XC9236 on nRF9160 UVLO when supplying power to external devices

[Jared]

I have the occasional issue where connecting VDD to VTG either on P20 or shorting SB24 will cause the nRF9160 DK to brown out. This happens when an external device to be programmed is connected. The issue appears to be related to the XC9236 itself. I took a quick capture of what's happening. See below:

My guess is there is something internal to the XC9236 which is turning off the output. First guess is the UVLO but since the drop is about 700mV. It never recovers though.. My best guess is the culprit is D10.

I'm using PCA10090 v1.1.0.

[Jared]

Removed D10 and shorted. Still happens.

[Kaja]

Hi Jared,


sorry for the delay, I will have a look at this tomorrow. 

Best regards,
Kaja

[Kaja]

Hi Jared, 

I don’t have access to a DK now, so I can’t reproduce it at the moment. 

But VTG is a detect pin, to select to connect the debugger to en external device. 

Can be that when you connect to the VTG, while power is on, you get an unsupported stat, so that you need to repower again. Because you switch what device the debugger is connected to. 

Best regards, 

Kaja

[Jared]

Hi Kaja 

Sorry but this isn't what I'm talking about.

The XC9236 latches off and does not power back on without a power cycle. Please try on an DK as I've described before.

[Stian Røed Hafskjold]

Hi Jared, thanks for the description. I will try this on a DK and get back to you soon.

[Jared]

Thanks Stian!

[Stian Røed Hafskjold]

Hi, I'm not able to reproduce this here. Are you able to measure the XC9236 output on TP50, to rule out that it's not the VIO_REF voltage follower that is the problem here? Thanks.

How is the external device connected to the DK? On P20 or the 10 pin SWD debug out connector, which pins are being used, etc. Does the external board use power from the DK, or is it powered from a separate supply, and if so, at which voltage? Or are you able to reproduce this without any connections to the external device?