Mamejay
Posts: 5
Joined: Fri Feb 20, 2015 2:55 am

Raspberry Pi 3B+ 3.3v short to ground

Sat Jul 21, 2018 4:31 am

Hi everyone,
I have a Raspberry Pi 3B+ which looks to have 3.3v shorted to ground.
This was a brand new unit which I was using for about 30 minutes. Switch it off and the next day tried to fire it back up again and it would not boot. Power LED comes on but that's it.
I did not connect anything strange to the GPIO's so not sure what is happening. Anyway further investigation shows 3.3v is shorted to ground.
From my reading the MXL7704-R3 chip does the voltage regulation on the 3B+. What I want to do is isolate the 3.3v easily to see which direction the 3.3v is shorted without removing the MXL7704-R3.
Is there any trace or passive component I can remove/cut to isolate it?
I did receive these for free with a offer from arrow.com so I am not too upset about it but would like to try and get this going. Or at least work out what failed.

Any help would be great. thanks

klricks
Posts: 5824
Joined: Sat Jan 12, 2013 3:01 am
Location: Grants Pass, OR, USA
Contact: Website

Re: Raspberry Pi 3B+ 3.3v short to ground

Sat Jul 21, 2018 4:51 am

Mamejay wrote:
Sat Jul 21, 2018 4:31 am
Hi everyone,
I have a Raspberry Pi 3B+ which looks to have 3.3v shorted to ground.
This was a brand new unit which I was using for about 30 minutes. Switch it off and the next day tried to fire it back up again and it would not boot. Power LED comes on but that's it.
I did not connect anything strange to the GPIO's so not sure what is happening. Anyway further investigation shows 3.3v is shorted to ground.
From my reading the MXL7704-R3 chip does the voltage regulation on the 3B+. What I want to do is isolate the 3.3v easily to see which direction the 3.3v is shorted without removing the MXL7704-R3.
Is there any trace or passive component I can remove/cut to isolate it?
I did receive these for free with a offer from arrow.com so I am not too upset about it but would like to try and get this going. Or at least work out what failed.

Any help would be great. thanks
I don't think there is any component that can be removed to isolate the 3V3 output short of lifting the 3V3 output pin 25.
Edit: Per the schematic at grid 2-D: https://www.raspberrypi.org/documentati ... educed.pdf
Either pin 25 or L7 would probably need to be removed as well in order to determine if the short is internal or external to the PMIC chip.

This thread might be of interest: viewtopic.php?f=28&t=218134
Last edited by klricks on Sat Jul 21, 2018 4:18 pm, edited 1 time in total.
Unless specified otherwise my response is based on the latest and fully updated Raspbian Stretch w/ Desktop OS.

hippy
Posts: 3750
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Raspberry Pi 3B+ 3.3v short to ground

Sat Jul 21, 2018 11:20 am

Mamejay wrote:
Sat Jul 21, 2018 4:31 am
From my reading the MXL7704-R3 chip does the voltage regulation on the 3B+. What I want to do is isolate the 3.3v easily to see which direction the 3.3v is shorted without removing the MXL7704-R3.
Is there any trace or passive component I can remove/cut to isolate it?
My guess, with reference to the thread klricks links to, is that the PMIC has probably stopped outputting 3V3 rather than there being a short.

It would probably be quite tricky identifying which tracks you should cut to allow the PMIC to generate 3V3, if it can, but not pass that to other parts of the circuit. Allowing it to generate other voltages but not 3V3 could damage the SoC and other components.

It would be worth checking if it is just 3V3 not present, or also 3V3A, 1V8, DDR_1V2 and VDD_CORE, what voltages are present on the PMIC's SEQ_EN, GLOBAL_EN, PG1 and PG2 pins. That would give a clearer picture of the failure mode.

As to what could have caused a 3V3 failure; that's difficult to say. Accidentally or inadvertently shorting 3V3 to 0V on GPIO or elsewhere could do that. We don't have full circuit diagrams so don't know exactly where 3V3 goes.

It could also be a failure on some other rail or elsewhere which has stopped 3V3 being generated. Lack of 3V3, appearing to be shorted to 0V, may just be a symptom.

drgeoff
Posts: 8483
Joined: Wed Jan 25, 2012 6:39 pm

Re: Raspberry Pi 3B+ 3.3v short to ground

Sat Jul 21, 2018 12:14 pm

With no power input to the RPi, measuring the resistance between the 3.3 volt rail and ground might be informative. Be sure to get the test polarity correct.

klricks
Posts: 5824
Joined: Sat Jan 12, 2013 3:01 am
Location: Grants Pass, OR, USA
Contact: Website

Re: Raspberry Pi 3B+ 3.3v short to ground

Sat Jul 21, 2018 2:02 pm

In my case I believe I briefly shorted 3V3 to ground. (Pin 1 to the nearby metal chip cover with the raspberry logo). The screen instantly went off and the RPi never booted again.

The 3V3 pin 1 to ground measures 0 Ohms.

With 5V0 power applied I found 5V0 in places but no other voltages near the PMIC. I probed caps near the PMIC as the pins on the chip are too small for me to probe.

I am guessing the PMIC will shut down all outputs when a short is detected.
Unless specified otherwise my response is based on the latest and fully updated Raspbian Stretch w/ Desktop OS.

hippy
Posts: 3750
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Raspberry Pi 3B+ 3.3v short to ground

Sat Jul 21, 2018 5:58 pm

klricks wrote:
Sat Jul 21, 2018 2:02 pm
I am guessing the PMIC will shut down all outputs when a short is detected.
That seems to be correct from my reading of the datasheet. Though as it's a programmable chip and what it actually does can be programmed, unless how it is programmed is known, it would not be possible to say what the exact behaviour is.

That's why I say lack of 3V3 may only be a symptom.

In your case you actually shorted the 3V3 and that may have damaged the 3V3 supply in such a way it won't come-up now. Or it could have collapsed the 3V3 rail, while leaving other rails up, which could have consequently damaged something else on the board. That damage may now be tripping something, possibly keeping everything from coming up, and causing 3V3 not to appear even though the 3V3 itself may be capable of coming up otherwise.

In earlier designs any power rail related fault condition would probably cause a cascade failure, ultimately take out the polyfuse, close down all power rails which feed from that, and, with a good wind, everything would recover after some time.

With a PMIC chip that seems to have changed. Only a 5V rail related fault would likely take out the polyfuse. Only an RPT engineer could tell us exactly what does or should happen under various power rail related fault conditions.

drgeoff
Posts: 8483
Joined: Wed Jan 25, 2012 6:39 pm

Re: Raspberry Pi 3B+ 3.3v short to ground

Sat Jul 21, 2018 8:17 pm

Momentarily* shorting the 3.3 volt rail to ground should not result in any device other than the PMC putting a permanent short to ground.

(* I use the word in its original sense, not as abused by some in the US who mean "very soon". https://en.wiktionary.org/wiki/momentarily)

hippy
Posts: 3750
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Raspberry Pi 3B+ 3.3v short to ground

Sat Jul 21, 2018 9:59 pm

drgeoff wrote:
Sat Jul 21, 2018 8:17 pm
Momentarily* shorting the 3.3 volt rail to ground should not result in any device other than the PMC putting a permanent short to ground.
But it's extremely unpredictable what happens when a device which uses multiple power rails loses one of those power rails. That can have knock-ons, can cause other things to misbehave or fail.

OmarAhmad
Posts: 1
Joined: Fri Sep 14, 2018 7:53 am

Re: Raspberry Pi 3B+ 3.3v short to ground

Fri Sep 14, 2018 7:57 am

drgeoff wrote:
Sat Jul 21, 2018 8:17 pm
Momentarily* shorting the 3.3 volt rail to ground should not result in any device other than the PMC putting a permanent short to ground.

(* I use the word in its original sense, not as abused by some in the US who mean "very soon". https://en.wiktionary.org/wiki/momentarily)
I am also facing the same problem (i.e. 3.3V rail and Gnd are shorted out), is there a solution?

LTolledo
Posts: 636
Joined: Sat Mar 17, 2018 7:29 am

Re: Raspberry Pi 3B+ 3.3v short to ground

Fri Sep 14, 2018 10:10 am

From the looks of this the only clear solution is the "US$ 35.00 solution".....
"Don't come to me with 'issues' for I don't know how to deal with those
Come to me with 'problems' and I'll help you find solutions"

Some people be like:
"Help me! Am drowning! But dont you dare touch me nor come near me!"

hippy
Posts: 3750
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Raspberry Pi 3B+ 3.3v short to ground

Sat Sep 15, 2018 9:27 am

OmarAhmad wrote:
Fri Sep 14, 2018 7:57 am
I am also facing the same problem (i.e. 3.3V rail and Gnd are shorted out), is there a solution?
Member 'deNiro' reports they replaced their PMIC chip and somewhat restored operation -

https://www.raspberrypi.org/forums/view ... 4#p1319930

That's likely not an easy job if you don't have the right tools and experience and as deNiro notes it's not ( so far ) a perfect solution.

As far as I know; no one has explained the failure mode, figured out why a short circuit the PMIC should mitigate causes the PMIC to 'brick itself' and never work again.
jbudd wrote:
Fri Sep 14, 2018 10:49 pm
"It's a 3B+!" was the sound that first came to their ears,
Not sure snarkiness is appreciated on this forum. <drumroll>

univok
Posts: 2
Joined: Mon Sep 24, 2018 11:59 am

Re: Raspberry Pi 3B+ 3.3v short to ground

Mon Sep 24, 2018 12:17 pm

I have exactly the same problem here, the MxL7704 is very likely shortened to GND at Pin 25 (VOUT1). I didn't replace it until now, it is hard to find, soldering will be a challenge and shipping exceeds the costs of a new Pi B3+ .
And I have to chhose: the MxL7704 as "-AQB 1.2V" vs. "-XQB 0.85V" for VOUT4 which is VDD_CORE on the Raspberry - any Idea which is the best?

hippy
Posts: 3750
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Raspberry Pi 3B+ 3.3v short to ground

Mon Sep 24, 2018 2:40 pm

univok wrote:
Mon Sep 24, 2018 12:17 pm
And I have to chhose: the MxL7704 as "-AQB 1.2V" vs. "-XQB 0.85V" for VOUT4 which is VDD_CORE on the Raspberry - any Idea which is the best?
I cannot answer that but if you have a working Pi 3B+ you can probably determine it from the power-up sequencing order, Figures 14 and 16 in the datasheet -

AQB : LDO, Vout1, Vout2, Vout4+Vout3

XQB : Vout4, Vout2, Vout3+Vout1+LDO

Looking at what the start-up sequence is for other Pi SoC's could give an indication as to which sequence is the best fit.

Added: https://www.raspberrypi.org/blog/pi-power-supply-chip
In designing the power chain for the Pi 2 and 3, the sequencing was fairly straightforward: power rails power up in order of voltage (5V, 3.3V, 1.8V, 1.2V).
This suggests AQB to me, LDO and Vout1 before the rest.

I did think Tables 7 and 8 values for the associated "L" coils might help -

AQB : V1 = 2.2uH, V2 = 1uH, V3 = 0.47uH, V4 = 0.47uH

XQB : V1 = 2.2uH, V2 = 2.2uH, V3 = 1uH, V4 = 0.47uH

But looking at the reduced schematics -

PCB : V1 = 2R2, V2 = 2R2, V3 = 470n, V4 = 470n

I'm assuming "2R2" means "2u2".

The V2 suggests it's an XQB but V3 suggests its an AQB. Go figure.

Maybe it's neither, is a special propriety variant designed for the Raspberry Pi ?

Maybe it's a transcription error on the schematic, like the "2R2". Maybe it's a design error as with the PoE HAT. Or maybe that's what turned out to work best despite what the datasheet suggests. Maybe the datasheet is wrong.

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 20447
Joined: Sat Jul 30, 2011 7:41 pm

Re: Raspberry Pi 3B+ 3.3v short to ground

Mon Sep 24, 2018 3:27 pm

If it broke without using the GPIO's or shorting anything out, and it's within warrantee, then it may simply have failed and you should approach the place you got it from for a replacement

If it was something you did (the most likely situation) sadly the easiest and cheapest fix is to buy a new Pi.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Please direct all questions to the forum, I do not do support via PM.

univok
Posts: 2
Joined: Mon Sep 24, 2018 11:59 am

Re: Raspberry Pi 3B+ 3.3v short to ground

Tue Sep 25, 2018 8:05 am

hippy wrote:
Mon Sep 24, 2018 2:40 pm
Added: https://www.raspberrypi.org/blog/pi-power-supply-chip
In designing the power chain for the Pi 2 and 3, the sequencing was fairly straightforward: power rails power up in order of voltage (5V, 3.3V, 1.8V, 1.2V).
This suggests AQB to me, LDO and Vout1 before the rest.
Great breakdown, thank you very much for your effort!
jamesh wrote:
Mon Sep 24, 2018 3:27 pm
... the easiest and cheapest fix is to buy a new Pi.
Yes, that's what I did already.
Because I am not sure that the Pi crashed on its own (I was using a Pimoroni Display-O-Tron when it happened). The board will be put on hold for inspection, and if I am sure that it is the PMIC that causes the short circuit, I'll try to replace it (have to buy a hot air soldering station before, but this will complete my tooling equipment anyway...).

Return to “Troubleshooting”

Who is online

Users browsing this forum: No registered users and 48 guests