What if the Power Input is not exactly 5v?


16 posts
by gallomimia » Sat Mar 03, 2012 5:09 am
What happens if this is true? I'm sure this on here somewhere but I couldn't find it.

I've heard it can be powered from AA batteries. If this is true there must be some amount of tolerance on the voltage range of the input. Since batteries inherently develop an internal resistance once they become partially discharged, this causes the voltage across the battery to drop depending upon how much current is drawn from it. Powered by batteries, a processor would draw a varying amount of current depending on how much use it is seeing; how many of its transistors are changing state in a given cycle, averaged over time.

So, lets' propose that I have connected four AA batteries to my pi board. It starts a voltage of 6 and drops with use, possibly as low as 4 before it might become totally unusable. And in the case of rechargeable cells it would start at 5V and drop from there. So I might even consider adding a 5th battery in this case.

So I ask formally, for an authoritative answer only, what is the range of input power which might be acceptable.

Also, in the case of USB connected devices which exceed the wattage of the power circuitry on the board, how might I add a powered hub to the battery arrangement? (The last powered hub I scrounged up backfed its power onto the host computer, very bad, but I cannot find anything better)
Posts: 3
Joined: Sat Mar 03, 2012 4:59 am
by jbeale » Sat Mar 03, 2012 5:33 am
 

 

Here's a relevant comment from Pete Lomas on another forum, he was involved with the RPi hardware design. Looks like you want 5.0V with +/- 5% accuracy. From http://www.element14.com/commu.....8;tstart=0


Re: What's Your Pi Plan??


It should be possible to power the RPi from a solar cell - BUT you really need to regulate the voltage down to between 4.75 and 5.25 volts - plenty of LDO Regulators on element14 that would do that. Something with an output accuracy of 5% or better.  Ideally you want some form of power storage otherwise passing clouds could give you a brownout!

The 5V on the USB micro USB port goes straight into the processor - so 6V could be uncomfortable - we don't want you to toast it!  There is a resetable fuse and a 5V Transil but importantly there is no reverse voltage protection diode as such - so please take care when wiring up.

You can also feed power to the board through the GPIO connector (P1) - Pins 2 & 4 are wired to +5.

User avatar
Posts: 2076
Joined: Tue Nov 22, 2011 11:51 pm
by SN » Tue Mar 06, 2012 9:50 pm
perhaps a 9v battery via a 5v regulator is the answer - I am keen to find a battery powered solution but would need to understand what the powerdraw means for battery life - possibly some solution where there is solar charging into one battery whilst another is powering the raspi and it is able to tell when it needs to be swapped from one cell to the other

just a thought...

reminds me of mates dad who built a sensor circuit for his battery powered Acorn Atom circa 1981 so that the Atom was able to warn that it was, in the words of the Gauntlet Arcade Game, "......you are about to die......"
Steve N – binatone mk4->intellivision->zx81->spectrum->cbm64->cpc6128->520stfm->pc->raspi ?
User avatar
Posts: 1008
Joined: Mon Feb 13, 2012 8:06 pm
Location: Romiley, UK
by drgeoff » Tue Mar 06, 2012 10:05 pm
There is a good chance that the RP can be powered directly from 4 rechargeable NiMH cells in series.  I'd rate the chance with alkalines somewhat lower.

For best results, a higher voltage feeding a switching regulator is the way to go.  I have already ordered and received two of these http://www.ebay.co.uk/itm/3306.....9.l2649.  Now waiting for an RP to use one with  :-)
Posts: 2924
Joined: Wed Jan 25, 2012 6:39 pm
by nullstring » Tue Mar 06, 2012 10:29 pm
I am actually curious about what will happen when we feed it extra voltage.

Hopefully a protection circuit will stop it getting fried, right?
Posts: 178
Joined: Sun Oct 02, 2011 3:05 pm
by drgeoff » Tue Mar 06, 2012 10:39 pm
nullstring said:


I am actually curious about what will happen when we feed it extra voltage.

Hopefully a protection circuit will stop it getting fried, right?


There is a regulator between the 5 volt input and the Broadcom SoC plus PoP RAM combination.  The 5 volt input also goes to the USB sockets and hence to things plugged into them. I'm not sure which supply the USB/LAN chip is powered from - the 5 volts may go in there too.

If you'd like to do some experiments and report back your results of what goes bang at what voltage, I'm sure many people here would appreciate it. :-)
Posts: 2924
Joined: Wed Jan 25, 2012 6:39 pm
by Gert van Loo » Tue Mar 06, 2012 10:40 pm
Up to a point. There is an over-voltage diode but if your supply does not shut down, but instead puts enough energy through it, it will eventually blow up. Shortly after that you will have a baked Pi.
User avatar
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 2077
Joined: Tue Aug 02, 2011 7:27 am
by nullstring » Tue Mar 06, 2012 11:09 pm
Gert said:


Up to a point. There is an over-voltage diode but if your supply does not shut down, but instead puts enough energy through it, it will eventually blow up. Shortly after that you will have a baked Pi.



So, if the diode blows up then it will be shorted out?

I am less worried about the RPI than a $2000 TV getting fried by an overvolted HDMI port.

There is no way to make it instead disconnect the circuit?
Posts: 178
Joined: Sun Oct 02, 2011 3:05 pm
by SN » Tue Mar 06, 2012 11:42 pm
drgeoff said:


There is a good chance that the RP can be powered directly from 4 rechargeable NiMH cells in series.  I'd rate the chance with alkalines somewhat lower.

For best results, a higher voltage feeding a switching regulator is the way to go.  I have already ordered and received two of these http://www.ebay.co.uk/itm/3306.....9.l2649.  Now waiting for an RP to use one with  :-)


Uh oh more expense - they look right for me though - I can throw 12v straight out of a std cigaretter lighter socket at them and they guarantee to step me down to 5v
Steve N – binatone mk4->intellivision->zx81->spectrum->cbm64->cpc6128->520stfm->pc->raspi ?
User avatar
Posts: 1008
Joined: Mon Feb 13, 2012 8:06 pm
Location: Romiley, UK
by johnbeetem » Tue Mar 06, 2012 11:57 pm
nullstring said:


Gert said:


Up to a point. There is an over-voltage diode but if your supply does not shut down, but instead puts enough energy through it, it will eventually blow up. Shortly after that you will have a baked Pi.


So, if the diode blows up then it will be shorted out?

I am less worried about the RPI than a $2000 TV getting fried by an overvolted HDMI port.

There is no way to make it instead disconnect the circuit?


Here's my understanding from the preliminary version of the power supply schematic.  (There's a link to this schematic at the RasPi wiki: http://elinux.org/RPi_Hardware ):

If the supply voltage goes over 5V, zener diode D17 quickly clamps it to 5V and if the resulting current exceeds 1A then self-resetting fuse F3 opens the circuit.  So there's a reasonable chance the RasPi and attached 5V equipment will survive.  However, my understanding is that zener diodes are not very precise over a wide temperature range and fuses take a while to blow so I wouldn't like to depend on this circuit for everyday use.

You are much better off using a switching power supply to convert 6V or higher battery voltage to precision 5V.
User avatar
Posts: 942
Joined: Mon Oct 17, 2011 11:18 pm
Location: The Coast
by Gert van Loo » Wed Mar 07, 2012 12:16 am
nullstring said:


Gert said:


Up to a point. There is an over-voltage diode but if your supply does not shut down, but instead puts enough energy through it, it will eventually blow up. Shortly after that you will have a baked Pi.


So, if the diode blows up then it will be shorted out?

I am less worried about the RPI than a $2000 TV getting fried by an overvolted HDMI port.

There is no way to make it instead disconnect the circuit?


You should use a standard micro-USB 5V supply. If you do not do so but instead start connecting non-standard power supplies the consequences are your own!
User avatar
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 2077
Joined: Tue Aug 02, 2011 7:27 am
by drgeoff » Wed Mar 07, 2012 1:11 am
SN said:


drgeoff said:


There is a good chance that the RP can be powered directly from 4 rechargeable NiMH cells in series.  I'd rate the chance with alkalines somewhat lower.

For best results, a higher voltage feeding a switching regulator is the way to go.  I have already ordered and received two of these http://www.ebay.co.uk/itm/3306.....9.l2649.  Now waiting for an RP to use one with  :-)


Uh oh more expense - they look right for me though - I can throw 12v straight out of a std cigaretter lighter socket at them and they guarantee to step me down to 5v


If you are buying these adjustable ones, do not rely on them being preset to 5 volts.  You need a meter and perhaps a dummy load to measure the voltage while you twiddle the potentiometer screw before connecting to your RP.
Posts: 2924
Joined: Wed Jan 25, 2012 6:39 pm
by ato » Fri Mar 09, 2012 4:22 pm
What about this one?

http://www.hi-fun.com/it/acces.....hi-energy/

I just oredered one for other uses, but maybe it  could be useful for Rpi too.

What about battery draining converting 3v to 5v ?
Posts: 2
Joined: Sun Nov 06, 2011 1:16 pm
by drgeoff » Fri Mar 09, 2012 4:34 pm
ato said:


What about this one?

http://www.hi-fun.com/it/acces.....hi-energy/

I just oredered one for other uses, but maybe it  could be useful for Rpi too.

What about battery draining converting 3v to 5v ?


Specs say 550mA output which is too low for a Model B.+
Posts: 2924
Joined: Wed Jan 25, 2012 6:39 pm
by RaTTuS » Fri Mar 09, 2012 5:06 pm
http://www.catb.org/esr/faqs/smart-questions.html <- ask smart Questions
"That's not right, the badgers have moved the goalposts."
1QC43qbL5FySu2Pi51vGqKqxy3UiJgukSX - Prosliver FTW
User avatar
Posts: 5307
Joined: Tue Nov 29, 2011 11:12 am
Location: North West UK
by gallomimia » Sat Aug 11, 2012 1:15 pm
drgeoff wrote:There is a good chance that the RP can be powered directly from 4 rechargeable NiMH cells in series.  I'd rate the chance with alkalines somewhat lower.


The nominal cell voltage of NiMH is 1.2V, times four that gives you 4.8V before internal losses, discharging through use, and cable/polyfuse losses. So the chance that it'll run off of 4 NiMH is slim to none. Maybe if you add some more of these batteries plus a battery eliminator circuit (which makes a constant voltage out of a battery which lowers constantly in voltage as you drain it) it might run the pi.

The nominal cell voltage of Alkaline non-rechargable batteries is 1.5V, times 4 gives 6V. Minus losses will probably put it in the right range, but a battery eliminator circuit is probably a good idea here too.

I'd like to post a concluded answer for my own question based upon the responses here: What happens? You fix it or it blows up your pi. Nuff said. It must be within the given voltage range. It must not drop due to increased load as some cheaper supplies do, or not enough batteries would do. It must not be over a given range or you will explode components. What happens? Just don't do it.
Posts: 3
Joined: Sat Mar 03, 2012 4:59 am