Demisi
Posts: 6
Joined: Thu Feb 14, 2019 10:29 pm

The Big (little) Yellow Bolt of Doom

Thu Feb 14, 2019 11:01 pm

OK so I've been literally all over the internet looking for a solution to the yellow lightning bolt that seems to have plagued so many people over the last several years. power cord are short (six inches)everything is stock only devices plugged in is a pi touch screen and a WiFi keyboard dongle. the original wall charger i was using was rated at 2 Amps and for some reason the Pi WiFi would not work so i got tired of everyone telling me to shorten my cables and just got a breadboard USB jack. as i write this I'm am sitting next to a tekpower TP300ST D.C. regulated power supply running at 5.1 Volts and 3 Amps, connected to the USB jack, connected to my Pi. this seems to have cleared up the Pi WiFi issue as i have finally updated from the original noobs install. However i still have a wonderful yellow lightning bolt telling my i am under-powered. it should be noted that the largest power draw I've seen so far on the power supply was 1.4 Amps during the upgrade. my big question after all this, is there any file in the Raspian OS that would force an under-powered even even when it has enough and where would i find it? If anyone has thoughts on my hardware setup or thinks i may have set something up inappropriately let me know. i can also provide picture of the setup if needed and am running a Rapberry Pi 3 B+.

User avatar
davidcoton
Posts: 4137
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: The Big (little) Yellow Bolt of Doom

Thu Feb 14, 2019 11:55 pm

Have you measured the voltage at the Pi's GPIO header 5V and Gnd pins? Be careful not to short anything (ideally use dupont wires) as doing so will kill the Pi.

What OS are you using?
Signature retired

User avatar
Imperf3kt
Posts: 2769
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 12:29 am

Have you tried the official power supply? It's the only power supply you'll be recommended here, for good reason.
Other supplies may work, but aren't guaranteed.
55:55:44:44:4C
52:4C:52:42:41

Demisi
Posts: 6
Joined: Thu Feb 14, 2019 10:29 pm

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 12:52 am

I'm using noobs downloaded about a month ago and just updated and upgraded last night, and I havnt botherd to get an official power supply as yet mostly becouse I have a 6 inch power cord and a dc power supply running to it through a breadboard, if 3 amps over that short a distance wont give me sufficent power I dont think an official power supply will fix my problem. I will say that the splitter that came with the screen has me a bit worried but everything powers on just fine and I dont have an inline voltage checker might be able to rig something up on the breadboard though.

User avatar
Imperf3kt
Posts: 2769
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 1:14 am

The general consensus is that those types of power supplies either don't operate at the rated specs (blatant marketing lies or they are showing their maximum rating which isn't the 'constant' rating), or they are noisy (transient spikes) or your cabling is introducing too much resistance, resulting in voltage drop.

I have been asked not to give advice, so I'll leave it up to you what to make of that information.
55:55:44:44:4C
52:4C:52:42:41

fruitoftheloom
Posts: 20694
Joined: Tue Mar 25, 2014 12:40 pm
Location: Delightful Dorset

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 2:34 am

Imperf3kt wrote:
Fri Feb 15, 2019 12:29 am
Have you tried the official power supply? It's the only power supply you'll be recommended here, for good reason.
Other supplies may work, but aren't guaranteed.

+1 any other Power Supply or Phone/Tablet Charger is a Lottery.

Also ensure you have an up to date Raspbian Stretch if running a 3B+ or 3A+ !
Retired disgracefully.....

Ernst
Posts: 1231
Joined: Sat Feb 04, 2017 9:39 am
Location: Germany

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 7:42 am

Demisi wrote:
Fri Feb 15, 2019 12:52 am
... I havnt botherd to get an official power supply as yet mostly becouse I have a 6 inch power cord and a dc power supply running to it through a breadboard, if 3 amps over that short a distance wont give me sufficent power I dont think an official power supply will fix my problem. I will say that the splitter that came with the screen has me a bit worried but everything powers on just fine and I dont have an inline voltage checker might be able to rig something up on the breadboard though.
see above, that is most likely the cause of your problems.

It will help if you give an accurate description of your configuration, if possible with photos. From your posts I get the impression that this is not a straightforward configuration.
The road to insanity is paved with static ip addresses

User avatar
rpiMike
Posts: 904
Joined: Fri Aug 10, 2012 12:38 pm
Location: Cumbria, UK

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 9:12 am

Buy the official power supply.

User avatar
RaTTuS
Posts: 10458
Joined: Tue Nov 29, 2011 11:12 am
Location: North West UK

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 10:48 am

or a decent micro USB lead - if it's made out of CCA or wet string your not going to have much luck anyway
a decent thick copper only cable will be fine
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

Idahowalker
Posts: 445
Joined: Wed Jan 03, 2018 5:43 pm

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 12:17 pm

Demisi wrote:
Fri Feb 15, 2019 12:52 am
I'm using noobs downloaded about a month ago and just updated and upgraded last night, and I havnt botherd to get an official power supply as yet mostly becouse I have a 6 inch power cord and a dc power supply running to it through a breadboard, if 3 amps over that short a distance wont give me sufficent power I dont think an official power supply will fix my problem. I will say that the splitter that came with the screen has me a bit worried but everything powers on just fine and I dont have an inline voltage checker might be able to rig something up on the breadboard though.
Ah, the good ole breadboard connection to power the RPi. What's the bread board current rated at? Using those premade 22ga jumper wires? Did you opt for the low cost hobbyist breadboard or did you get a high dollar engineering prototyping breadboard? These little things mage a difference.
Without knowing why you are deleting my postings, I will not know how...

Demisi
Posts: 6
Joined: Thu Feb 14, 2019 10:29 pm

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 12:40 pm

I'll be posting some pictures when I get home tonight, maybe even throw in some flir shots if I have the time to find it

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 12218
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 1:32 pm

The only way to measure if a PI receives stable and enough power is to measure directly on the board NOT on the power supply itself.
even with a perfect power supply, a bad or marginal cable will give rise to dips in received voltage, and if these dips go below 4.65V (plus or minus a few milli-volt detector tolerance) then an under-volt signal will be generated. (note the OS has nothing to do with it, the detection level is NOT programmable, but hard wired).

This happens because cables physically are not perfect conductors, with no resistance, and no inductance.
Real cables have both resistance, and also a little bit of inductance. The resulting effect is that variations in current draw, will induce variations of voltage loss. Something WiFi adapters, with their analog HF systems have much problems with.

Charger cables were not designed with low resistance in mind, as the only consequence of relatively high resistance is that it takes a few minutes more to charge up a battery, so there will be no consumer reaction to these cables being "bad".

User avatar
davidcoton
Posts: 4137
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: The Big (little) Yellow Bolt of Doom

Fri Feb 15, 2019 2:06 pm

mahjongg wrote:
Fri Feb 15, 2019 1:32 pm
(note the OS has nothing to do with it, the detection level is NOT programmable, but hard wired).
Different OSs may not have been properly adapted for the correct GPIO assignment to control the undervolt warning on the Pi3B+.

However, it does seem more likely that the PSU (a bench model that SHOULD be adequate, unless the spec is seriously over-hyped) is not actually delivering enough voltage to the Pi itself. Hence my other request, for an actual measurement of the on-board voltage at the GPIO header.
Without the right diagnostic information all suggested solutions are at best a guess -- though often a well-informed one.

I find it strange that a bench PSU is available, but not a multimeter to check the voltage :roll:
Signature retired

Demisi
Posts: 6
Joined: Thu Feb 14, 2019 10:29 pm

Re: The Big (little) Yellow Bolt of Doom

Sat Feb 16, 2019 2:07 am

First off thank you to Imperf3kt. I am here to learn what i did wrong or how i screwed something up not where i should spend my money so if you have advice i will always listen, it was helpful. Next another thanks to Ernst and Idahowalker, i have never heard that about breadboards before and it ultimately lead to the solution i found. basically i had a drop from the PSU connecting lead to the breadboard, it was minor 5.1 volts at PSU to about 4.95 volts going into the breadboard. then across the breadboard to the USB connector there was quite a significant drop to about 4.63 volts, then on to the USB cable. To those who suggested the breadboard I am using a Velleman SD35N. i found that if i bump the voltage up to 5.8 Volts on the PSU the power issues clear up and i get a reading of about 5.2 to 5.3 volts at the USB plug. i have a few pictures of the setup but apparently they are to big for this site so if anyone still want to see them let me know and ill find a hosting site to put them.

Finally davidcoton and mahjongg, i have seen this on a few different web site but I am a bit confused do i just test on any old 5 volt and ground pins or is there a pin pair i should be testing on the GPIO?

User avatar
Burngate
Posts: 6005
Joined: Thu Sep 29, 2011 4:34 pm
Location: Berkshire UK Tralfamadore
Contact: Website

Re: The Big (little) Yellow Bolt of Doom

Sat Feb 16, 2019 10:54 am

Demisi wrote:
Sat Feb 16, 2019 2:07 am
Finally davidcoton and mahjongg, i have seen this on a few different web site but I am a bit confused do i just test on any old 5 volt and ground pins or is there a pin pair i should be testing on the GPIO?
davidcoton wrote:
Fri Feb 15, 2019 2:06 pm
... measurement of the on-board voltage at the GPIO header.
The major voltage drops will be in cables, etc. connecting to the μUSB input socket, and in cables connecting to the USB sockets and GPIO header.
On the board itself, both 5v pins on the header can be considered identical, as can all gnd pins and metalwork on the various sockets.

So use either 5v pin, and any metal bit.
Just be careful to not short the 5v to any other pin with your probe!

User avatar
davidcoton
Posts: 4137
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: The Big (little) Yellow Bolt of Doom

Sat Feb 16, 2019 11:13 am

Demisi wrote:
Sat Feb 16, 2019 2:07 am
. i found that if i bump the voltage up to 5.8 Volts on the PSU the power issues clear up and i get a reading of about 5.2 to 5.3 volts at the USB plug.
Don't do that. When you shut down, the load (current) will decrease, so the voltage drop will decrease, so the voltage at the Pi will increase, and 5V8 is well outside the maximum voltage for a Pi.
Tackle the real problem, arrange the power wiring so that the voltage drop is minimised (thick, short wires by the most direct route possible). Run the PSU at no more than 5V25.
Or the alternative answer, suggested several times already (even if naively): get an official Pi PSU.
Signature retired

Demisi
Posts: 6
Joined: Thu Feb 14, 2019 10:29 pm

Re: The Big (little) Yellow Bolt of Doom

Sat Feb 16, 2019 1:38 pm

Ok thanks again davidcoton ill try making a new cable with shorter but much thicker wires if that dose not work then off to buy an official power supply. Also thanks burngate I'll set up some jumper wires and start checking at the board.

Idahowalker
Posts: 445
Joined: Wed Jan 03, 2018 5:43 pm

Re: The Big (little) Yellow Bolt of Doom

Sat Feb 16, 2019 3:22 pm

Demisi wrote:
Sat Feb 16, 2019 2:07 am
First off thank you to Imperf3kt. I am here to learn what i did wrong or how i screwed something up not where i should spend my money so if you have advice i will always listen, it was helpful. Next another thanks to Ernst and Idahowalker, i have never heard that about breadboards before and it ultimately lead to the solution i found. basically i had a drop from the PSU connecting lead to the breadboard, it was minor 5.1 volts at PSU to about 4.95 volts going into the breadboard. then across the breadboard to the USB connector there was quite a significant drop to about 4.63 volts, then on to the USB cable. To those who suggested the breadboard I am using a Velleman SD35N. i found that if i bump the voltage up to 5.8 Volts on the PSU the power issues clear up and i get a reading of about 5.2 to 5.3 volts at the USB plug.
You are welcome. You might want to put in a fuse and a crowbar circuit with the gate sensing as close as to the RPi as you can get.
Without knowing why you are deleting my postings, I will not know how...

User avatar
HawaiianPi
Posts: 4597
Joined: Mon Apr 08, 2013 4:53 am
Location: Aloha, Oregon USA

Re: The Big (little) Yellow Bolt of Doom

Sun Feb 17, 2019 2:04 am

Demisi wrote:
Sat Feb 16, 2019 2:07 am
i found that if i bump the voltage up to 5.8 Volts on the PSU the power issues clear up and i get a reading of about 5.2 to 5.3 volts at the USB plug.
That is clearly a power delivery problem, and increasing the voltage is the wrong way to solve it. The Pi does not have an on-board 5V regulator, so feeding more than the recommended maximum of 5.25V risks damage, and you can't count on resistance of the power wires to reduce voltage to a safe level, because voltage varies with load.

This is the reason we keep recommending the official PSU. Trying to get an unknown and untested source working is not only a hassle, it risks damaging the Pi.

If you want to work with what you have, you need a way to deliver the power with minimal resistance (dual power and ground wires connected to the GPIO expansion header), and you need to properly load test the PSU to see if it can actually meet its specs. You'd be surprised how many don't (I have a 5V/2.4A charger that drops well below 5V at only a 1A load). Many so-called 5V "power supplies" are re-purposed phone chargers that don't maintain a stable 5V under load (phones don't care about that, computers do).

Note that measuring the voltage with a typical multi-meter won't show transient voltage drops (you need a scope to see voltage transients), but the Pi will see them and activate the low voltage warning. Also note that the low voltage warning indicates less than 4.63 volts, while some USB devices need at least 4.75V to work properly. So it's possible to have power problems without a low voltage warning.
My mind is like a browser. 27 tabs are open, 9 aren't responding,
lots of pop-ups...and where is that annoying music coming from?

Demisi
Posts: 6
Joined: Thu Feb 14, 2019 10:29 pm

Re: The Big (little) Yellow Bolt of Doom

Sun Feb 17, 2019 4:03 am

Thanks Hawaiianpi dose the power warning kick in only if the power supply drops below 4.63 volts or will draw from USB devices kick it in to? What you said sounds like the USB would be isolated and that didnt make sense to me. And yes I started seeing a power warning again while stress testing at about a 1.6 to 1.7 amp draw. That is after I abandoned the breadboard and connected the psu directly to vcc and ground on the USB breakout board and dropped the voltage to 5.2 volts.

User avatar
HawaiianPi
Posts: 4597
Joined: Mon Apr 08, 2013 4:53 am
Location: Aloha, Oregon USA

Re: The Big (little) Yellow Bolt of Doom

Sun Feb 17, 2019 7:21 am

USB devices add to the system load and can cause a low voltage warning if the PSU can't keep up, or you try to pull more than 1.2A from the USB ports (total for all 4 ports, with no per-port limits).

The USB ports are controlled by the USB+Ethernet hub chip, which is behind a poly-fuse when power is supplied by the micro USB jack. So they output the input voltage minus some small circuit losses.

The GPIO expansion header pins 2 & 4 are connected directly to the 5V rail and should equal the input voltage. If it's significantly less, especially under load, then the wires connecting the power have too much resistance. You could try powering your Pi from the GPIO expansion header using a pair of wires to both +5V pins, and another pair to any two ground pins. Just be aware that this bypasses the poly-fuse protection.
pi3_gpio.jpg
pi3_gpio.jpg (88.86 KiB) Viewed 971 times
My mind is like a browser. 27 tabs are open, 9 aren't responding,
lots of pop-ups...and where is that annoying music coming from?

drgeoff
Posts: 9804
Joined: Wed Jan 25, 2012 6:39 pm

Re: The Big (little) Yellow Bolt of Doom

Sun Feb 17, 2019 11:53 am

HawaiianPi wrote:
Sun Feb 17, 2019 7:21 am
The GPIO expansion header pins 2 & 4 are connected directly to the 5V rail and should equal the input voltage.
The polyfuse is between the micro-USB power input socket and everything. The voltage at the GPIO header pins 2 & 4 will be slightly less than the input voltage.

Return to “Troubleshooting”