• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 29 subscribers
  • Views 437 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

DRV8402/8432 - AGND and output protection

IGW84
Prodigy 50 points
Other Parts Discussed in Thread: STRIKE

Hello all,

I am finalising a design based on the DRV8402 and DRV8432.  The 8432 would be the preferred part but availability concerns means I have to design for the 8402 as well. Factory fit options for supporting component such as OC_ADJ resistor is fine.

I have a couple of questions:

1) AGND-GND connection

The 8432 suggests a 1 ohm resistor to provide a little isolation between AGND and GND (pins 8 and 9).  The 8402 does not have this.  Would it present a problem if the 8402 build had the 1 ohm resistor?  Or alternatively we could easily enough fit a 0R for the 8402 variant.  Any problem with this?

2) Output protection

My application has long cable runs (up to 40 metres) that go to externally mounted motors.  These are isolated from direct weather but normal "outside" problems can be present - lightening mainly.  Wiring is floating but induced currents could be significant.  I am not expecting to protect against direct strike to the (pole mounted) systems.  I would like to provide reasonable level of protection.

Unidirectional 1500W transient suppressors in an SMC pkg from each output to GND would be a start, possibly also DIN rail mounted lightening surge arrestors (but possibly too slow to be of huge benefit).  Would the single diode drop breakdown in the reverse direction present any problem?  It will appear in parallel to the low side MOSFETs body diodes.  Which raises the issue of whether the tansient suppressors should go on the inboard or outboard side of the current-limiting inductors.  I am thinking inboard side.

I will probably be including fusing as well (resettable or PCB mount .. not sure yet) but this will mainly be to catch load wiring errors as an adjunct to the cycle-by-cycle limiting of the chip.

Motor supply is 48V regulated.  Most of the motors (this is multi-channel) will be driven in simple full-bridge mode.  A few will need higher current and will be driven in parallel full bridge mode.  Motor control is not needing to be very precise in this application. I will be being very generous in the bulk PVDD decoupling caps (more than EVM) and the local 100nF decouplers will be very close to the chip.  I was going to make provision for snubbers (3.3R+10nF) from each output to ground (on the external side of the current limiting inductors).

Would the transient suppressors cause any issue?  Inboard or outboard side?

Any suggestions on anything more I can do to increase reliability?

Thanks,

Ian

  • 0
    TI__Guru 55605 points

    Ian,

     

    1)  This is no problem.  We added the 1ohm resistors to the DRV8432 layout based on lessons learned on the DRV8402.  The 1ohm can stay in there for the DRV8402.

    2)  TVS directly on the outputs work well and we have seen success with these components in the past in limiting spike voltages on the outputs.  I have attached a series we have used in the past.....not nearly as large as the ones you are describing, but the same idea in principle.  

    As you also mentioned, the snubbers are effective, but I would recommend placing them directly on the outputs as well for maximum effectiveness in limiting the fast rising edges.  

    0763.TVS-SMAJseries.pdf

     

    One other suggestion would be to make sure the heatsink on the top of the package is connected to GND.  The substrate of the die is exposed to the top side PowerPAD and should have a connection to board GND.  If the heatsink is anodized, scrape off the portion in contact with the exposed metal pad.  

     

     

  • 0 in reply to Ryan Kehr
    Prodigy 50 points

    Thanks Ryan,

    You wrote:

    "As you also mentioned, the snubbers are effective, but I would recommend placing them directly on the outputs as well for maximum effectiveness in limiting the fast rising edges."


    When you say "outputs" were you suggesting of the system (motor side of series output inductors) or outputs of the '8402/'8432?  Just wanting to be clear on your comment.

    I am in two minds here - I also need to be considering radiated EMC issues as well in my case.  So putting a shunt snubber (few ohms in series with a a few 10's nF) on the output of the system has some advantage here.   However, the reciprocal will be the case for any conducted spikes being picked up on the external lines.  In this case a snubber will be more effective on the inboard side of the series inductors.

    Resistor has to be quite power unfortunately in my case, a few watts, which is a nuisance. due to high power rail and I don't want too low a value in order to not create a high Q filter (with the motor inductance (motor is about 6 ohms and 6.8 mH).

    Thanks,

    Ian

     

  • 0 in reply to IGW84
    TI__Guru 55605 points

    Ian,

     

    I was suggesting the "outputs" of our device, but the tradeoffs that you mention are certainly spot on in terms of the placement of the snubber.  Most of my customers are more concerned with conducted, but they also don't have the long wire runs that you have.  

    You could, of course, have the stuff options for both in order to have the options open later when it is certification time.