PCM6260-Q1: For MIC+ line shorted to battery fault; any potential risk on V_MICBIAS?

Hi Daveon, 

Thanks for your reply! 

But I don't think shutting down MICBIAS will resolve the hardware risk, since shutting down the output doesn't mean cutting it off from the rail and MICBIAS will be still shorted to VBAT.

And worse still, I'm concerned that it would damage the BSTVDD port via the built-in boost converter.

Any thoughts on that?

Thanks,

Mingrui

Hi Mingrui,

You're correct, the MICBIAS pin does not have the ability to sync current so in the event it is shorted to a higher voltage (i.e VBAT) the pin will saturate and pump current back into the HVDD/Boost. This can damage the device, please allow me some time to consult with design team on best practices to protect device from any reliability concerns. I will update thread this week.

Regards,

Hi Mingrui,

Just to provide an update, there is validation data over the response to MICBIAS short to VBAT condition, but the engineer who conducted is out of office this week.

If urgent, a zener diode at the output of MICBIAS may be the only additional circuitry needed, but please let me confirm with engineer next week.

Best Regards,

Hi Daveon,

Any updates now? Thanks!

Hi Mingrui,

There isn't any reliability concern in the device with reverse current, the current flow should be reduced to uA's by an internal clamping mechanism.

We have extra leakage path inside the device on MICBIAS which gets Auto triggered in case of input short to VBAT to ensure reliability if needed.

In addition, shown on page 97 of the d/s, a micbias high current threshold feature is available. Customer can set the load current thresholds in steps of 0.5mA and accordingly limit it from within the device and know in real time if the current crosses the threshold.

My zener diode comment is still applicable if wanting additional protection.

Regards,

Hi Daveon,

Could you please elaborate a bit more on the internal clamping and the extra leakage path mechanism? The customer might be interested in this.

Thanks!

Hi Mingrui,

I've presented the question to our systems team, I will update the thread following their feedback.

Regards,

The maximum voltage on the HVDD Line is 12v . If VBAT goes to 16 v then there can be a leakage path to this 12v clamp through the Bias Resistor as shown.

We can limit this current to less then 10ma by choosing a suitable Bias resistor. A value of 470 ohms should be  fine. 

Alternatively  ,a schottky diode in series with the Bias Resistor can be used as shown.

Hi Sanjay,

In the PCM62x0 devices there is no HVDD input. For these devices with the integrated boost, is this 12V clamp still applicable? 

Thanks,

Alex

I've changed ownership of this question to Sanjay, please wait for him to update the thread.

Regards,

Yes this is still applicable.

The Maximum Micbias that is programmable is 9v and the Boost converter keeps the BSTOUT higher by about 1V at 10v. During norml operation this pin never reaches 12v

Hi Sanjay,

Thank you for confirming. As we are working through worst case analysis, does this 12V clamping diode have any tolerance we should consider? Is the 10mA the max current that should be allowed to pass through?

Thanks,

Alex

You can keep the 12v as the Clamp value. Its best to keep the worst case current less then 10ma