigy 90 points

Hello Hector

I`ve went through this datasheet, and I basically followed evaluation board to design my board, so the structure is a little bit different from EVA board, and I am pretty sure it is the thermal shutdown problem, but I don't have a propal tool to measure the temperature as I tested it, it can bear 2.5A for 70s, 3A for 7s, 3.3A for 3s, but for 2A, it is a lot longer as I only tested it for 5mins.

Thank you

0 Hector Hernandez Luque over 5 years ago in reply to Ziqi Yang

TI__Expert 8985 points

Hi Ziqi,

Can you indicate the differences between your board and the EVM in terms of IMODE pin?

0 Ziqi Yang over 5 years ago in reply to Hector Hernandez Luque

Prodigy 90 points

Hi Hector

Thanks for reply, IMODE pin is the same with EVM as it go through 0 OHM jumper to GND.

Thanks

0 Hector Hernandez Luque over 5 years ago in reply to Ziqi Yang

TI__Expert 8985 points

Hi Ziqi,

If you have debugged the UVLO, VCPUV, and OCP fault table conditions and your EVM does not meet them when the fault is generated, then it can be thermal.

Please, let me know if you have proved those voltages and currents right before the fault triggers to therefore try to debug thermal issues.

As a first hand check, can you confirm the thermal pad soldering of the IC follows general guidelines for that package? If the thermal pad has impedance to the ground, the heat dissipation can be greatly affected.

0 Ziqi Yang over 5 years ago in reply to Hector Hernandez Luque

Prodigy 90 points

Hello Hector

My design`s thermal pad has impedance to the ground, and can you give me advice on how to test UVLO, VCPUV, and OCP ??

0 Hector Hernandez Luque over 5 years ago in reply to Ziqi Yang

TI__Expert 8985 points

Hi Ziqi,

From the GND pour that I am able to see, there is worry about how much fragmentation there could be on your VM to OUT1/2 to GND pours for the heat dissipation and the current handling path. If you can, please image the flow of current across all of those nodes. There must be a good current path with low impedance from VM to GND and back from GND to VM.

As for the UVLO, VCPUV, and OCP, put a probe on nFAULT, and probe VM, VCP, and your OUT1/2 current. Set a negative slope trigger on nFAULT and lower the scale to around 10 us/div. Highest sample rate possible. So, when nFAULT trips, you can see if any those voltages or the current pass the fault thresholds.

0 Ziqi Yang over 5 years ago in reply to Hector Hernandez Luque

Prodigy 90 points

Hello Hector

After test I can exclude UVLO, VCPUV, and OCP fault, so how can I deal with the thermal problem now as the board has already been printed out, would heat sink make a great difference to it?

Thank you

Ziqi

0 Hector Hernandez Luque over 5 years ago in reply to Ziqi Yang

TI__Expert 8985 points

Hi Ziqi,

If the problem is thermal and you can prove it is thermal, then a heat sink will help increase heat dissipation from the IC.

0 Ziqi Yang over 5 years ago in reply to Hector Hernandez Luque

Prodigy 90 points

Thanks Hector

0 Hector Hernandez Luque over 5 years ago in reply to Ziqi Yang

TI__Expert 8985 points

Hi Ziqi,

I will close this thread for now. Please, reopen this thread or start a new one if further developments arise.

Motor drivers

Motor drivers forum

DRV8874-Q1: DRV8874-Q1 thermal dissipate problem