occupied
Posts: 42
Joined: Mon Apr 25, 2016 3:50 pm

Current Consumption Figures

Tue May 10, 2016 8:15 pm

Hello all, I'm new to the forum Image

I ran some checks on the current consumption of various Raspberry Pi models as a precursor to a battery-powered project. I thought I would post the results here in case they are useful to someone else.

Raspberry Pi Models
  • Raspberry Pi 1 Model B
    Image
  • Raspberry Pi 2 Model B
    Image
  • Raspberry Pi 3 Model B
    Image
Multimeter
  • I used a MUKER-TM103:
    Image
  • I ran each test for 30 minutes then doubled the displayed mAh figure e.g. 75mAh after 30 minutes = 150 mA
  • The only exception was the raspivid test because the 8GB card wasn't big enough to store 30 minutes with the full Jessie distro. Therefore the raspivid tests were ran for 20 minutes then tripled.
  • I have no idea how accurate the MUKER-TM103 is. The precision seemed good enough for my needs.
Distros
  • Raspbian Jessie (2016-03-18)
  • Raspbian Jessie Lite (2016-03-18)
  • Diet Pi (v116)
Stock Idle
  • Fresh image on Toshiba Exceria 8GB card
    Image
  • Updates done
Disable HDMI Disable USB Bus
  • Pi 1B

    Code: Select all

    echo 0 | sudo tee /sys/devices/platform/soc/20980000.usb/buspower
  • Pi 2B / Pi 3B

    Code: Select all

    echo 0 | sudo tee /sys/devices/platform/soc/3f980000.usb/buspower
Disable Power LED (Pi 2B)
  • Code: Select all

    echo 0 | sudo tee /sys/class/leds/led1/brightness
Connect Ethernet
  • Cat 6 cable connected to LAN
Connect EW-7811UN (Pi 1B / Pi 2B)
  • Image
  • Connected to LAN
Connect TL-WN823N (Pi 1B / Pi 2B)
  • Image
  • Connected to LAN
Raspivid 1080p30
  • Code: Select all

    raspivid -t 0 -o test.h264
  • Raspberry Pi Camera Module v1
    Image
  • Camera was pointed down the same corridor, no movement, constant lighting
Results
Pi1B.png
Pi1B.png (19.48 KiB) Viewed 5156 times
Pi2B.png
Pi2B.png (19.89 KiB) Viewed 5156 times
Pi3B.png
Pi3B.png (18.41 KiB) Viewed 5156 times

User avatar
MarkHaysHarris777
Posts: 1820
Joined: Mon Mar 23, 2015 7:39 am
Location: Rochester, MN
Contact: Website

Re: Current Consumption Figures

Wed May 11, 2016 4:43 am

occupied wrote: I ran some checks on the current consumption of various Raspberry Pi models as a precursor to a battery-powered project.
I question your results; especially the PI 3B. Did you think of loading the PI (all four cores doing something... )? The 3B easily exceeds 2000ma in normal usage, and often goes as high as 2500ma. In my robot I'm using a 3amp DC-DC 12v -> 5v converter and an 80amphr gel cell. (its an R2D2 look-alike, fairly heavy, and frankly most of its mass is going to be the 12v gel cell).

Anyway, based on your chart above, I think you are going to be way under-estimating your power consumption.
marcus
:ugeek:

tweak42
Posts: 498
Joined: Tue Mar 22, 2016 3:16 am
Location: Los Angeles, California

Re: Current Consumption Figures

Wed May 11, 2016 5:28 am

MarkHaysHarris777 wrote:
occupied wrote: I ran some checks on the current consumption of various Raspberry Pi models as a precursor to a battery-powered project.
I question your results; especially the PI 3B. Did you think of loading the PI (all four cores doing something... )? The 3B easily exceeds 2000ma in normal usage, and often goes as high as 2500ma. In my robot I'm using a 3amp DC-DC 12v -> 5v converter and an 80amphr gel cell. (its an R2D2 look-alike, fairly heavy, and frankly most of its mass is going to be the 12v gel cell).

Anyway, based on your chart above, I think you are going to be way under-estimating your power consumption.
I was going to agree with you initially but went back and re-read the post. I think the figures are accurate as a baseline for a pi camera recording a static view - and that's all. For a off-the-grid camera recording project it makes sense.

I'm wondering if underclocking / undervolting would have any effect on the power draw.

occupied
Posts: 42
Joined: Mon Apr 25, 2016 3:50 pm

Re: Current Consumption Figures

Wed May 11, 2016 6:59 am

MarkHaysHarris777 wrote:
occupied wrote: I ran some checks on the current consumption of various Raspberry Pi models as a precursor to a battery-powered project.
I question your results; especially the PI 3B. Did you think of loading the PI (all four cores doing something... )? The 3B easily exceeds 2000ma in normal usage, and often goes as high as 2500ma. In my robot I'm using a 3amp DC-DC 12v -> 5v converter and an 80amphr gel cell. (its an R2D2 look-alike, fairly heavy, and frankly most of its mass is going to be the 12v gel cell).

Anyway, based on your chart above, I think you are going to be way under-estimating your power consumption.
Thanks for your reply. I have no way of ascertaining how accurate the MUKER-TM103 is, but really what I was after was a comparison between the models so I know which to work with (in my case, Raspberry Pi 2B). I did note that my figures seem to tally up with others I have seen elsewhere e.g. the saving by disabling HDMI and USB bus.

Accuracy unknown, precision good ;)

I didn't load the cores because I have no need for that in my project. Each test takes about 5 hours to complete so I had to limit the scope.

Pithagoros
Posts: 580
Joined: Wed Nov 12, 2014 8:16 pm

Re: Current Consumption Figures

Wed May 11, 2016 7:02 am

MarkHaysHarris777 wrote: The 3B easily exceeds 2000ma in normal usage, and often goes as high as 2500ma.
Got to ask what is your definition of "normal usage".

Mine doesn't go anywhere near [email protected] in normal use.

occupied
Posts: 42
Joined: Mon Apr 25, 2016 3:50 pm

Re: Current Consumption Figures

Wed May 11, 2016 7:05 am

tweak42 wrote: I was going to agree with you initially but went back and re-read the post. I think the figures are accurate as a baseline for a pi camera recording a static view - and that's all. For a off-the-grid camera recording project it makes sense.

I'm wondering if underclocking / undervolting would have any effect on the power draw.
Thanks for your reply. I did consider underclocking the pi but I had read the effect was negligible. However, now I have narrowed things down to the Pi 2B I'll most likely explore the effects of underclocking. My project won't be CPU intensive.

occupied
Posts: 42
Joined: Mon Apr 25, 2016 3:50 pm

Re: Current Consumption Figures

Wed May 11, 2016 7:10 am

Pithagoros wrote:
MarkHaysHarris777 wrote: The 3B easily exceeds 2000ma in normal usage, and often goes as high as 2500ma.
Got to ask what is your definition of "normal usage".

Mine doesn't go anywhere near [email protected] in normal use.
For what it is worth, I've run my Raspberry Pi 3 on a sub-2A power supply without any issues. To get over 2A I'd imagine you'd need to be connecting things via USB

Pithagoros
Posts: 580
Joined: Wed Nov 12, 2014 8:16 pm

Re: Current Consumption Figures

Wed May 11, 2016 7:19 am

occupied wrote: For what it is worth, I've run my Raspberry Pi 3 on a sub-2A power supply without any issues. To get over 2A I'd imagine you'd need to be connecting things via USB
Indeed, I haven't bothered to get any 2.5A power supplies for my 3Bs. I have found that a PSU that has a clean output and can respond quickly to changes in current demand is FAR more important than the max rated output.

User avatar
Gerd
Posts: 66
Joined: Wed Mar 16, 2016 10:48 am
Location: Europe

Re: Current Consumption Figures

Wed May 11, 2016 12:00 pm

Thanks for sharing your results. Using an A+ or *cough* Zero would shave off annother 100-120 mA current, leading to double uptime! So why not getting an used A+ for little money?

Massi
Posts: 1691
Joined: Fri May 02, 2014 1:52 pm
Location: Italy

Re: Current Consumption Figures

Wed May 11, 2016 12:18 pm

definitely i'm gonna disable hdmi
i've never used it and i'd never evaluated its consumption

mikerr
Posts: 2774
Joined: Thu Jan 12, 2012 12:46 pm
Location: UK
Contact: Website

Re: Current Consumption Figures

Wed May 11, 2016 12:47 pm

Unfortuneately those usb ammeters are often wildly inaccurate - especially at less than 100mA

E.g. your figure of 28mA for "powered down" is quite far off - 80mA is more realistic for all pi.
Android app - Raspi Card Imager - download and image SD cards - No PC required !

occupied
Posts: 42
Joined: Mon Apr 25, 2016 3:50 pm

Re: Current Consumption Figures

Wed May 11, 2016 1:00 pm

Gerd wrote:Thanks for sharing your results. Using an A+ or *cough* Zero would shave off annother 100-120 mA current, leading to double uptime! So why not getting an used A+ for little money?
You're most welcome.
I would love to try a Zero (for a different project) but they seem hard to come by.
If I get access to the A+ I'll test it and add the results.

occupied
Posts: 42
Joined: Mon Apr 25, 2016 3:50 pm

Re: Current Consumption Figures

Wed May 11, 2016 1:03 pm

mikerr wrote:Unfortuneately those usb ammeters are often wildly inaccurate - especially at less than 100mA

E.g. your figure of 28mA for "powered down" is quite far off - 80mA is more realistic for all pi.
Thanks for your reply. I suspect that the accuracy may not be a strong point. What I find interesting is the precision. Repeat the tests and I get the same figures, time after time, consistent across each model and varying significantly between models.

I'd love to have a bench power supply but I can't warrant the expense.

The main purpose of what I did was for comparison purposes between models, rather than getting accurate figures. I'm doing some rundown tests on battery packs now which should give me some more data.
Last edited by occupied on Wed May 11, 2016 1:10 pm, edited 1 time in total.

occupied
Posts: 42
Joined: Mon Apr 25, 2016 3:50 pm

Re: Current Consumption Figures

Wed May 11, 2016 1:06 pm

Massi wrote:definitely i'm gonna disable hdmi
i've never used it and i'd never evaluated its consumption
It is quite a nice saving if you run off batteries :)

User avatar
MarkHaysHarris777
Posts: 1820
Joined: Mon Mar 23, 2015 7:39 am
Location: Rochester, MN
Contact: Website

Re: Current Consumption Figures

Wed May 11, 2016 3:14 pm

Pithagoros wrote:
MarkHaysHarris777 wrote: The 3B easily exceeds 2000ma in normal usage, and often goes as high as 2500ma.
Got to ask what is your definition of "normal usage".
Mine doesn't go anywhere near [email protected] in normal use.
First, I don't believe you've even measured it! (place a good ammeter in series with the 5v power feed so that you can actually measure the current while using your PI)

Second, normal use is completely subjective, isn't it? If I tell you what normal use is, you'll disagree and tell me that I'm wrong... who cares. I actually use my PIs... all of them are functioning as one server or another, all of them are remote desktop PCs, two of them are also GPIO controllers, all of them have usb connections, two of them also have ethernet connections, and most of them are using at least two cores all of the time... they get used! They are real computers (not toys) and they actually process data all the time... for me that's normal use.

I used to use multiple desktop towers to do this same work (mostly e-machines). I have completely replaced my compliment of towers with Raspberry PIs. My entire functioning network of personal computers (besides my mac, and my HP notebook) are Raspberry PIs... and they are the network workhorses. Again, they are being used.

I have actually measured the PI current on the 5v supply during operations (PI 3B) and they routinely exceed 2000ma/ and they get hot (all of them have heatsinks) but without throttling. My 2.0A PSU was not enough to keep the 3B running right. They are on dedicated 2.5 or 3.0 amp PSUs now.

You may claim what you wish... I am finding it hard to believe that my situation is any different than anybody else's setup, with the exception of Rive; no, I don't not deliberately overdrive my PI just to experiment with running it at the high end (not even)... I just use them, in a normal way (as personal computers).
marcus
:ugeek:

ElenoreLaw
Posts: 1
Joined: Wed May 11, 2016 4:23 pm

Re: Current Consumption Figures

Wed May 11, 2016 4:39 pm

Hi...i am new to this forum and i have to ask did you think of loading the PI? The 3B easily exceeds 2000ma in normal usage, and often goes as high as 2500ma. In my robot I'm using a 3amp DC-DC 12v -> 5v converter and an 80amphr gel cell. Anyway, based on your chart above, I think you are going to be way under-estimating your power consumption.

IC programming
Last edited by ElenoreLaw on Tue May 17, 2016 9:19 pm, edited 1 time in total.

User avatar
davidcoton
Posts: 4010
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: Current Consumption Figures

Wed May 11, 2016 5:11 pm

MarkHaysHarris777 wrote: You may claim what you wish... I am finding it hard to believe that my situation is any different than anybody else's setup, with the exception of Rive; no, I don't not deliberately overdrive my PI just to experiment with running it at the high end (not even)... I just use them, in a normal way (as personal computers).
Why do you disbelieve other people's results just because they are not the same as yours?
1) Clearly there are differences in the measurement conditions, which will affect the results. These have been (to some extent) documented. As you say, "normal" does not mean "standard".
2) Clearly the measurements have mainly been taken on uncalibrated instruments. I make electrical measurements professionally, and my instruments are not routinely calibrated -- however they are regularly checked against a standard "checkbox" -- any significant variation in readings indicates a need for recalibration. Is yours ammeter calibrated, or otherwise checked, or do you just believe a piece of test equipment you happen to own?
3) Anecdotal, and maybe totally wrong: despite the above difficulties in comparing results, the power consumption of Pi3s does seem to vary significantly. Some seem to overheat (throttle) easily, some need to be pushed hard before needing a heatsink. Investigating this further is (to me) the interesting project -- particularly trying to do it as a community, without one person or organisation needing to buy dozens of Pi3s at different times. For example, can we define a standardised test that will get different Pis working at a comparable load, in comparable environmental conditions, with comparable power supplies? How do we measure consumption (or the related heat rise) in a comparable manner without calibrated instruments? Can we then gather enough results for statistical validity?
Signature retired

dgordon42
Posts: 760
Joined: Tue Aug 13, 2013 6:55 pm
Location: Dublin, Ireland

Re: Current Consumption Figures

Wed May 11, 2016 5:39 pm

Interesting to contrast the reported power consumption figures in this thread with those of the Foundation here.

Must make some measurements myself.

Dave.

occupied
Posts: 42
Joined: Mon Apr 25, 2016 3:50 pm

Re: Current Consumption Figures

Wed May 11, 2016 9:37 pm

To throw another cat amongst the pigeons - real world figures.

RAVPower Element 10,400 mAh power bank:
Image

Pi 1B displaying raspivid output via HDMI plus the edimax wifi running.

At the start, the output was shown as 5.05 V with a fairly stable 0.60 A.
The voltage and mAh figures were noted at various intervals.
At 11 hours the voltage was down to 5.00 V.
At 11:14 the Raspberry Pi powered down, showing 6,787 mAh.
Graph.png
Graph.png (7 KiB) Viewed 4359 times
Please feel free to correct my calculations:

10,400mAh @ 3.7V = 7,696mAh @ 5V
The Pi powered down at 6,787mAh, which is around 12% less.
No published figures on conversion losses for this product that I could find.

gordon77
Posts: 4111
Joined: Sun Aug 05, 2012 3:12 pm

Re: Current Consumption Figures

Wed May 11, 2016 9:47 pm

My Pi3B with pi camera working, and with wireless keyboard and mouse, and using a python script and raspistill taking photos at approx 5fps, is OK on a ipod 1A supply.

User avatar
MarkHaysHarris777
Posts: 1820
Joined: Mon Mar 23, 2015 7:39 am
Location: Rochester, MN
Contact: Website

Re: Current Consumption Figures

Wed May 11, 2016 10:23 pm

davidcoton wrote:
MarkHaysHarris777 wrote: You may claim what you wish... I am finding it hard to believe that my situation is any different than anybody else's setup, with the exception of Rive; no, I don't not deliberately overdrive my PI just to experiment with running it at the high end (not even)... I just use them, in a normal way (as personal computers).
Why do you disbelieve other people's results just because they are not the same as yours?
I reject the premise of your inquiry. I disbelieve the measurements because they are out-of-line with experience of not only myself but others. I have several PI 3B from two sources and three vendors... and they ALL behave precisely the same way. Other scientists on this forum have reported similar data (experiential and otherwise). The OPs results are not consistent with that experience. (there is nothing personal nor arrogant about it)
davidcoton wrote: ... the power consumption of Pi3s does seem to vary significantly. Some seem to overheat (throttle) easily, some need to be pushed hard before needing a heatsink. Investigating this further is (to me) the interesting project -- particularly trying to do it as a community, without one person or organisation needing to buy dozens of Pi3s at different times. For example, can we define a standardised test that will get different Pis working at a comparable load, in comparable environmental conditions, with comparable power supplies? How do we measure consumption (or the related heat rise) in a comparable manner without calibrated instruments? Can we then gather enough results for statistical validity?
I agree with most of what you have said... with the exception that I reject (at this time) the assertion that the PI3B has been manufactured with such a wide variance of engineering tolerances (varying components, &c) that there could possibly be such a wide range of measurements on the board... if this is indeed the case, then there is a significant quality control problem which should concern the RPF.

As for the rest of your comment, I concur; and I will add that the quicker everyone gets to the bottom of this the better.
marcus
:ugeek:

boyoh
Posts: 1314
Joined: Fri Nov 23, 2012 3:30 pm
Location: Selby. North Yorkshire .UK

Re: Current Consumption Figures

Wed May 11, 2016 10:46 pm

Tests carried out would only prove accurate when don under test bench conditions
With instrument calibration checks, Testing two identical Pi might give different results
Then you are left with a problem , witch is correct.

Best way might be to put a 10ohm high wattage resistor in the power supply lead
And check the volt drop on each test ( Ohms Law)
BoyOh ( Selby, North Yorkshire.UK)
Some Times Right Some Times Wrong

User avatar
davidcoton
Posts: 4010
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: Current Consumption Figures

Wed May 11, 2016 10:47 pm

MarkHaysHarris777 wrote:
davidcoton wrote: ... the power consumption of Pi3s does seem to vary significantly. Some seem to overheat (throttle) easily, some need to be pushed hard before needing a heatsink. Investigating this further is (to me) the interesting project -- particularly trying to do it as a community, without one person or organisation needing to buy dozens of Pi3s at different times. For example, can we define a standardised test that will get different Pis working at a comparable load, in comparable environmental conditions, with comparable power supplies? How do we measure consumption (or the related heat rise) in a comparable manner without calibrated instruments? Can we then gather enough results for statistical validity?
I agree with most of what you have said... with the exception that I reject (at this time) the assertion that the PI3B has been manufactured with such a wide variance of engineering tolerances (varying components, &c) that there could possibly be such a wide range of measurements on the board... if this is indeed the case, then there is a significant quality control problem which should concern the RPF.
I prefer not to reject a possible explanation until there is evidence against it. Yes, such a variation should not happen. That does not mean it has not happened. Batteries should not catch fire. Vehicle parts should not need to be recalled. Food should not be contaminated when sold. All three have happened.
Signature retired

User avatar
davidcoton
Posts: 4010
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: Current Consumption Figures

Wed May 11, 2016 10:55 pm

boyoh wrote:Tests carried out would only prove accurate when don under test bench conditions
With instrument calibration checks, Testing two identical Pi might give different results
Then you are left with a problem , witch is correct.
My "thought experiment" is to find a way of making valid, comparable measurements under distributed conditions. Conventional wisdom says it is not possible. Many other scientific and engineering practices were not possible until someone cracked it with new insights. Unfortunately I don't have the necessary insight yet...
Best way might be to put a 10ohm high wattage resistor in the power supply lead
And check the volt drop on each test ( Ohms Law)
If the current draw is around 1 amp, a 10 ohm resistor will drop 10 volts (and dissipate 50 Watts) -- but not from a 5V supply. The resistor value needs to be nearer 0.1 ohms. I don't think the connections to it could be made consistently good enough not to affect results, not everyone can solder. That sort of issue is at the heart of the problem.
Signature retired

MarkTF
Posts: 297
Joined: Tue Mar 03, 2015 4:59 pm

Re: Current Consumption Figures

Thu May 12, 2016 3:45 am

mikerr wrote:Unfortuneately those usb ammeters are often wildly inaccurate - especially at less than 100mA

E.g. your figure of 28mA for "powered down" is quite far off - 80mA is more realistic for all pi.
I'd expect that they're probably good to a few percent of the full scale reading. Since there is no subranging, that means that with a 3 Amp full scale USB meter that any measurement is realistically going to be +/- a few ten's of milliamps uncertainty at best. This likely shows as a bias rather than a random (per measurement) error so the results would be repeatable as the original poster noted.

Return to “General discussion”