I/O ports connected to nSLEEP and nRESET broken

Herve NONON

Other Parts Discussed in Thread: DRV8814, DRV8846

Hi Everybody,

For a second time, I/O ports connected to nSLEEP and nRESET (DRV8814) and nSLEEP, nENBL (DRV8846) broke.

First time, I made the error connecting nSLEEP (which disables regulator) to V3P3OUT (on DRV8814). This time, all these signals are connected to the microcontroller (Atmel ATxmega128A3U). With the first electronic card, DRV8846 works fine, and after correcting the error and change of DRV8814, all seems good. With this second card, all I/O ports connected to nSLEEP, nRESET (DRV8814) and nSLEEP, nENBL (DRV8846) broke. DRV8814 seems to operate (ports are permanent high states). I can't test DRV8846 because nENBL is permanent high too.

I think, this could come from order of power supply. DRV8814 is power from 24V supply, DRV8846 from 12V supply and microcontroller from 5V via 3V3 regulator.

Does anybody know what the problem is and how to correct it ?

Thanks for your responses.

Hervé

over 9 years ago

0 Wilson Zuo over 9 years ago

TI__Mastermind 26065 points

Hi Herve,

So you mean both the I/O of the MCU and the nSLEEP, nENBL pins are all damaged?

Normally there should be no problem even we don't care about the power on sequence too much.

Could you post your schematic? Did the issue happen to all boards, or just a small failure rate.

Best regards,

0 Herve NONON over 9 years ago in reply to Wilson Zuo

Prodigy 150 points

Hi Wilson,

and thanks for reply

I can't test DRB8846 because nENBL is high and i hope to repair the card without cutting signal lines.
This card is a prototype : second version. But It is sailed and will be really used on 12-15-2015.

below are pieces of schematics for DRV8814 and DRV8846.

3V3_24 is derived from 24V supply

GNDIO is microcontroller gnd

+3V3 is microcontroller 3V3

xxx_FAULT signals are connected to microcontroller inputs with pullup

xxx_PWM and MOT_IMAX signals are connected to microcontroller PWM outputs

other signals are microcontroller outputs

Do you see something wrong ?

Regards

Hervé

0 Herve NONON over 9 years ago in reply to Herve NONON

Prodigy 150 points

I noticed 220R resistors on DRV8814EVM schematics. Perhaps it could be the solution ?

0 Wilson Zuo over 9 years ago in reply to Herve NONON

TI__Mastermind 26065 points

Yes. It will be help to use serial resistor which will reduce any surge current on the pins even ESD event.
But the root cause of your case maybe other reason, so better to confirm it with more reproducing test.

Best regard,

0 Herve NONON over 9 years ago in reply to Wilson Zuo

Prodigy 150 points

Hi Wilson,
Do you know what happen if nRESET and nSLEEP (DRV8814) are high impedance ? ( or nSLLEP and nEN for DRV8846)
It is the case before the first MCU programming (or during MCU reset). Can it cause "the breaking state" ?

Best Regards
Hervé

0 Rick Duncan over 9 years ago in reply to Herve NONON

TI__Guru** 114905 points

Hi Herve,

The inputs have internal pulldowns. In general, there should be no problem.

A MCU reset could cause damage the device if the current in the windings cause the voltage to spike at VM.

0 Herve NONON over 9 years ago in reply to Rick Duncan

Prodigy 150 points

Sorry to have bothered you !

It was a mistake : output ports were not broken, just masked by an unintended JTAG interface !

Thank you very much for help !

Hervé

0 Rick Duncan over 9 years ago in reply to Herve NONON

TI__Guru** 114905 points

Hi Herve,

Thanks for following up with the reason for this. I hope it will help someone else.

Motor drivers

Motor drivers forum

I/O ports connected to nSLEEP and nRESET broken