How to estimate the battery life of a product. In this example, the BM786 multimeter.
How to measure battery current accurately, avoiding Burden voltage, and looking at battery datasheets characteristic curves. The difference between alkaline and lithium primary batteries.
If you find my videos useful you may consider supporting the EEVblog on Patreon:
Why use AA's instead fo 9V batteries in multimeters?
00:00 - What is the battery life of the BM786 multimeter?
01:55 - DaveCAD - Linear Regulators are Constant Current loads.
03:49 - The Rohde & Schwarz NGA100 PSU is pretty schmick
05:00 - The Burden Voltage problem
06:34 - Current consumption in AC mode and backlight
07:58 - The battery cutout voltage and hysteresis
08:48 - Why do multimeters waste battery capacity?
09:21 - Alkaline Battery Datasheet time
10:06 - DC-DC Converters are Constant Power loads
12:00 - How good are Energizer Lithium batteries?
14:26 - Temperate effect on lithium batteries
15:30 - Different types of characteristic curves
16:36 - Extraporlating the curve
18:46 - Conclusion
Web Site:
Other channels:
#ElectronicsCreators #Battery #Measurement

Hi Somebody on the EV blog Forum asked what is the battery life of the Bm786 multimeter and well, the answer is I don't precisely know Um, and of course it depends on the type of batteries uh, you're using in it. Of course the uh, the specs in the manual are just that. it is just the current consumption I.E 8 milliamps in most normal functions, 10 milliamps in AC plus DC So anyway, I Thought we'd actually measure that and then actually look at how you would estimate the battery life of a product. If you can't, simply plug the batteries in and leave it on and have a timer there.

you know if you've got something that's got hundreds of hours battery life, uh, then you know that could be a problem. Sometimes you want to get a nice quick estimation of a battery life. so let's see how we do that. And I'll link a video up here and down below if you haven't seen it.

It's from my second channel from quite a few years ago on. uh, why the 121 GW multimeter uses uh, double A batteries instead of nine volt batteries and that has some interesting stuff in it if you're interested. So the Bm786 it uses three AAA batteries like this in a quite unusual, uh, vertical, uh battery configuration. Like that, it's got a spring and terminal down there, which makes it a real pain in the butt to get in there and actually, uh, measure So we could of course skip this measurement and just use the data sheet value.

But let's double check to see um, what Bryman actually say the power consumption is. They say eight milliamps. Well, let's actually measure it. Put some bar tape in there.

Just a temporary thing like that, and there you have it on DC volt mode. It's five milliamps. It's not 8 milliamps, so it's less than we thought. Uh, three AAA is, of course, a nominal 4.5 volts.

We can actually, uh, go above that and it's going to stay the same current and it should stay the same current. When we drop it like this, let me explain why. Let's go to Dave CAD here and I'll show you how this particular load ie. this multimeter actually works.

It's going to have a linear regulator inside. none of that switching regulator rubbish. It might have multiple linear Regulators but for the purpose of this discussion, it doesn't matter. We've got our battery over here.

Okay, And then we've got our current. uh, draw from the battery which is what we just start measured that five milliamps or so. And then we've got current which goes into the load, which is all the circuitry for the multimeter. Now, of course, the low can switch and you can have, like, you know, pulse currents and all sorts of things.

But a multimeter like this is like a fairly consistent five milliamp load there, as you uh, saw. So this linear regulator, of course, is going to have a little bit of quiescent current. Uh, it's called down here, which is why I've labeled it IQ Like this, but generally that's in like the micro amp region or something like that, whereas your load is in the milliamp regions. So as a rule of thumb in electronics, anything more than like two orders of magnitude.
so you know a hundred, one, one hundredth of something, you can pretty much rule it out. That's like one percent error doesn't matter, and if you get lower than that, well, it, you know it just becomes insignificant. So you can rule out IQ So for a linear voltage regulator, and this is something you should know as a basic building block thing for. Uh Regulators Like this is your output current, your load current is equal to the input current.

So effectively, because the linear regulator is a fixed voltage across, essentially a fixed load. Your effect essentially getting a fixed current or a constant current. Essentially so, as you'll see in a minute in the battery are data sheets we're drawing from the battery. essentially a constant current load.

If you're using a linear regulator, that's what you've got switching. Regulators Very different thing. anything with changing loads. Very different again.

But in this case, we've got essentially a simple constant current load. Which makes things really easy when we're trying to estimate battery life, as we'll see shortly. Just as an aside, the reason I'm using this rodent Schwartz NGA 100 power supply is because it's the only power supply I've got here that has a real low current mode. It's got like Precision measurement k capability ideal for measuring low current devices like this.

As you can see, it's got one micro amp resolution there which is absolutely incredible and I can actually go in there and modify that to like a fixed high range for example. and you'll see that even in the Uh 2 Amp range, it can get 10 microamps resolution. Terrific stuff. Not all power supplies have this and if your power supply doesn't very likely then you'll have to add a multimeter in the in series with the output to measure the current.

And here is where you can come a gutter Because here's where burden voltage becomes a big problem. Then this is why I developed specifically the microcurrent is so that it minimizes the burden voltage of your multimeter because the absolute last thing you want when you're trying to measure the voltage Dropout point of your product over here because we have to adjust this four and a half volts until we get down to where you know the low battery warning comes on or the product stops working in. So we need to know what that voltage is precisely. So I'll illustrate the problem here.

You've got your power supply. You've set it to precisely 4.00 volts. Okay and then, but it doesn't have an accurate current meter so you have to put an external multimeter in here and the current shunt inside the multimeter has a resistance and that's going to have a voltage drop when the current flows through it. It's just simple Ohm's law.

So depending on the current range you choose in the multimeter depends on which value shunt resistor I have. You're going to get a voltage drop when you pass current through your device under test. Now, if you're at very low currents, it may not matter, right? but then you don't get the resolution on your current range on your multimeter like you might want. So you might think that you're at four and a half volts.
uh, battery. And that's where you see your battery uh Dropout sign. Come on. But you're actually it's like three volts or three and a half.

Some multimeters can have like a one volt uh, burden voltage on them or more. Now of course, to overcome this, you can put in a second Digital model meter in here and you can actually measure the voltage uh, actually right at the input terminals. but you're going to need two multimeters for that. As I said, always have two multimeters so just be careful.

You can really come a gutter with that and your measurements won't be accurate. Now you could of course have a low burden voltage multimeter like the 121. Uh GW. There's not many on the market that have low burden capability, but it just be aware.

just because it's low burden, it's it's lower than a regular multimeter, but might not be low enough for your particular requirements. Just be aware of that. Read the manual, know what your burden voltage is and for those playing along at home in AC plus DC mode, Here, we're measuring about nine milliamps and if you want to know what happens if I turn the backlight on, there you go backlight. Oh, it's a killer 42 milliamps.

Oh, anyway, let's measure and then estimate the battery life of this product. uh, without having to time the thing. So we'll take that as five milliamps and we start out at the maximum voltage there: four and a half volts. It could go above that, but you know it won't stay that uh far for long.

for Alkaline R batteries. for example, if you're using Lithium, it could go higher. but as I said, we can actually go in there. and if we adjust the voltage 5 volts for example, there you go.

It's still only going to be drawing five milliamps because as I said, it's a constant current load. Then now lower this until our battery uh icon comes on. or however, you determine that your product stops working. Now, just be aware there's maybe some lag on the time it takes for the low battery uh symbol to come up.

So I know it's not going to be four volts so there's no need to muck around there, so it's going to be somewhere below that. There we go I saw it come on there at 3.5 something. So at 3.6 Oh yeah, 3.5 out. Now there could be some hysteresis here I.E when it actually switches back off, but that doesn't really seem to be.

Yeah, so let's definitely say 3.58 volts there. Okay, and we're going to assume that the meter is still accurate just at that point where it comes on. Now, of course, you could go into more advanced stuff, like actually measuring the performance of your instrument, but that should have been part of your design process rather than just, you know, uh, like estimating the, uh, battery life, which is what we're doing here now. 3.58 divided by three.
Get the confuser out. Uh, that's a smidge under 1.2 volts. So we definitely aren't extracting all of the energy from these AAA batteries because I've done a ton of battery videos and I'll link in the playlist. Uh, down below, and batteries do contain usable energy right down to 0.8 volts.

But in products like this that use a linear regulator like this, or multiple linear Regulators then So yeah, it is common to weigh some energy in your batteries like this, and that's just an unfortunate side effect of a product like a multimeter that you don't want to have a switch in power supply in here because that can screw things up. You know, this is a Precision measurement device, so the last thing you want in this is a switching converter to power your like dual slope a to d inside this thing, you know Precision measurements? not yeah, no thanks. So it's common in products like multimeters like this to sacrifice uh, some of the energy in the batteries for measurement performance and uh, simplicity. So let's look at some data sheets.

Let's just take a Duracell ultra power AAA here and we know we've got a constant current load here and this will go into the different varieties and sure enough, look at this constant current graph. They've got two constant current graphs here, but one one one Murphy's Law Wouldn't you know they've got one milliamp and 10 milliamps. They don't have 5 milliamps and as you can see, it's quite a large discrepancy between these, so you could actually like, try and guesstimate it like by sort of matching that but like it's probably not going to be in the middle, it's more likely to be like further over to here it's going to be a non-linear thing going from one milliamp year to 10 milliamps here and then. as I mentioned, there are different load types.

like if you had a DC to DC converter in your product for example, then all things being equal. it's essentially a constant power thing. but of course it has to do with the draw like the efficiency of your DC to DC converter over the load range and all sorts of things right? But essentially if you've got a decent diesel converter, you'd be looking more at the constant power graphs than you would for the constant current graphs. Now, unfortunately, we can't just use Ohm's lowering.

You remember the voltage is dropping like that. So yeah, um, it's not the correct thing to use. And also a thing to note that is these: All these things will not only depend on the manufacturer of the battery, they'll depend on the batch of the battery, they'll depend on the type of the battery, they'll depend on the chemistry of the battery, and they'll depend on the temperature as well. And you can see here for the same uh draw they don't.
They assume it's like constant, uh, Power draw. Look, there's quite a difference in service hours from like three hours up to nine hours here of battery life from -10 to 21 degrees. So that's only a 30 degree. well, 30 degrees.

There's a lot of temperature differential, but that is a huge difference in battery life. Two and a half hours to nine hours just for temperature, right? So we go over to an Energizer over here. Well, they've got a milliamp hour capacity here. Continuous, uh discharge at a constant discharge current.

So constant current. But why only 25 milliamps? That doesn't help us. We need five milliamps and then down here on the industry standard tests, these are actually a resistive load. 5 ohms, 24 Ohms.

They've got one down here, which is a constant current 250 milliamps. They've got another one which is 50 milliamps down here. But they're like milliamp hours per day and stuff like that. So yeah, nah, it's not what we want now.

Of course, if we want the ultimate battery life in a product like this, we would use a Lithium primary battery. Check it out our discharge profile. We don't get five milliamps. We only get one milliamp and 10 milliamps.

But look at the differential right in these graphs. Unfortunately, it's not rendering. My PDF viewer is not odd there. It didn't render that properly I Found a bug in drawboard PDF which is a software I Use A lot of people actually ask me which software I use Anyway, you can see that it is somewhere between these two lines here, right? And as you can see the height, the longer up you go, the more battery you discharge.

There's essentially no difference between 1 milliamp and 10 milliamp draw. Essentially, so we're essentially good to go. We can accurately use this. So where did we, uh, determine? It was just under 1.2 volts.

If you remember per cell, you remember we've got three cells in series. Uh, so we were 3.58 volts divided by three. so you can see down at 1.2 If we extrapolate that over here, you can see that we're really extracting pretty much at the absolute maximum energy out of a Lithium battery. Beauty We're not wasting any, as we'll see that we will in an alkaline.

So yeah. Fantastic. So let's just call that like 1200 milliamp hours. So we get the confuser out.

Uh, 1200 milliamp hours divided by 5 milliamps there. 240 hours of battery life for the Bm786 on uh, using these Lithium primaries, so that's going to be reasonably accurate and does and what the temperature should have lesser effect on these as well, which is one of the advantages that they actually have and you can really see. Look at this, you can really see the ESR just like go through the roof right there. This is how batteries die.

Their internal series resistance just suddenly goes and yeah, and it just doesn't give you the volts anymore. It Ohm's law Like, well, you're talking uh, Kirchhoffs now and they do have a nice graph over here. and as I mentioned before, that non-linearity right? They actually show you it's they use a log graph here. This is constant Uh current here.
So so Two, three, four, five, so five up there. Uh yeah, we're talking about that 240 odd hours we were talking about before. so you know that that Line's a bit Rough and Ready. But yeah, you can get an estimate from that.

And yep, they give us a temperature effect on capacity for constant current. Unfortunately, only at uh, 25 milliamps there. but you can see that the 25 milliamp is the thick one there. so at the lower currents and five milliamps, that'd be like flat as attack, right? So there's effectively no change uh, in the battery life over temperature with these Lithiums, But that's what you expect.

that's why you pay a huge premium for it. Anyway, We want uh, Alkaline or Manganese um, dioxide, Zinc Manganese dioxide here for those uh playing along at home. So let's go to the Energizer again. but this is the Energizer Max.

You can see that this Energizer Max data sheet is significantly different. It's got milliamp hour capacity and then constant power performance. constant current performance here. whereas before we got like the these industry standard tests.

So the max version is you know, significantly different data sheet. We do have constant current performance. What do we got here? What lowest value we've got is 10 milliamps here. So 5 milliamps is off the graph so our 1.2 volts cut out.

You can see how all of these manufacturers, like even the same manufacturers can give you totally different curves. You know you've got to be able to interpret all these different types of graphs. This one gives you a characteristic curve of voltage here. like this: a plotted against the Uh discharge in milliamps and giving you the surface hours.

So we would have to extend this graph right up here like this to try and get our five milliamps here. But as I said, it's a logarithmic scale anyway, so we do know the figure is somewhere above 100 and should at least do that so we can go to like a Phillips brand here. They don't give us anything. Let's try the Germans Vata Shall we discard? Does type load five items 24 Once again, this industry standard stuff here.

That's all they got. Let's go for a Panasonic jobby shall we? And once again, we've got like significantly different characteristic. We've got load in Milliamp like this is real old school. Like they've like plotted this like on an actual pen plotter or something.

But once again, the five Milliamps that we want if we extrapolate that up, the curves don't go that far. Once again, it's the end voltage per cell. So we're looking at the 1.2 volt curve here. So yeah, like eyeballing that it's somewhere between 150 hours and 200 hours? I'd say it's not more than 200 hours.
Ah, like we've got nothing that's absolutely precise yet. Then we've got this Panasonic Alkaline Handbook here. Once again, you can see that they stop at 10 Milliamps here, but you know once again, 1.2 volts, this is the lowest curve here I Don't know I that's a bit how you doing, isn't it? Five Milliamps go up there like that and yeah, we're talking 180 hours. So as it turns out on that same EV blog Forum thread which I'll uh Link in down below for the Bm786 uh Joe Smith uh, following his videos down below, you know it does Multimedia destruction.

Multimeter tense: extensive multimeter testing. That's pretty much all he does on his channel. Um, he measured 118 hours on the Bm786. I you know it's like 118 is like is down here I can't see a tail in off like that.

but as I said, huge discrepancy temperature battery type battery Brands Uh, slightly different electrochemistries between manufacturers. All sorts of things apparently. Uh, some people have measured like eight Milliamps on the Bm786, so there seems to be some variability in the actual meter itself. So anyway, it looks like it's easily going to get 100 hours.

We could split it, say 150 hours. Something like that. I Think mine at Five Milliamps will actually last probably 100 1380 hours. Something like that.

I can do another follow-up video. leave it in the comments down below if you want me to do that anyway. There's like more you can do to this and obviously nothing beats actually putting it in the product itself. especially if you get any sort of pulse current.

And as it turns out, uh, that. uh, Roden Schwartz Power Supply I just press the log button there. It is like the Five Milliamps because you can see just in DC Volts mode. There's not really a huge discrepancy in the current there.

and I think most other modes will probably similar when it flashes the backlight and stuff like that. The backlight seems to be you know a major thing, but easily get 240 hours for those Energizer Lithiums over here. But as I've shown you in many uh, other videos, the voltage that you set over here for like the minimum Dropout voltage for your product, that determines how much energy you actually waste and in this particular case, the energy. You have to extend this all the way down to there, all the way down to zero volts down here, this just drops off like a rock.

Yeah, but all of this under here is the energy. The area under the curve is the energy. So all that energy is wasted. So compare the area of that compared to the area under this side of the curve here.

and that's how much energy you're wasting in alkaline batteries Where you don't waste that in the Energizer Lithium ones you saw. There is no area under the curve, right? You're wasted practically nothing of that battery. It's it's fully discharged so your better eyes are ain't going to extract your 800 percent. It's not even going to extract an extra eight percent.
Really. So I hope you enjoyed that video and found it useful. If you did, please give it a big thumbs up. As always discussed down below, catch you next time.

Avatar photo


25 thoughts on “Eevblog 1533 – how to estimate product battery life”
  1. Avataaar/Circle Created with python_avatars Hobby Electronics says:

    Why you design it for 1 cell more so you can use much more energy ?

  2. Avataaar/Circle Created with python_avatars Paula Abdul says:

    Thank you Lord Jesus for the gift of life and blessings to me and my family $14,120.47 weekly profit Our lord Jesus have lifted up my Life!!!🙏❤️❤️

  3. Avataaar/Circle Created with python_avatars Dino says:

    Thanks for the informative video Dave. On the topic of estimating product battery life, have you ever come across the Nordic Semiconductor "Power Profiler Kit 2"? It's a great little tool with a GUI for measuring periodic power consumption in IoT/embedded systems. It's great for bluetooth systems. The GUI lets you highlight parts of the captured waveform and it determines the amount of charge used. It has some limitations, but it's way cheaper than an expensive DC power analyser and useful for many applications. Maybe in a future video you could compare its accuracy with a DC power analyser.

  4. Avataaar/Circle Created with python_avatars WizardTim says:

    I really wish battery manufacturers were required by law to clearly mark on packaging and the batteries themselves the typical Wh or mAh capacity, all of the "SUPER HEAVY DUTY" "EXTRA HEAVY DUTY" "ADVANCED" "ULTRA ADVANCED PLUS" "3x LONGER LIFE" "MAX PLUS" that's currently used is worse than useless.

  5. Avataaar/Circle Created with python_avatars Frank Bose says:

    Lol I ordered a negative feedback Hoody and by the time I get it it will be summer here lmao

  6. Avataaar/Circle Created with python_avatars Mike Adler says:


  7. Avataaar/Circle Created with python_avatars jimkarlock says:


  8. Avataaar/Circle Created with python_avatars Florida House says:

    The conclusion determined in this video is why I have not understood the reasons for using 3 cells instead of 4 when the dropout is so high at 1.2V per cell. A choice of 4 cells would be usable down to 0.9V (x4) since the 3.6V was the lower limit. The space requirements for 4 AAA cells would not be an issue when designing a device like this Berryman looking at the case. If they used 4 cells, then the wasted energy in the graph would now be available from all 4 cells in addition to the extra capacity of the 4th cell so adding a cell would add 50%-80% more battery life while only adding 33% more volume. Less waste overall for everyone at a zero to nominal cost and much improved battery life.

  9. Avataaar/Circle Created with python_avatars John USP says:

    Only 1.2V? Will be a problem for NiMH batteries.

  10. Avataaar/Circle Created with python_avatars John USP says:

    1.2V is too bad. My DMMs works for years with 2AA batteries up to 0.9V

  11. Avataaar/Circle Created with python_avatars Lagittaja says:

    In case you're unaware, there's an awesome website by a danish fella who among other thing tests batteries and battery chargers. It's called lygte-info.
    He has published two articles testing low current discharge.
    Duracell Plus Power AAA at 5mA constant delivers 987mAh with a 1.2V cutoff. If discharged down to 0.8V you get 1396mAh.
    The Energizer Ultimate LIthium AAA at 5mA constant delivers 1271mAh with a 1.2V cutoff. Down to 0.8V you get 1287mAh.

  12. Avataaar/Circle Created with python_avatars Ph Pn says:

    Love these mouldy meaters

  13. Avataaar/Circle Created with python_avatars Christian Schneider says:

    Really interesting topic. We are currently developing a measurement device at work that sits in sleep mode and consumes almost nothing but now and then it wakes up and draws in the region of 1.5A@3V3. I got the task of making a back of the envelope calculation about the battery life and really struggled because when it‘s on, the ESR suddenly becomes important and that can vary greatly over temperature (the product is intended for outdoor use). I also found these battery simulator power supplies. Maybe someone at R&S or Tektronix wants to send one of these to Dave to play around? I‘d love to see more battery videos.

  14. Avataaar/Circle Created with python_avatars Andy Barnard says:

    On the service hours graph with the batteries with 1mA and 10mA curves you should calculate mAh capacity to 1.19V at 21C. The service hours to 1.19V varies enormously but the mAh figures are much closer. Then for 5mA take a figure in between the 1mA and 10mA curve.

  15. Avataaar/Circle Created with python_avatars Кирилл Рагузин says:

    Actually if you have at least a half-decent lab power supply and not some cheapest bottom of the barrel stuff then there is a somewhat risky but overall much better way of completely eliminating the pesky burden voltage of your current measuring multimeter from the equation! You just disconnect sense terminals of your power supply from the power outputs and connect them to the wires running directly to the load after all your extra measurement gear and connections.

    Dave's R&S comes with a connector on the back and a small green terminal block for hooking up your wires for the purpose. And if you are lucky enough to have something like a good old TEC series PSU from soviet Bulgaria then you will just have an extra couple of "sense" banana sockets right on the front panel for extra convenience. Just make sure to connect everything very securely and preferably do not move the wires too much while the power is on or the open loop voltage of your PSU will most likely instantly kill you DUT.

  16. Avataaar/Circle Created with python_avatars Bert Blankenstein says:

    I find temperature a huge factor. At -15C the alkaline batteries last less than half of when it is warm, as indicated in the video. The batteries do not come back to life when they are warmed up.

  17. Avataaar/Circle Created with python_avatars danosdotnl says:

    just picked up this DMM, the test leads are simply a god send, so delicious

  18. Avataaar/Circle Created with python_avatars Ghost Banana says:

    Dave have you seen the post on the eevblog forum in "power/renewable energy/ev" section , the post title is "I caught exact moment when akaline battery started leaking – It makes a sound!"

  19. Avataaar/Circle Created with python_avatars Elliott Veares says:

    Does this method need to be adapted for high current devices, like RC cars, flash cameras, Bluetooth speakers, etc…?

  20. Avataaar/Circle Created with python_avatars Thomas Ives says:

    Hey Dave, ever given a thought to making a rechargeable battery pack for the Brymen? The voltage range sounds about right for a single cell. The form factor is probably better suited to a pouch cell, but maybe an 18650 could fit in there. I have retrofit a couple of no-name DMMs that use a 9V battery to use 2x 18650s, so I almost never buy 9V cells. I've just been using off-the-shelf battery packs glued on or shoehorned into the case, but I've been considering 3D printing an entirely new case cover (back) to accommodate the 18650s – with the Brymen, that would be easier since it is a plug-in module.
    Totally agreed: the Energizer Lithium cells are amazing and I use them in any instrument/tool/camera that costs >$20. It just makes economic sense: the price difference is about $1.50-ish per cell, so it is $4.50 more to power with lithiums. If a very expensive tool (say $100) gets destroyed because of leaking batteries, that's 20+ battery changes worth of value lost. Plus, the lithiums will easily last 10 years just sitting around, so they have a lot higher storage-vs-cost potential.
    Great video, thanks for all you do for the electronics community! Cheers!!

  21. Avataaar/Circle Created with python_avatars name not shown says:

    Can you do a video on dc motor controllers that use an inductor in the circuit.
    Recently opened up a car radiator fan controller and found inductor inside, with one schotkey diode/two mosfets and a few large capacitors.

  22. Avataaar/Circle Created with python_avatars bachger says:

    Hi Dave, why are considering the 5mA current consumption from only one battery cell? How I can interpolate/estimate the lifetime in case of 3x cells? because the sum of Ri is higher.

  23. Avataaar/Circle Created with python_avatars Paul Denisowski says:

    In my experience, a lot of engineering students and young engineers don't appreciate and/or don't understand the importance of battery life testing — despite the fact that so much of their world actually runs on batteries — so this is a very timely and useful video. Thanks!

    (and nice power supply :))

  24. Avataaar/Circle Created with python_avatars Albert says:

    Hello ! Dave, For referencing. 👍

  25. Avataaar/Circle Created with python_avatars Mickey Filmer says:

    Lithium batteries in the UK are riughly £1.30/unit as opposed to £0.60/unit for Alkaline so Lithium are more than double the price but not quite double the lifetime- therefore are they really worth it?

Leave a Reply

Your email address will not be published. Required fields are marked *