This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

DRV8704: Operating Temperature Concerns

Part Number: DRV8704

Hello,

I have some questions regarding the typical operating temperature of the DRV8704. The device is being used to drive a Peltier. I am within the operating conditions and have the gate drive settings set correctly.

My setup uses +24V as its “VM”. It has come to my attention that the device gets warm (around 45C) while “awake”, but with no PWM inputs applied. I suspect this is due to the internal LDOs that must drop from the input voltage to 1.8V, 5.0V, and ~10V. I have the device mounted on a 6-layer PCB with 3-oz internal layers. The exposed copper pad on the underside is stitched into the ground plane. Originally, I was operating at +48V input but the device was unable to drive any significant load as the device temperature would rise to its absolute maximum rating. Using a fan is a last resort.

One thing that concerns me is that the absolute maximum operating virtual junction temperature (Tj) is 150C, however the thermal shutdown temperature (Ttsd) is characterised as being 150C minimum, 180C maximum. Doesn’t this mean that the device could be damaged even before it gets chance to show a fault in the status register and pull its fault(n) line low?

I see that the recommended operating temperature of the device (Ta) is 85C maximum. My driver runs at around this temperature without a lid on the unit, however once a lid is added the temperature of the driver continues to rise. I measured around 110C before I shut the device down to prevent any possible damage.

I plan to revise the PCB where I will try and improve the thermal distribution by possibly mounting the device on the underside of the board, and then using an SMD heatsink on the topside, placing lots of thermal vias to transfer the heat from the exposed copper pad up to the heatsink through the  ground plane. I suspect that these internal LDOs are acting as very localised hot-spots, and therefore it may be difficult to effectively spread the heat. If I can better distribute the heat, would you think that it is safe to run the device for prolonged periods at say 110C?

Do you have any other advice or recommendations as to how best to improve the thermal performance of these devices? A driver that could be powered separately to the bridge it is driving would be ideal, but I can’t see any in your range that have this arrangement.

Kind regards,

Stephen

  • Hi Stephen,

    The internal regulators can be contributing to the higher temperature, but there may be something else.

    Would it be possible to provide the layout? We can make arrangements to receive it privately.

    If the DRV8704 cannot solve the issue, the DRV835x may be an alternative assuming you are only using two half bridges of the DRV8704.

  • Hi Rick,


    Thank you for your reply. The DRV8704 driver is being used in full-bridge mode to drive the TECs, so one driver controls two TECs (driving eight MOSFETs).
    I am almost ready to revise this design and I think I could improve the location of the driver somewhat and also adding extra thermal vias. A passive heatsink is an option.

    I can send a partial design across via PM if you're happy to do so?


    Kind regards,


    Stephen

  • Hi Stephen,

    Instructions have been sent regarding how to share the design.

    We will mark this thread as closed.