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.

DRV8305-Q1: Problem With Driver Output

Part Number: DRV8305-Q1

Hello, to whom this may concern,
I am trying to develop a custom board using the DRV8305N and I'm facing some problems. Although nFault pin is high, I get no response on the output, except for a dc of 5.6v on every output pin.
Some infos about my hardware:

PVDD = 10v

INHA = PWM around 20kHz, 50% duty-cycle

INLA = INHB = INLB = INHC = INLC = 0 vols.

This setup consumes around 23 mA.

There's no Mosfet soldered yet, only the minimum hardware to get the driver working.

Please, I would like to understand what that means, I've tryed to apply 8v and 6v and all the outputs keep as 5.6v fixed. The driver emits a noise that resambles commutation, but there's no mosfet soldered, as I've said before.  

  • Emanuel,

    I would suggest using the device EVM. FETs are needed because the protection features will be checking for Vds and Vgs when attempting to switch. 

    Regards,

    -Adam

  • Can  I add 2 mosfets only while I test the driver features? Or do I need to add all 6?

  • Emanuel,

    All of them are needed as the protection features will check all 6 outputs and all 3 phases.

    Regards,

    -Adam

  • Adam,

    Other question: I am using 22 ohm resistors between the gate and the GHx and GLx ports (in all of the mosfets), is that a big issue?

    Sincerely,

    Emanuel

  • Adam,

    The chip seems work fine.The interesting thing is that with only 1 half bridge welded it worked fine. The only problem now is the temperature, the chip gets stable, but quite hot, is that normal?


    Sincerely,
    Emanuel  

  • Hi Emanuel, thanks for the feedback.

    The chip should not be running hot. Is the temperature comment based on the EVM or on your custom board ?

    If this your custom board, then we would recommend checking the thermal pad soldering since the IC relies on this for heat dissipation. If this is the case, we have several PowerPad design recommendations. As goo starting point is the application note below

    . PowerPad Made Easy

    Regards

    Anuj Narain

  • Anuj Narain,

    That was about my custom board. I did not design the thermal pad it was supposed to have, cause my board do not have much space. 
    Although it is still running well, with stable current consume. 

    Regards,
    Emanuel

  • Hi Emanuel, 

    Great. That sounds like it is the problem. As you start to design your board we have a PCB layout guide that other customers have found useful. You know where to find us if you need help. 

    [FAQ] Best Practices for PCB layout of Motor Drives

    Also, 

    If your issue has been resolved, please help the E2E community by clicking the green “This Resolved My Issue” button.

    If you have a related question, please use the orange "Ask a related question" in the top right corner of the page.

     

  • I have a feedback about the driver: The chip (DRV8305N) seems work fine, it finally responds to the SPI communication, tested with no motor terminal plugged in. 

    With no motor plugged in, although the outputs on the 3 terminals, each between 2 mosfets in the same half-bridge, have right behave according to the datasheet, when I plug the motor, it makes the driver burn. It stopped to respond to the SPI data, and stopped working. The only thing I did was to plug the motor. Before this, it was just working fine.

    Notice that the DC bus connected to the Drain of the High side of the 3 half-bridges is connected also to the PVDD.

    To someone that could help me.  

  • Emmanuel, 

    Do you know what was damaged? Have you taken any measurements of the damaged IC?

    Did you improve your power pad?

    Regards,

    -Adam

  • Do you know what was damaged?

    - I don't know exactly, but it's heating much more than previously, and I can't access the SPI registers anymore.

    Have you taken any measurements of the damaged IC?

    - What you meant? During the continuity test it was ok.

    Did you improve your power pad?

    - the power pad continues the same. But I've added a cooler, so the temperature is not a problem.

  • Emanuel,

    The device needs to be replaced. When you remove the device from the PCB, I would measure it's pins and compare them to a known good unit.

    Even if the cooling is not an issue, the power pad and it's grounding are still needed.

    Regards,

    -Adam

  • The problem is that it was working just before plugging the motor. What kind of event could have happened to cause its failure? My question is that it was not supposed to happen, considering that all power is driven to the load by the mosfets, as the driver control it's gates.

  • Emanuel,

    Without the motor installed, the Bridge won't be carrying any current. Once the current is introduced the FETs and board GND behave very differently.

    Regards,

    -Adam

  • Let me get things more clear: the when I try to access the chip using SPI, it responds with a number that corresponds to all bits set to 1, I meant, 0xFFFF.

    Comparing 2 boards, one working properly the other dead, both seems the same, when checked the resistance on each pin, when both not powered.

    Comparing both powered, the dead one demanda around 200mA as the other don't go above 70mA.

  • What kind of test should I execute so that you guys could help me to solve this problem?

  • Emanuel,

    Please remove the damaged IC from the PCB and check the resistance of each pin to GND, then repeat the same for a known good IC which is not attached to a PCB, likely you will find some differences.

    Regards,

    -Adam

  • Emanuel,

    I have seen no update on this thread, did your issue get resolved?

    I will mark this thread as closed for now but please post again if you need further assistance.

    Regards,

    -Adam

  • I didn't find differences between the two chips, i.e., a good one and a burned one, but the response to SPI communication is always 0xFFFF, for all of the registers.

    This happened because I've tryed to use a motor with a high demand for current during the start. Does it make sense? To avoid this should I use another motor, with lower power or do the chip can handle with big inductive loads? Although, the Mosfets are great, not even got warm.

  • Emanuel,

    The chip that is taking more supply current and returning FFFF on the SPI is obviously damaged. I tend to feel that the DRV is getting damaged due to the PCB and the issues that need to be resolved. 

    Regards,

    -Adam

  • The power pad underneath the chip, as you said, is important to be solded. But it's just a good practice in terms of heat dissipation or it's needed, indeed?

  • Emanuel,

    It is needed both electrically and thermally.

    Regards,

    -Adam

  • Emanuel,

    I have seen no update on this thread, did your issue get resolved?

     

    I will mark this thread as closed for now but please post again if you need further assistance.

     

    Regards,

    -Adam

  • Hello, again.

    I have solded the thermal pad to GND and it worked for a while but suddenly the chip died when the duty-cycle was increased from 10% to 50%.

    The test: Using the 1 pwm mode, with a motor plugged in the system, I've connected the hall sensor output to the INLA,INHB,INLB. In the INHA I've put a 10khz pwm (initially, using 10% it was working, and the motor was spinning very slow. Then, using 50% duty, it burned the DRV8305N); the set up for the registers was: for Gate Drive Control Register (Address = 0x7), the 11 bits were: 01101111111 (MSB ... LSB).

    Not used: Capacitor between DC source, what could have caused momentarily a drop on the system's voltage. 

    Question 1: I was writing the same value on the same register (address 0x7) each 200ms, for debugging porpuses, while the motor spins . Could it have caused for any reason the burn of the chip? It is recommended to disable the "enable gate driver pin" during register writing?

    Question 2: A motor that requires a certain amount of current could burn the chip? (my motor is not so big, but it demands a lot of current, which, I suppose, have nothing to do with the chip, but with the MOSFETS current sourcing capability).

    I need help to understand what is the problem with the chip behave.

    Sincerely,

    Emanuel

  • Emanuel,

    I suggest you try spinning the motor with one of our EVMs and see if that works well, please check here:

    https://www.ti.com/product/DRV8305-Q1/toolssoftware

    Question 1: It is not required but it is recommend to configure the device via SPI while the EN_GATE pin is low. You could try a new chip with this method to see if this is causing the issue.

    Question 2: Higher current will not damage the DRV directly as this current is not through the DRV itself but instead through the FETs as you mentioned. One possibility is that higher bridge current causes higher transient voltages which the DRV will see on GND, SHx, GLx, and GHx; these could damage the DRV.

    Can you check the GHx, GLx, SHX nodes while the motor is spinning? Look for any voltage spikes or anything that looks out of place.

    On a damaged device that has been removed from the board, could you check the resistance of the output (GL/GH/SL/SH) pins to GND? Then compare this to a known good unit that's not soldered to a board.

    Regards,

    -Adam

  • Adam,

    - All the pins GL/GH/SL/SH related to GND have the same resistance, comparing a good unit to a  damaged one. If the DRV was damaged through this pins, it would indicate different resistance, which does not occurs, is this analysis correct? 

    Q1: To test the transient effect I need to know: If the Hall sensors output pins are connected to INLA, INHB and INLB, with INHC = 0, setting registers to 1-PWM MODE, it is supposed to see the motor spinning according to INHA PWM duty-cycle, right?

    Q2: The current sense amp: the lower the gain applied, the more likely to burn the chip throught this pins (analysing amp circuit)?

    Sincerely,
    Emanuel 

  • Emanuel,

    This is not a 100% guarantee that these pins were not damaged but it shows that there is not a gross failure with these pins.

    Q1: Yes you are correct, please see datasheet pages 19, 20, and 21, also figure 8:

    Q2: Can you do the same resistance check on a good and bad unit on the CSA pins? (SN1/2/3, SP1/2/3, SO1/2/3, VREG)

    Regards.

    -Adam