Testing 5V USB supplies for RPi usage
I have several 5V USB supplies, which if the labels on them are to be believed will supply anything from 1 to 2.5A, although I have serious doubts that I can believe the labels. Based on how sowly they charge my phone I think some that are labelled 2.5A are more like 200mA... They're from various sources, which include the likes of eBay. Some will power my RPi, but only if I don't try to plug in a USB device as well. Assuming very little in the way of electrical knowledge (which I guess is obvious as I don't know the answer to my question), what's the "best" way to see what current a power supply can deliver? I have a multimeter, and could doubtless obtain some resisters (or other components) if I need to. I also have a shiny new temperature controlled soldering iron that's never been used if that helps :-) -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) @ 13 Clarke Rd, Milton Keynes, MK1 1LG
On 25/01/13 17:41, Mark Rogers wrote:
I have several 5V USB supplies, which if the labels on them are to be believed will supply anything from 1 to 2.5A, although I have serious doubts that I can believe the labels. Based on how sowly they charge my phone I think some that are labelled 2.5A are more like 200mA... They're from various sources, which include the likes of eBay. Some will power my RPi, but only if I don't try to plug in a USB device as well.
Assuming very little in the way of electrical knowledge (which I guess is obvious as I don't know the answer to my question), what's the "best" way to see what current a power supply can deliver?
Read the spec. It's the voltage which determines the speed of charging, not the rated amps - but if the amperage at the required voltage isn't sufficient to cope, either the delivered voltage will drop, or the supply will lose all its magic smoke.
I have a multimeter, and could doubtless obtain some resisters (or other components) if I need to. I also have a shiny new temperature controlled soldering iron that's never been used if that helps :-)
Not really. Just choose the supply which delivers enough and then a bit. You *could* look for a circuit diagram online, visit somewhere like Maplin or go online again to (say) Farnellto get some Veraboard and components, and build yourself a better PSU - that's if you want to use your soldering iron... Really - it's not that difficult. -- Tony
On 25/01/13 18:06, Anthony Anson wrote:
On 25/01/13 17:41, Mark Rogers wrote:
I have several 5V USB supplies, which if the labels on them are to be believed will supply anything from 1 to 2.5A, although I have serious doubts that I can believe the labels. Based on how sowly they charge my phone I think some that are labelled 2.5A are more like 200mA... They're from various sources, which include the likes of eBay. Some will power my RPi, but only if I don't try to plug in a USB device as well.
Assuming very little in the way of electrical knowledge (which I guess is obvious as I don't know the answer to my question), what's the "best" way to see what current a power supply can deliver?
Further, and which might save you a lot of time and money, use your multimeter to check the voltage of the supply (across the + and - lines) before and during charging. If the voltage drops below what is required, *that's* the prompt to raise the output capability. It's the impedance/resistance of the device being supplied which determines the amperage, not the rated output of the PSU. -- Tony
On 25 January 2013 18:06, Anthony Anson <tony.anson@girolle.co.uk> wrote:
Read the spec.
That only works if you believe the spec. The problem I have now is that I have a mixture of chargers, most of them "branded", some of which it is become clear are likely fakes, but without a way to know which are which I don't know which to trust. (Eg I have a mixture of Sony Ericsson chargers, a least one of which is genuine, but if any are fakes I can't tell them apart.)
It's the voltage which determines the speed of charging, not the rated amps - but if the amperage at the required voltage isn't sufficient to cope, either the delivered voltage will drop, or the supply will lose all its magic smoke.
So puting it another way: if a PSU says it can deliver 1A @5V, then I should be able to stick a 5ohm (R=V/I) resister across the terminals and the voltage shouldn't drop substantially below 5V? And provided I pick a suitable 5ohm resistor (5W? P=VI?) then I shouldn't get any loud bangs, burning smells, or tripped power?
Not really. Just choose the supply which delivers enough and then a bit.
The problem is I have picked supplies which *claim to* deliver enough but I don't believe them. Not least I'd like to prove they don't and get me money back! -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) @ 13 Clarke Rd, Milton Keynes, MK1 1LG
On 25/01/13 18:38, Mark Rogers wrote:
On 25 January 2013 18:06, Anthony Anson <tony.anson@girolle.co.uk> wrote:
Read the spec.
That only works if you believe the spec.
The problem I have now is that I have a mixture of chargers, most of them "branded", some of which it is become clear are likely fakes, but without a way to know which are which I don't know which to trust. (Eg I have a mixture of Sony Ericsson chargers, a least one of which is genuine, but if any are fakes I can't tell them apart.)
Plug them in and test the voltage across the (low voltage) supply. That will tell you which ones cut the mustard.
It's the voltage which determines the speed of charging, not the rated amps - but if the amperage at the required voltage isn't sufficient to cope, either the delivered voltage will drop, or the supply will lose all its magic smoke.
So puting it another way: if a PSU says it can deliver 1A @5V, then I should be able to stick a 5ohm (R=V/I) resister across the terminals and the voltage shouldn't drop substantially below 5V?
Theoretically. Remember that the resistance of a component is often temperature sensitive, and a resistor's value is probably measured at expected working temperature. While it's unlikely to vary a lot over the range, if the PSU is working at its limit, it could be put at risk. Better to measure the supplied potential difference with the device connected. If it's within sensible limits (the spec of the device should tell you that if you dig deeply enough), that's good enough. If the measured voltage is too low, you risk magic smoke from the PSU. If it's too high,look for the smoke to come from your device as it chars.
And provided I pick a suitable 5ohm resistor (5W? P=VI?) then I shouldn't get any loud bangs, burning smells, or tripped power?
There is a difference between resistance and impedance - not always numericlly. I'd forget about using a resistor.
Not really. Just choose the supply which delivers enough and then a bit.
The problem is I have picked supplies which *claim to* deliver enough but I don't believe them. Not least I'd like to prove they don't and get me money back!
If they don't deliver enough power the potential difference between the + and - lines will be below the stated voltage figure. This will not harm your device, but may destroy your charger - given time. -- Tony
On 25 January 2013 19:16, Anthony Anson <tony.anson@girolle.co.uk> wrote:
If they don't deliver enough power the potential difference between the + and - lines will be below the stated voltage figure. This will not harm your device, but may destroy your charger - given time.
Thanks. I'll cut an old Micro-USB cable up and take some voltage readings while charging a phone and see how they compare. -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) @ 13 Clarke Rd, Milton Keynes, MK1 1LG
On 25/01/13 20:49, Mark Rogers wrote:
On 25 January 2013 19:16, Anthony Anson <tony.anson@girolle.co.uk> wrote:
If they don't deliver enough power the potential difference between the + and - lines will be below the stated voltage figure. This will not harm your device, but may destroy your charger - given time.
Thanks. I'll cut an old Micro-USB cable up and take some voltage readings while charging a phone and see how they compare.
I don't expect you'll find any PSU will be too weak - remember, 5 watts is about the power used by the front bulb of a bicycle, or IIRC, some motor vehicle rearlight bulbs. A Raspberry Pi (I would guess) would use only a fraction of that. -- Tony
(Sorry if this appears twice.) When I needed to test 5volt supplies I made a simple load resistor out of 21W 12v car bulbs. putting 1 across the 5v supply should draw 0.729A 2 in parallel should draw 1.458A 3 in parallel should draw 2.188A 4 ... 2.92A etc. As long as I've got my sums right ! As long as the psu is able to supply that amount of current then the voltage should still measure 5V. If not the psu is duff. Obviously chose the load to match the current rating of the psu you're testing. Nev On 25/01/13 20:49, Mark Rogers wrote:
On 25 January 2013 19:16, Anthony Anson <tony.anson@girolle.co.uk> wrote:
If they don't deliver enough power the potential difference between the + and - lines will be below the stated voltage figure. This will not harm your device, but may destroy your charger - given time. Thanks. I'll cut an old Micro-USB cable up and take some voltage readings while charging a phone and see how they compare.
On Fri, Jan 25, 2013 at 09:21:05PM +0000, nev young wrote:
(Sorry if this appears twice.)
When I needed to test 5volt supplies I made a simple load resistor out of 21W 12v car bulbs. putting 1 across the 5v supply should draw 0.729A 2 in parallel should draw 1.458A 3 in parallel should draw 2.188A 4 ... 2.92A etc.
As long as I've got my sums right !
Ah, but lamp bulbs *do* change their resistance significantly from cold to hot/alight. -- Chris Green
On 25/01/13 20:49, Mark Rogers wrote:
On 25 January 2013 19:16, Anthony Anson <tony.anson@girolle.co.uk> wrote:
If they don't deliver enough power the potential difference between the + and - lines will be below the stated voltage figure. This will not harm your device, but may destroy your charger - given time.
Thanks. I'll cut an old Micro-USB cable up and take some voltage readings while charging a phone and see how they compare.
Have a look at www.maplin.co.uk/ and see if (offers) N66JU; A92JH; A83LA float your boat. -- Tony
On 27 January 2013 01:04, Anthony Anson <tony.anson@girolle.co.uk> wrote:
Have a look at www.maplin.co.uk/ and see if (offers) N66JU; A92JH; A83LA float your boat.
Just to make it clear (because I don't think I did from the outset), it's not that I'm having a problem sourcing supplies that will keep the RPi going - I have some that do and some that don't, it's that the current method of testing (plug it in and see whether it crashes) is a bit inexact, and the correct method (read the label to see what current it's rated at) isn't working because the labels can't be trusted on some of them. Also, there's the option of complaining and getting a refund on ones that are no good, and having a measurement to back me up always helps. And finally, perhaps more importantly for me, its the opportunity to learn how to get the measurement. -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) @ 13 Clarke Rd, Milton Keynes, MK1 1LG
On 28/01/13 11:53, Mark Rogers wrote:
On 27 January 2013 01:04, Anthony Anson <tony.anson@girolle.co.uk> wrote:
Have a look at www.maplin.co.uk/ and see if (offers) N66JU; A92JH; A83LA float your boat.
Just to make it clear (because I don't think I did from the outset), it's not that I'm having a problem sourcing supplies that will keep the RPi going - I have some that do and some that don't, it's that the current method of testing (plug it in and see whether it crashes) is a bit inexact, and the correct method (read the label to see what current it's rated at) isn't working because the labels can't be trusted on some of them.
Also, there's the option of complaining and getting a refund on ones that are no good, and having a measurement to back me up always helps. And finally, perhaps more importantly for me, its the opportunity to learn how to get the measurement.
Somebody has had similar thoughts to you. See :- http://www.raspberrypi.org/phpBB3/viewtopic.php?f=9&t=5311
On 28 January 2013 13:20, Chris Walker <cdw_alug@the-walker-household.co.uk> wrote:
Somebody has had similar thoughts to you. See :- http://www.raspberrypi.org/phpBB3/viewtopic.php?f=9&t=5311
Thanks for this, I'm going to knock together something fairly similar and see how I get on. -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0844 251 1450 Registered in England (0456 0902) @ 13 Clarke Rd, Milton Keynes, MK1 1LG
On 28/01/13 11:53, Mark Rogers wrote:
On 27 January 2013 01:04, Anthony Anson <tony.anson@girolle.co.uk> wrote:
Have a look at www.maplin.co.uk/ and see if (offers) N66JU; A92JH; A83LA float your boat.
Just to make it clear (because I don't think I did from the outset), it's not that I'm having a problem sourcing supplies that will keep the RPi going - I have some that do and some that don't, it's that the current method of testing (plug it in and see whether it crashes) is a bit inexact,
It does have points to commend it...
and the correct method (read the label to see what current it's rated at) isn't working because the labels can't be trusted on some of them.
Also, there's the option of complaining and getting a refund on ones that are no good, and having a measurement to back me up always helps. And finally, perhaps more importantly for me, its the opportunity to learn how to get the measurement.
be sure, when you have found how, that your multimeter is reasonably accurate. -- Tony
On Fri, Jan 25, 2013 at 07:16:40PM +0000, Anthony Anson wrote:
So puting it another way: if a PSU says it can deliver 1A @5V, then I should be able to stick a 5ohm (R=V/I) resister across the terminals and the voltage shouldn't drop substantially below 5V?
Theoretically. Remember that the resistance of a component is often temperature sensitive, and a resistor's value is probably measured at expected working temperature. While it's unlikely to vary a lot over the range, if the PSU is working at its limit, it could be put at risk.
Any 'purpose made' resistor will have (for the case in point) a negligable temperature coefficient. It's only things that *aren't* 'resistors' that will change value significantly with temperature. Even a crude bit of 'resistance wire' (as in constantan, I think that's one of the names) will have a temperature coefficient of a tiny fraction of a percent per degree and will maintain essentially the same resistance up to red heat. -- Chris Green
On 26/01/13 10:48, Chris Green wrote:
On Fri, Jan 25, 2013 at 07:16:40PM +0000, Anthony Anson wrote:
So puting it another way: if a PSU says it can deliver 1A @5V, then I should be able to stick a 5ohm (R=V/I) resister across the terminals and the voltage shouldn't drop substantially below 5V?
Theoretically. Remember that the resistance of a component is often temperature sensitive, and a resistor's value is probably measured at expected working temperature. While it's unlikely to vary a lot over the range, if the PSU is working at its limit, it could be put at risk.
Any 'purpose made' resistor will have (for the case in point) a negligable temperature coefficient. It's only things that *aren't* 'resistors' that will change value significantly with temperature.
Even a crude bit of 'resistance wire' (as in constantan, I think that's one of the names) will have a temperature coefficient of a tiny fraction of a percent per degree and will maintain essentially the same resistance up to red heat.
Try testing a wire-wound resistor... -- Tony
On Sat, Jan 26, 2013 at 12:37:56PM +0000, Anthony Anson wrote:
On 26/01/13 10:48, Chris Green wrote:
On Fri, Jan 25, 2013 at 07:16:40PM +0000, Anthony Anson wrote:
So puting it another way: if a PSU says it can deliver 1A @5V, then I should be able to stick a 5ohm (R=V/I) resister across the terminals and the voltage shouldn't drop substantially below 5V?
Theoretically. Remember that the resistance of a component is often temperature sensitive, and a resistor's value is probably measured at expected working temperature. While it's unlikely to vary a lot over the range, if the PSU is working at its limit, it could be put at risk.
Any 'purpose made' resistor will have (for the case in point) a negligable temperature coefficient. It's only things that *aren't* 'resistors' that will change value significantly with temperature.
Even a crude bit of 'resistance wire' (as in constantan, I think that's one of the names) will have a temperature coefficient of a tiny fraction of a percent per degree and will maintain essentially the same resistance up to red heat.
Try testing a wire-wound resistor...
... I'm sure you'll find the resistance doesn't vary by any amount you'd be able to notice. E.g. take a look at:- http://www.rapidonline.com/Electronic-Components/5W-Axial-ceramic-resistors-... the temperature coefficient is 300ppm per degree centigrade so even if you take it to 200 degrees (starting near zero) that would change the resistance by 200*300/1000000 which is 60000/1000000, 6% is it? Not very significant in the sorts of measurements we're talking about and you shouldn't really get them that hot anyway! -- Chris Green
participants (5)
-
Anthony Anson -
Chris Green -
Chris Walker -
Mark Rogers -
nev young