Power Issue Hell.

[FrustrationPi]

Hi there,

I've bough 2 Pi's, 3 usb hubs, 2 sd cards, 2 keyboards, 2 power supplies and have tried a bunch of stuff I had laying around.

I can't get the pi to have more than 4.2 volts with a usb item plugged in.

I'm ready to throw this thing out the window to be honest, so frustrated am I. Every time I've hoped a new purchase would fix my problems... I'd boot the pi and either have problems right away or it would lock up or do weird stuff as soon as I tried to dabble with some python.

I'm at my wits' end. Is there anything you could recommend? Anything I could try? I've googled, I've gotten out the multimeter, I've tried everything except ripping off the micro sd connector and soldering the wired of the psu right on the board.

HELP!!!!

[Dweeber]

3 usb hubs, 2 sd cards, 2 keyboards, 2 power supplies.

And the specific items you got for each of those were? Hard to provide any feedback when you don't provide specifics.

[mahjongg]

At one point you probably blew the power input polyfuse (F3), that can happen if, for example, you hot plugged a badly regulated power supply into the PI, and its overvoltage protection diode tripped.

The polyfuse will only recover very slowly, and only if you leave it be for a few days at least.

With a bad polyfuse, unlike a regular fuse, some power will still reach the PI (enough to light the power LED), but not enough for the PI to work. Thats why you see only 4.2 volt (measured between TP1 and TP2 on the PI) if you measure the voltage across the polyfuse, you will find the missing 0.8V there.

if you temporarily short the fuse with pair of tweezers, or just with a bent paperclip, then you will find that your PI will probably boot (if you haven't corrupted the image on the SD card).

I suggest you read the troubleshooting page of the wiki, here:
http://elinux.org/R-Pi_Troubleshooting

[FrustrationPi]

Thank you for taking the time to answer. I have gone through the trouble shooting and I eventually bought peripherals from the compatibility list. So when it comes to that I am officially 'all good', and yet both my pi's are very very wonky.

I just dabbled with trying different sd cards as I read that some use more power than others. But I see no difference in power when I exchange the cards and the one I use is from the compatibility list.

[mahjongg]

Just try to find out why your PI isn't getting the 4.9V it needs (well in any case the USB devices need).
IMHO there are just two reasons, (if your PSU is able to deliver 5V @ 700mA)
1) Your polyfuse has gone bad
or
2) There is power loss in the cable from the PSU to the PI

Obviously its also possible, but unlikely, that the PI is broken and draws so much current that the PSU cannot cope.

The wiki is a community effort, and what it contains isn't gospel, it as good as what people put into it. Most often the info is reliable, but some people just test a PSU and then declare its "working", there is no guarantee it is well regulated.

[tufty]

First thing to do is measure the voltage drop across the polyfuse (f3, on the other side of the board to the micro-usb connector). Measure the voltages between gnd (usb connector casing is a good one to use) and either side of the fuse, with the pi powered up and no usb devices or sd card inserted, post them up. Also, post up the specs of your psu(s).

Simon

[FrustrationPi]

Measuring from f03 to the housing of the micro usb gives me practically nothing. 0.01v - so either I'm doing that wrong or there's nothing going on there.

[obcd]

On one site of the polyfuse you should measure the input voltage of your supply, approx. 5V
On the other site, you should measure the TP1 - TP2 voltage as one of those testpoints is gnd and the other one is connected to that polyfuse.
Maybe there is an isolation barrier on the solder so that your probes don't give good contact to them.
Or maybe the shielding you used isn't connected to gnd. Take TP1 or TP2 instead as gnd point. (Not sure which one, but you can find that on the schematics.)

[johnbeetem]

To expand on mahjongg's comment, one possible culprit is the Micro USB cable. Some Micro USB cables have very thin conductors and result in a voltage drop high enough to make RasPi fail. Measure the voltage on both sides on F3 with voltmeter ground connected to TP2. If they're approximately the same (4.2V at TP1, <4.3V at the other side of F3) then the voltage drop is probably across your Micro USB cable. (This assumes your power supply is putting out a good 5V +/- 5%, which is probably the case since you tried multiple supplies.) If the F3 voltages differ by more than that, then F3 is bad or has blown and needs to heal itself.

[FrustrationPi]

Both sides of F03 to TP1 give me the same as TP1 to TP2.

The micro usb cable is the 1 thing I have never replaced. I bought one just for the Pi and when I looked in to replacing it all the available cables seemed equally flimsy.

[mahjongg]

There are two ways to measure the voltage across the polyfuse, the most obvious way is to put one probe of the voltmeter on one end of the polyfuse, and the other probe on the other end (with power on, obviously) then you will measure the voltage directly, if its bigger than 0.1 Volt the polyfuse is bad, for example 0.7V is bad.

The other way, which some people, like johnbeetem, seem to prefer, (for reasons that i'm not aware of), is to connect the black lead to GND, that is to the metal shell of either the USB, HDMI or ethernet port (or TP2), then put the red lead on one end of the polyfuse, and write down what you measure, then on the other end and repeat. If the voltage on one end is say 4.9V and on the other end 4.2V the voltage over the polyfuse is 4.9 - 4.2 = 0.7 volt and that is bad, it should be less than 0.1 Volt.

If both sides give 4.2 Volt, then the loss isn't in the polyfuse, logically it should be in the cable, but you may try measuring the voltage over the polyfuse directly to check.

[FrustrationPi]

If there was 0.7V over the polyfuse would the Pi boot?

I'm no electronics expert but it seems signs are pointing towards a 'bad' cable. Since I haven't tried to replace it yet but have replaced everything else several times over it might be the simplest for me to order a new one. I don't see how I'll be guaranteed to get a good one though, as I said.. they all look equally flimsy.

Is there a conclusion on the Pi power issues? Is this a design flaw? Will future Pi's have a different psu solution?

[davidmam]

The apparent strength of the cable is down to the plastic, not the wire. I bought some cheap (3 for £2) micro USB cables and I might as well have burned the cash. They cause a 0.4 V drop which is enough to screw the whole thing up. Fine for data, no good for power.

It also caused me much frustration as I too learned that not all cables are created equal.

[mahjongg]

copper is expensive relative to plastic, so some chinese trash makers use hair thin copper wires, or worse even hair thin aluminium wires, just to save 0.1 cent.

[obcd]

It's a design flaw.
The universal usb chargers that the Pi is supposed to work with should output a voltage between 4.85V and 5.25V between no load and maximum load conditions. (European recommended regulations)
4.85V is the minimum voltage you need for a stable Pi operation.
There is no such thing as an ohmless cable, so there will always be an additional small voltage drop over the cable. The same counts for the input polyfuse.
While it's known that chinese suppliers don't care much about the specs, even with supplies following the specs you can run into problems.
The alpha boards had a step down power supply on the board. Their input voltage was something like 7 - 12V. It was left out to cut the price.
Besides that, the Pi should be low cost since you can reuse the devices you have lying around. For that reason, the 5V power adapter was choosen. They only started to use those adapters for cellphones less than a year ago. So, I don't think much people have one lying around. You can use the one of your cellphone if you are lucky, but besides for first time experimenting, this is not very practical.

I know people will argue, and some will say they had the Pi running down to 4.3V. Apparantly usb doesn't work at those voltages, but to prove the Pi isn't having issues with lower voltages, everything is allowed. I had 3 identical power adapters. One had a voltage that was 0.02V lower than the other. The TP1 - TP2 voltage was approx. 4.83V. The one with the lower voltage caused interference on the monitor. It went away by boosting the hdmi signal in the config.txt file. You can blame the monitor..It keeps this Forums alive and kicking...But others might use the same monitor...So, I am not going to tell them everything is fine if they only measure 4.80V between TP1 and TP2. And I will also not tell that the Pi works stable down to 4.75V It's simply not always the case.
Don't get me wrong, this is no complain or whining. It's just a conclusion I think people should be aware about.

[tufty]

You're not the only one to think that the micro-usb PSU thing was a bad idea from the get-go. However, going over that old ground again doesn't help the OP.

So, FrustratedPi. Sounds a lot like you have USB cable woes. Just for completeness, though, could you post the actual measurements you made TP1-TP2 and the two GND -> F3? And, for kicks, what's marked on your PSU.

Simon

[jamesh]

It's a design flaw.
The universal usb chargers that the Pi is supposed to work with should output a voltage between 4.85V and 5.25V between no load and maximum load conditions. (European recommended regulations)
4.85V is the minimum voltage you need for a stable Pi operation.
There is no such thing as an ohmless cable, so there will always be an additional small voltage drop over the cable. The same counts for the input polyfuse.
While it's known that chinese suppliers don't care much about the specs, even with supplies following the specs you can run into problems.
The alpha boards had a step down power supply on the board. Their input voltage was something like 7 - 12V. It was left out to cut the price.
Besides that, the Pi should be low cost since you can reuse the devices you have lying around. For that reason, the 5V power adapter was choosen. They only started to use those adapters for cellphones less than a year ago. So, I don't think much people have one lying around. You can use the one of your cellphone if you are lucky, but besides for first time experimenting, this is not very practical.

I know people will argue, and some will say they had the Pi running down to 4.3V. Apparantly usb doesn't work at those voltages, but to prove the Pi isn't having issues with lower voltages, everything is allowed. I had 3 identical power adapters. One had a voltage that was 0.02V lower than the other. The TP1 - TP2 voltage was approx. 4.83V. The one with the lower voltage caused interference on the monitor. It went away by boosting the hdmi signal in the config.txt file. You can blame the monitor..It keeps this Forums alive and kicking...But others might use the same monitor...So, I am not going to tell them everything is fine if they only measure 4.80V between TP1 and TP2. And I will also not tell that the Pi works stable down to 4.75V It's simply not always the case.
Don't get me wrong, this is no complain or whining. It's just a conclusion I think people should be aware about.

Actually, the uUSB charger was first accepted in 2007, and ratified in 2009 - in Europe at least, so chargers have been around since 2007. I noticed the other day that I have more uUSB cables than mini's, and had to really dig to find a B connector cable!