• State
  • Replies 6 replies
  • Subscribers 30 subscribers
  • Views 202 views
  • Users 0 members are here
Support feedback
Options
Options
Related

DRV8244-Q1: Rapid Field Decay with DRV8244 and TVS2700

Brad Matheus
Prodigy 10 points
Part Number: DRV8244-Q1
Other Parts Discussed in Thread: TVS2700,

Tool/software:

Hello,

I have been working on a board that controls a solenoid using the DRV8244 motor driver and a TVS2700 for surge protection. We had failures occur while "coasting" the solenoid (Hi-Z), which is done to increase the rate at which the field decays. Two failures have been observed on our boards: at 16.8V (a 4-cell LiPo battery), the DRV8244 seems to have blown at the Vm pin (first image). Another tester reported that the TVS2700 failed with a higher battery voltage of 25.2V (a 6-cell LiPo battery), but the same conditions otherwise (second image). The solenoid used is 14mH and 2.3Ω, and controller was using a basic on/off control (not pwm). Unsure why these chips could have failed, any thoughts?

 

  • 0
    TI__Mastermind 23990 points

    Hi Brad,

    1. Can you provide this boards schematic? I would like to see anything that is connected to the DRV8244-Q1.
    2. How long is the solenoid being driven for in this application?

    Initial thought is that the current through the body FET diode after thermal shutdown activated was too high and at this point the device became damaged, but more analysis will be needed to make sure this is the case.

    Regards,

    Joshua

  • 0 in reply to Joshua Horlander Cruz
    Prodigy 10 points

    sample_schematic.pdf

    Here is the schematic of connected components, solenoid was driven high for 20ms and low for 35ms. Current through the body diode being too high and causing overheating makes sense to me, I'm thinking actively driving the solenoid in reverse until the current decays sufficiently should work as a fix without requiring changes to the circuit. Does that sound reasonable?

    Thanks,

    Brad

  • 0 in reply to Brad Matheus
    TI__Mastermind 23990 points

    Hi Brad,

    With the time that you are driving the solenoid, the device should actually be able to handle the thermal rise without getting too close to 150C at 85C ambient. I had assumed you may have been driving the solenoid for closer to 1 second which would have pushed the device closer to the thermal shutdown temperature.

    I checked with my team and the bulk capacitance on VBAT in the schematic appears to be a bit small for the current that is being driven which could be leading to VM rising above the abs max rating and leading to damage. We generally recommend 1uF per watt of power as a rule of thumb so with the 25.2V battery and the solenoid we would recommend going above ~280uF bulk capacitance. If you have a non-damaged board, can you supply VM with ~8V so that the current is around 3A to avoid damaging the new device and capture OUT1, OUT2, VM, and the output current? We want to check for a rise on VM during the coast state and we want to observe the high side FETs body diode conduct on the OUTx pins.

    Regards,

    Joshua

  • 0 in reply to Joshua Horlander Cruz
    Prodigy 10 points

    Hello,

    Attached is a single pulse. Channel 1 is output current, captured through the IPROPI pin with a 680Ω resistor for a peak calculated current of ~3.5A. Channel 2 is OUT1, channel 3 is OUT2, and channel 4 is VM.

    Below is a close up of VM right as it enters the coast state. It appears to have reached steady state on the right.

    Let me know if any other views would be helpful.

    Thanks,

    Brad

  • 0 in reply to Brad Matheus
    TI__Mastermind 23990 points

    Hi Brad,

    I will review this information with my team and I will aim to provide a response tomorrow.

    Regards,

    Joshua

  • 0 in reply to Joshua Horlander Cruz
    TI__Mastermind 23990 points

    Hi Brad,

    From the captures, the outputs and current look ok but that VM oscillation while not an issue at this current. Can you place more bulk capacitance on the VM node to help stabilize VM?

    Brad Matheus said:
    I'm thinking actively driving the solenoid in reverse until the current decays sufficiently should work as a fix without requiring changes to the circuit. Does that sound reasonable?

    This should help keep the thermal generation down. Although if the issue is due to VM pumping the device may sill become damaged so I would recommend keeping an eye on the VM voltage at least during testing if you test this.

    If you end up doing another test can you capture VM and the load current? This will help narrow down what is causing the damage with the load. 

    Regards,

    Joshua