User avatar
apurva_v8
Posts: 24
Joined: Sat Jan 04, 2020 5:24 pm
Location: Pune, Maharastra
Contact: Website Twitter

Dead after 5.5V Input at micro USB

Tue Feb 18, 2020 1:42 pm

I was using my 3B+ usually Like a normal day. I Had Xfce installed And I had a power button also customized to turn off the RPI. So I was using a
5 V 3A adapter.

But I went Somewhere and came back after 30 mins and my RPI stopped working. When I checked the output at the micro USB port with multimeter It was showing 5.5V and the current was fluctuating as well. Btw I used female Micro USB to check this not on the Raspberry pi board.

I checked the poly fuse. I even Checked by bypassing polyfuse with a jumper. Still, my RPI won't boot. I checked after that my 3.3V and GND were Shorted. I am really Sad. I used to use it as my desktop as well I used it for coding purposes. Can anything be done in this scenario? Can I fix it? With any modification. I am a technical Guy so all help to this are welcome.

I tried reinstalling OS multiple times but the green light won't blink. It's just showing a steady Red Light. What might have happened?

I wonder why they cut so many things to make it cheaper. There should be reverse polarity protection as well. I think things should be more tinker able and protected.

thanks for the help in Advance :geek:
I hope my reply helps you in some way.

Apurva
  • Embedded Engineer

User avatar
RaTTuS
Posts: 10559
Joined: Tue Nov 29, 2011 11:12 am
Location: North West UK
Contact: Twitter YouTube

Re: Dead after 5.5V Input at micro USB

Tue Feb 18, 2020 1:53 pm

probably your wiring has had a meltdown
how had you connected everything to the RPi
show all the links i.e.
I had a power button also customized to turn off the RPI
^^ how

it's not the foundations problem it's something you have done what make of RPi ?>
#
How To ask Questions :- http://www.catb.org/esr/faqs/smart-questions.html
WARNING - some parts of this post may be erroneous YMMV

1QC43qbL5FySu2Pi51vGqKqxy3UiJgukSX
Covfefe

User avatar
apurva_v8
Posts: 24
Joined: Sat Jan 04, 2020 5:24 pm
Location: Pune, Maharastra
Contact: Website Twitter

Re: Dead after 5.5V Input at micro USB

Tue Feb 18, 2020 2:00 pm

I used this https://howchoo.com/g/mwnlytk3zmm/how-t ... spberry-pilink to make a button. It was working fine. I even tried removing everything. I used jumpers and I didn't see any meltdown of wires. Everything was in place.
the button was isolated properly and was fixed in the case by drilling a hole.
I have official RPI 3B+ (element14)
Currently, the board is at my office. I am still looking to do some modifications to fix it.
I hope my reply helps you in some way.

Apurva
  • Embedded Engineer

User avatar
davidcoton
Posts: 5020
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK
Contact: Website

Re: Dead after 5.5V Input at micro USB

Tue Feb 18, 2020 2:51 pm

The Pi3B+ does have overvoltage and reverse polarity protection. Part of the overall protection scheme is the polyfuse. If you apply power via the GPIO header you bypass the polyfuse, so if the TVS diode protection trips, it will probably sacrifice itself to save your Pi.

Bypassing the polyfuse is a very poor diagnostic test, if it has tripped for a good reason then bypassing it will cause more damage.

Something is wrong if your 5V supply gives 5V5 under load -- it might be acceptable (but not good) as an open circuit voltage.

Get a different, known working PSU, connected by the microUSB port, and preferably the official 5V1 2A5 version. Try the Pi with that. If is doesn't work, and you are good with a soldering iron, remove the TVS diode and try again. If is works, shut down and fit a new TVS diode before using it again -- mistakes/faults without a TVS diode fitted will destroy your Pi.

If that doesn't work, it's probable that some external fault involving the 3V3 GPIO pin has killed the PMIC (occasionally PMICs seem to die for no obvious reason, but this is rarer than owners would like to think). If you have SMT rework tools you can replace the PMIC, but you need the right version for the Pi. May be available from RPT in quantities of 100.
Signature retired

User avatar
apurva_v8
Posts: 24
Joined: Sat Jan 04, 2020 5:24 pm
Location: Pune, Maharastra
Contact: Website Twitter

Re: Dead after 5.5V Input at micro USB

Tue Feb 18, 2020 3:08 pm

I tried bypassing polyfuse with a good power adapter giving 5V 2.5A. My 3v3 and GND are short. will changing TVS diode fix it?
I didn't apply a voltage through GPIO. I did with the Micro USB port
Can you please share me the link to the PMIC I wanna know where it's available.
I have all the tools available like SMT station and I can solder SMD components also. I need to know if it will possibly fix the 3v3 and GND short after changing the TVS diode.
supply was giving 5.5V in no-load, checking through the multimeter.

Thanks for the help in advance :geek:
I hope my reply helps you in some way.

Apurva
  • Embedded Engineer

User avatar
rpdom
Posts: 17164
Joined: Sun May 06, 2012 5:17 am
Location: Chelmsford, Essex, UK

Re: Dead after 5.5V Input at micro USB

Tue Feb 18, 2020 6:10 pm

apurva_v8 wrote:
Tue Feb 18, 2020 3:08 pm
My 3v3 and GND are short. will changing TVS diode fix it?
If your 5V reads 5V and your 3.3V reads 0V on a 3B+, the PMIC chip has died and you need to replace the Pi.
Can you please share me the link to the PMIC I wanna know where it's available.
It isn't available. It is a custom version of the chip for the Pi.

A representative of Raspberry Pi Trading has said that they may be able to supply the custom chip in quantities of 100 or more for a price. I don't know if anyone has taken them up on this offer.
Unreadable squiggle

User avatar
davidcoton
Posts: 5020
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK
Contact: Website

Re: Dead after 5.5V Input at micro USB

Tue Feb 18, 2020 9:52 pm

Read my previous answer again, and rpdom's post. Try to understand what is written, not what you think it says.

Bypassing the polyfuse is not a good idea.

With no power supply, you can measure the resistance of the polyfuse. It should be low (virtual short circuit). If not, give it time to recover and re-check. If it trips again, try the TVS stuff.

You cannot get meaningful results by measuring the resistance between 3V3 and GND.

With power connected, check that 5V is present after the polyfuse (eg at the GPIO header). Same for 3V3. If you have no 5V, try the TVS stuff. If you have 5V and not 3V3, it's the PMIC at is dead.

Search this forum for posts from jamesh about PMIC supply.
And finally, partial schematics are available here.
Signature retired

renice123
Posts: 198
Joined: Tue Oct 15, 2019 3:56 pm

Re: Dead after 5.5V Input at micro USB

Wed Feb 19, 2020 7:00 am

Sorry, if your adapter produces 5.5 V, then you have a bad adapter and most likely you will not be able to replace PMIC
The adapter, which suddenly started working on 5.5 V, is garbage found somewhere in the wilds of the Amazon from the remnants of an airplane from the Second World War. Or you are disastrously out of luck. What power supply did you use?

For some reason, I think that you used the Chinese "adaptive adapter", which are used because of a low-quality tuning resistor to suddenly change the voltage.
Start by buying a good power supply with good stabilization. And most likely you will not replace PMIC - it is a simple chip, but the installation of Raspberry is small and it takes at least a certain kind of skill to replace PMIC. Well, I have not heard of this among amateurs. But perhaps you are a high-class specialist, but judging by the fact that you used the wrong devices (maybe), you are new to electronics. Just buy a new Raspberry and a regular power supply.

PS Reverse polarity protection costs $ 1 (if you don’t do it yourself), overvoltage protection costs from 13 to 60 (and higher) dollars, and thermal protection and undervoltage protection will also be included in the service. There is no problem adding protection of any complexity to Raspberry, in particular, with measuring current and voltage and automatic shutdown at the slightest malfunction. You can even work continuously with UPS, which eliminates accidental shutdown.
 There is no secret in what happened to you: your power supply has deteriorated, and it has generated an increased voltage.

This simple circuit has been working for radio amateurs for decades.
It can be soldered in five minutes! But now you are able to buy or solder a much more reliable circuit with tremendous speed and galvanic isolation

Image
Also buy Arduino and connect the power supply through
https://www.tindie.com/products/yesyes/ ... es-001ohm/

Image
Arduino will also provide you with thermal protection (if you connect a temperature sensor).

In addition to additional protection, you also get the opportunity to log power.

Good luck in understanding the electronics! But it’s better to start with a simple one than to try to fix something that requires industrial equipment.

hippy
Posts: 7713
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Dead after 5.5V Input at micro USB

Wed Feb 19, 2020 5:25 pm

apurva_v8 wrote:
Tue Feb 18, 2020 3:08 pm
Can you please share me the link to the PMIC I wanna know where it's available.
Google would be the best bet. You want an "MxL7704-R3" - that "-R3" is the important but.

There was a post on the forum from someone saying they had purchased one, replaced the failed PMIC on their Pi, and it was working again but there's no guarantee of success if anything else has been damaged.

Found it : viewtopic.php?t=240280#p1524527

I wouldn't have expected 5V5 to destroy the PMIC or damage anything else, so the cause may be something other than that, or you just got unlucky.

User avatar
Imperf3kt
Posts: 3620
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: Dead after 5.5V Input at micro USB

Wed Feb 19, 2020 7:09 pm

Not sure where that poster found the exact part they needed, but the chip in question cannot be bought from Amazon as it is a custom chip fabricated solely for the Raspberry Pi and sold only to the Raspberry Pi foundation.
55:55:44:44:4C
52:4C:52:42:41

renice123
Posts: 198
Joined: Tue Oct 15, 2019 3:56 pm

Re: Dead after 5.5V Input at micro USB

Wed Feb 19, 2020 9:11 pm

Link to Russian video is a link to the site of a specialist radio engineer. He is repairing electronics. Since not everyone understands the Russian language, I will translate what the master talks about.
He says that one cannot simply “solder” PMIC without special devices - for example, without a special device for heating the entire board. “Balloons will float” under the processor or the board will be twisted by a screw. That's exactly what he says. So it is better to immediately forget about the repair without at least quite expensive specialized devices.
This is the first point. And it’s just that the repair requires qualifications (which the author of the video blog has - he is engaged in the repair of radio electronics, this is his business).
The second observation is even simpler. There is a datasheet on the PMIC, the voltage limit is 5.5 V (this means that the stabilizer will work, but not necessarily stably and reliably). But this is precisely PMIC. There is a raspberry passport, where voltage limits are also indicated. Of course, you can overestimate the tension you like, but why then complain that “Raspberry is burnt”? 5V / 2.5A DC power input written in black and white
https://www.raspberrypi.org/products/ra ... el-b-plus/
It is clear that there are strength limits, but why test the device at the limit values? What's the point? See if it burns or not burns?

And finally. If the power supply began to produce chaotic voltage, then this means that such a power supply can produce either 5.8 V or 4.3 V. In other words, we can’t talk about stable power, the power supply is also ejected.
A stabilized power supply should always give the same voltage, and not jump either to 5 V or to 6. Otherwise, such a power supply is considered to be faulty (or unstabilized)

User avatar
apurva_v8
Posts: 24
Joined: Sat Jan 04, 2020 5:24 pm
Location: Pune, Maharastra
Contact: Website Twitter

Re: Dead after 5.5V Input at micro USB

Thu Feb 20, 2020 10:27 am

He says that one cannot simply “solder” PMIC without special devices - for example, without a special device for heating the entire board. “Balloons will float” under the processor or the board will be twisted by a screw. That's exactly what he says. So it is better to immediately forget about the repair without at least quite expensive specialized devices.
This is the first point. And it’s just that the repair requires qualifications (which the author of the video blog as - he is engaged in the repair of radio electronics, this is his business).
I am an engineer myself. I work at a good firm and I have everything from micro solder, flux, heat gun, tweezers, power supply, etc,..., every workbench utility we have here as we also do the production. I have worked with SMD components and ICs so I don't think that would be a problem for me. But I think I will get a replacement for my RPI from the adapter company as it was in warranty. :geek:
I hope my reply helps you in some way.

Apurva
  • Embedded Engineer

hippy
Posts: 7713
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Dead after 5.5V Input at micro USB

Thu Feb 20, 2020 11:07 am

Imperf3kt wrote:
Wed Feb 19, 2020 7:09 pm
Not sure where that poster found the exact part they needed, but the chip in question cannot be bought from Amazon as it is a custom chip fabricated solely for the Raspberry Pi and sold only to the Raspberry Pi foundation.
They say they bought their part from AliExpress and there are a number of sources claiming to sell MxL7704-R3 which is why I pointed towards Google. There are similar sources on Amazon which turn up with a Google search. I don't know if any of those are legitimate or as advertised which was another reason for pointing towards Google. Emptor Caveat, but the linked poster claims their replacement worked for them.

The Foundation has said it will sell MxL7704-R3 chips with, I recall, a MOQ of 100 so the sellers could have bought from the Foundation and be reselling their excess. Which is what I would expect anyone who wanted one but had to buy a hundred would do.

renice123
Posts: 198
Joined: Tue Oct 15, 2019 3:56 pm

Re: Dead after 5.5V Input at micro USB

Thu Feb 20, 2020 4:57 pm

Any manufacturer likes to spread myths about some “special technologies”. People like to believe that they have “special designs.” Raspberries were collected, including in China, make your own conclusions whether the Chinese have access to PMIC =)
Finding the right chip is not a problem at all. They are sold even in our country and cost $ 3. But, of course, you can continue to believe that there are unique microcircuits created specifically for an inexpensive device and available to VIPs. However, there are such microcircuits - Soviet microcircuits created for military purposes)
Image
But now such microcircuits are sold in radio markets in Russia (they are bought because of gold)
Ali is full of PMIC offers, and this is understandable: any major manufacturer is required to provide repair shops with spare parts. Even Apple is forced to abandon proprietary decisions, you know?
Any company must provide parts for repair. Of course, if this is not a Chinese shed of Uncle Liao's family.
https://aliexpress.ru/i/33012950545.html
https://aliexpress.ru/item/33049459362. ... archweb0_0, searchweb201602_, searchweb201603_
There are a lot of offers, by the way.
If you have all the necessary equipment, mainly a board heating device and temperature-controlled soldering irons, then you will replace this chip in about ten minutes. It is also implied that you really know how to work with smd components. On the video and in the accompanying article, the necessary temperatures and even the composition of the solder for repair are indicated. So this is just a trivial howto
Fortunately, RPIs do not varnish their microcircuits and do not use very advanced motherboard manufacturing technologies (as in some cell phones that turn into disposable ones, and the components are glued with a compound).
Here are the repair data, if anyone is interested.
Chips MXL7704 (RE is just another designation, not a “special chip for Raspberries, engineer writes that you can take the MXL7704-AQB-T QFN32): https://vk.cc/9B0NWV and https://vk.cc/9B0O30
The master used the Quick 857DW + soldering station - anti-static gloves, tweezers, homemade LBP, MS8268 household multimeter, 7 "separator (UYUE 948/988 - this is expensive in our country, they are often heated on gas separators), a microscope and a T12 soldering station , microwave tip for T12, flux NC-559, solder 0.8 mm 100 g with flux, copper braid for removing solder

User avatar
Imperf3kt
Posts: 3620
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: Dead after 5.5V Input at micro USB

Thu Feb 20, 2020 7:32 pm

I based my reply on this.
https://www.raspberrypi.org/blog/pi-power-supply-chip/
This
https://www.maxlinear.com/product/power ... cs/mxl7704
And several forum posts I cannot currently find
55:55:44:44:4C
52:4C:52:42:41

renice123
Posts: 198
Joined: Tue Oct 15, 2019 3:56 pm

Re: Dead after 5.5V Input at micro USB

Thu Feb 20, 2020 8:32 pm

After reading this article, some kind of too sad thoughts come.
A huge team of engineers and managers made a “universal reliable device”, but for some reason it was silent in every way that this device died and apparently died like a fly in spider webs, and even fixing this “ingenious PMIC” is not more difficult than replace the memory chip in the smartphone. But for this you need to buy 100 memory chips! (if you don’t turn to Chinese merchants who understood everything for a long time)
There are many beautiful idioms in the Russian language to call such a situation truthful words
Honestly, I am now more disappointed than ever.

SteveMann
Posts: 78
Joined: Sat Dec 08, 2012 10:39 pm

Re: Dead after 5.5V Input at micro USB

Thu Feb 20, 2020 8:58 pm

It's unlikely that the reset button fried your Pi- if it were to do so, it would have failed instantly. Not "after a while".
It is also unlikely that 5.5V would damage it, either.
What else is connected to the Pi 3.3V?

You said: " I checked after that my 3.3V and GND were Shorted."

This is nonsense. Measuring the resistance of a power rail to ground is usually meaningless. What voltage did you measure?
SteveMann
Still a Novice after four years...

klricks
Posts: 7154
Joined: Sat Jan 12, 2013 3:01 am
Location: Grants Pass, OR, USA
Contact: Website

Re: Dead after 5.5V Input at micro USB

Fri Feb 21, 2020 12:00 am

SteveMann wrote:
Thu Feb 20, 2020 8:58 pm
It's unlikely that the reset button fried your Pi- if it were to do so, it would have failed instantly. Not "after a while".
It is also unlikely that 5.5V would damage it, either.
What else is connected to the Pi 3.3V?

You said: " I checked after that my 3.3V and GND were Shorted."

This is nonsense. Measuring the resistance of a power rail to ground is usually meaningless. What voltage did you measure?
I destroyed my RP3B+ by accidentally shorting 3V3 to ground (metal chip cover with RPi logo) with my meter probe. The RPi instantly shut down and never booted again.
I agree that measuring resistance of a power rail does not indicate much about function of the rail but It is usually some several K Ohms and may fluctuate as caps charge etc.
In my case I measured a steady 0.6 Ohms which I consider to be a dead short.
When powered I measured ~0 VDC on the 3V3 rail.
Unless specified otherwise my response is based on the latest and fully updated RPiOS Buster w/ Desktop OS.

hippy
Posts: 7713
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Dead after 5.5V Input at micro USB

Fri Feb 21, 2020 12:11 pm

SteveMann wrote:
Thu Feb 20, 2020 8:58 pm
You said: " I checked after that my 3.3V and GND were Shorted."

This is nonsense.
Technically yes, but colloquially that is how it appears to most people. They will work on the principle that "if it's shows 0V or there about it must therefore be shorted to 0V".

And, as I believe the MxL7704 has clamping MOSFET's potentially shorting to 0V on its outputs it could actually be the case.

cleverca22
Posts: 775
Joined: Sat Aug 18, 2012 2:33 pm

Re: Dead after 5.5V Input at micro USB

Fri Feb 21, 2020 4:13 pm

ive heard that with older pi a models, you could force-feed 3.3v into BOTH the 5v and 3.3v pins of the gpio header, and ignore the usb entirely

could similar things be used to bypass a dead PMIC safely? or could the PMIC have failed with the outputs shorted?

definitely would want to use a current-limited bench power-supply for such testing

may want to wait for extra input from an engineer before trying it, enless you dont care about the "dead" pi and dont mind it becoming deader

klricks
Posts: 7154
Joined: Sat Jan 12, 2013 3:01 am
Location: Grants Pass, OR, USA
Contact: Website

Re: Dead after 5.5V Input at micro USB

Fri Feb 21, 2020 4:24 pm

cleverca22 wrote:
Fri Feb 21, 2020 4:13 pm
ive heard that with older pi a models, you could force-feed 3.3v into BOTH the 5v and 3.3v pins of the gpio header, and ignore the usb entirely

could similar things be used to bypass a dead PMIC safely? or could the PMIC have failed with the outputs shorted?

definitely would want to use a current-limited bench power-supply for such testing

may want to wait for extra input from an engineer before trying it, enless you dont care about the "dead" pi and dont mind it becoming deader
No... The system communicates with the PMIC during boot and the PMIC generates several other voltages which will not not come up if the PMIC does not initialize properly.
See: https://www.raspberrypi.org/documentati ... educed.pdf
Unless specified otherwise my response is based on the latest and fully updated RPiOS Buster w/ Desktop OS.

cleverca22
Posts: 775
Joined: Sat Aug 18, 2012 2:33 pm

Re: Dead after 5.5V Input at micro USB

Fri Feb 21, 2020 4:29 pm

ah, maybe that trick only worked on the rpi1, it had a much simpler power setup i think, and the SoC itself did most of the buck regulation

it also depends on if the PMIC is 100% dead, or only lost a 3.3v rail

User avatar
rpdom
Posts: 17164
Joined: Sun May 06, 2012 5:17 am
Location: Chelmsford, Essex, UK

Re: Dead after 5.5V Input at micro USB

Fri Feb 21, 2020 4:43 pm

cleverca22 wrote:
Fri Feb 21, 2020 4:29 pm
ah, maybe that trick only worked on the rpi1, it had a much simpler power setup i think, and the SoC itself did most of the buck regulation

it also depends on if the PMIC is 100% dead, or only lost a 3.3v rail
The Pi 1 had three linear voltage regulators. One for 3.3V, one for 1.8V and one (I believe...hang on, why am I guessing? I've still got the schematics) for 2.5V. The rest of the voltages were generated inside the BCM2835 and the LAN9512 chips (although there was a clash between the 1.8V regulator and the LAN9512's internal 1.8V regulator on early versions!)

The latest Pis all use that PMIC to provide those voltages and start them up in the correct sequence (that's the "custom" bit - I believe the chip can be programmed to provide the voltage in the right order vi an i2c interface).
Unreadable squiggle

drgeoff
Posts: 10765
Joined: Wed Jan 25, 2012 6:39 pm

Re: Dead after 5.5V Input at micro USB

Fri Feb 21, 2020 5:09 pm

renice123 wrote:
Thu Feb 20, 2020 4:57 pm
However, there are such microcircuits - Soviet microcircuits created for military purposes
Image
If you are implying that the chip in that photo is a Soviet microcircuit created for military purposes you are incorrect. It was a standard catalogue UV erasable EPROM made by Intel. I think I still have a board with 12 of them somewhere in the loft.

cleverca22
Posts: 775
Joined: Sat Aug 18, 2012 2:33 pm

Re: Dead after 5.5V Input at micro USB

Fri Feb 21, 2020 5:31 pm

rpdom wrote:
Fri Feb 21, 2020 4:43 pm
cleverca22 wrote:
Fri Feb 21, 2020 4:29 pm
ah, maybe that trick only worked on the rpi1, it had a much simpler power setup i think, and the SoC itself did most of the buck regulation

it also depends on if the PMIC is 100% dead, or only lost a 3.3v rail
The Pi 1 had three linear voltage regulators. One for 3.3V, one for 1.8V and one (I believe...hang on, why am I guessing? I've still got the schematics) for 2.5V. The rest of the voltages were generated inside the BCM2835 and the LAN9512 chips (although there was a clash between the 1.8V regulator and the LAN9512's internal 1.8V regulator on early versions!)

The latest Pis all use that PMIC to provide those voltages and start them up in the correct sequence (that's the "custom" bit - I believe the chip can be programmed to provide the voltage in the right order vi an i2c interface).
yeah, so the rpi1 could likely make all of those if you feed it only 3.3v, assuming the 2.5 is an LDO linear

Return to “Troubleshooting”