Fire Alarm Batteries Apparently Dead/Dying but Multimeter Reads 9.58v--What Gives?

So the fire alarms started going haywire, going off every couple of days (then every couple of hours, then every few mins…) for no good reason. I figured the batteries were exhausted, and replaced them. That fixed it.

But… I also bought a cheapo multimeter for $5 at a dollar shop, and tested the removed batteries just to figure out how it works (I never used one, and don't actually know what I'm doing). The removed batteries tested as 9.58v (all but one, at 8.80v). What gives?? Are the batteries still good after all? Or just not good enough for fire alarms but good for other things? (I used one of them to power the multimeter, for instance, and that worked.) Or is cheapo multimeter just giving me false readings (I did see something about better multimeters applying load on the battery for a more accurate reading.)


  • +5

    My guess is that without a load applied the batteries will show close to their nominal voltage (eg 9V). But if they are flat (or partially exhausted) then the voltage will be lower when a load is applied. I use a multimeter to check batteries but use a resistor connected to the leads so there is a load applied to the battery.

    • +1

      How much resistance? I just tried with a piece of paper, but that was too much for even the brand new battery.

      • +1

        if you are aren't familiar with testing and using a multi-meter it is unlikely you will have a resistor laying about of an appropriate value.
        paper isn't a conductor( generally) and just putting a piece of wire across + and - will give you a spark/ heat…

        The higher the voltage / current the more dangerous this gets when you aren't sure / know what you are doing.

        play safe, and never try to use your multimeter on mains voltage or even high current DC. ( eg a car battery could melt steel if shorted)

        google "battery surface charge" and read up about batteries ( surface charge is mainly on lead acid) but it may give you some answers to your questions :)

      • +1

        I use a 100 ohm resistor which is useful for AAAs, AAs, button cells, 9Vs…
        Also, it is not the type you would commonly find soldered on a circuit board. It has a heatsink.

        • Also, it is not the type you would commonly find soldered on a circuit board. It has a heatsink.

          That's probably due to the wattage of the resistor. Which is also an important number.

          You can find any number of 100 ohm resistors. At 9V, they should pull 90mA which would then dissipate 0.81 Watts (as heat).

          If that resistor isn't rated for ~1W+, using it could damage it, your meter or you.

          • @Chandler: It's 10W. I bought it for this particular use, after researching the requirements for testing AAA to 9V.
            I was just trying to point out to OP not to grab any resistor he/she could find.

            • @snapper: Yeah, sorry my comment was mainly targeted at OP & others to ensure they consider the wattage of the resistor - you made me think of it with your mention of a heatsink.

              It's good having the right resistance value, but if it can't take the wattage you'll be having a bad time.

      • +4

        Have you tried 2kΩ paper? I think you may have used doubly insulated paper - everyone knows that won't work.

    • A smoke alarm should have such a low load that loading the battery really shouldn't make much difference.

      That said if they are truly crappy batteries with a high source impedance, then that may be it but I doubt it.

      I'd be replacing the whole thing. They are fairly cheap and will come with proper 10 year lifetime batteries - regular Energizers/Duracells are not rated for smoke alarms.

      Smoke alarms should be replaced after 10 years generally too.

      • Ours are only a couple of years old. It's a new building.

  • +1

    Same here. Tested at 8.77V. Guess it's normal.

  • +2

    Cheap multimeters are usually not accurate. Get one that is decent quality if you intend to use it in the future because inaccurate readings will couse lots of headaches. A fully charged 9v battery should read around 9.5 volts. If your multimeter is reading 9.58 volts on used batteries there is a problem. Also, a good way to test the battery is to try and load them while measuring the voltage. If you have a resistor that is around 200 ohms, connect it to the battery while measuring the voltage. If the voltage is lower than 6 volts, the battery is probably dead.

    • I'm really curious about the opinion of those who neg'd your comment now.

      Does the resistor go in serially or in parallel with the multimeter? Or it doesn't matter?

      • +6

        It's pretty safe to ignore negs on ozBargain. No-one has ever been able to work out why the idiots neg comments that are factually correct. I guess they have to do something to feel important.

        The resistor goes in parallel with the multimeter. So, connect the two multimeter leads to the two leads of the resistor, then connect the resistor across the battery only for long enough to get a stable reading. 470 ohms might be a better choice if you want to use a single resistor with any kind of consumer battery.

        • BTW Consumer truck batteries are good for welding if the cells are good.

          • @AndyC1: I don't think that truck batteries are really consumer items, and I wouldn't expect them to last too long if you used them for much welding.

            • @pjetson: I concur. Emergency short duty spot welding if you have the bits and know how in the field is possible with truck batteries, but for something more substantial, alternators(basically a 240v generator) can be used with a lot more know how and many more bits.

        • It's pretty safe to ignore negs on ozBargain. No-one has ever been able to work out why the idiots neg comments that are factually correct. I guess they have to do something to feel important.

          They should consider adding the same requirement to comments as deals - you want to neg? You have to put a comment.

          The resistor goes in parallel with the multimeter.

          If you're measuring voltage (which you are in this case) - if you want to measure current (Amps), you put it in series.

          Also - ensure your expected measurement is within the specifications of the meter. Don't want to fry/blow anything (including yourself).

          And if the voltage exceeds the specification of the meter, you probably shouldn't even be touching it - get an electrician.

          • @Chandler: Put the resistor in series on an unkown capacity 9v battery will give you volts through the resistor and the internal battery resistor. Great. Get as close as the same result without that external resistor.

            An external resistor in parallel will load down the battery and the voltage will drop accordingly. Not actually reading amps across or through the battery but the resulting voltage will give a good idea of the voltage capability. i.e. a large voltage drop like down to 7v or lower means that it's (insert your own description word/s for a failed battery).

            EDIT: Read and learn if you aren't aware of the internal resistance of a battery:

  • +1

    Testing battery unloaded does not give a true voltage when it is in a circuit. battery testers have a relative high ohm resister in them and as such effectively give a voltage that the battery would supply in a circuit.

    If you have a resistor between 1K & 2.2K put one between the battery terminals and then measure the battery voltage.

      • +1

        100 ohms across a 9v battery would draw 90mA. That's far too much - 9v batteries are designed for no more than about 15mA maximum. 100 ohms is also too low for AAA batteries and coin and button cells.

        Still, if you only draw that current for a second or two, you're not doing too much damage. You'll probably think a 9V or AAA or coin cell/button battery is flat before it actually is, though.

        • You are talking about a person who bought a cheap multi meter and does not know about electronics and you want them to only connect the circuit for a few seconds…… cross fingers and hope it does not overheat and………………..zzzzzzzzz

        • Have a look at the datasheet of this energizer 9v battery.
          It shows that the battery is rated to have a capacity of about 400mah at 50ma current (which is drawn with a 200 ohm resistor). Although drawing 50ma from a 9v battery continuously will cause it to have a lower capacity, it is good for testing purposes because the voltage reading with a 5ma current gives a better indication of the remaining capacity of the battery. This is because an almost fully discharged battery might still work if you draw 10ma but it won't if you draw 50ma.

  • +5

    What does the tongue test read?

    • +6

      Medium tingles

      • Good tingle I think for a few seconds until the effective capacitance charge holding the voltage high dissipates.

  • The removed batteries tested as 9.58v (all but one, at 8.80v). What gives??

    Voltage is easy to get but it has no Amps to flow.

    Did you ever got "electricity" from potatoes at school Physics?
    Voltage is OK, easy to obtain, but nothing will lit, nothing will work.

    Similar with 9V batteries (or any batteries), there is a potential differential between negative and positive but nothing else of use.

    No juice, just water.

    • The multi meter will be drawing micro or nano amps….

    • So the cheapo multimeter is useless? Or am I useless at using it?

      • The only real problem is that you don't know how accurate it is. It's likely that it will remain as accurate as it is now into the future, so you just need to compare it once with something that is known to be accurate to know how accurate your multimeter is. It's kind of like your car speedo. It's very unlikely that you're actually doing 100kph when the speedo says you are, but once you know how far off it is, you can compensate mentally.

        • For $52??

          Nah, I'd rather stay in electrical darkness 😅

          • @wisdomtooth: That eBay item is a cheap and nasty multimeter dressed in fancy multimeter jacket and sold at jacked up price. 😦 😞

      • Get a Fluke multi-meter and it will be allot more accurate than you need.

      • It used to be that you could buy tiny mercury based hearing aid batteries for a couple of dollars. This type of battery maintained its voltage very accurately for almost its entire service life (1.35 or 1.45v, I can't recall which atm). These were a very cheap way to check multimeter accuracy. Sadly, I don't think they're available any more due to the mercury. I think silver-oxide batteries might be similar, and I'm googling for information on that.

        • It seems that a silver-oxide coin battery will have a voltage of 1.55v over almost its entire life, and for your purposes, it would be a cheap and simple way to check the accuracy of your multimeter on the low voltage ranges. Your multimeter will probably have a 2v range and a 20v range, check it on both since multimeters often do not have the same accuracy on all ranges.

          A silver-oxide battery will probably have a part number starting with SR. Batteries with part numbers starting with LR or CR are not silver-oxide.

          It does appear that for the first few hours of use, a silver-oxide cell will measure slightly higher. Putting a 15K Ohm resistor across it for a day or so will get you past that initial slightly higher voltage period, and it should then read 1.55v for most of the rest of its life.

          • @pjetson: I have a few 3v CR batteries laying around. My meter gives 2.98 at 20 DCV, 1 at 2 DCV and 2.8 at 200 DCV. I'm finding it odd to say it's inaccurate, bc it gives the very same reading every time, even on different days (different readings for different batteries, of course).

            • @wisdomtooth: Inaccurate doesn't mean that the meter will show different readings each time - your meter will probably show the same voltage for years but that doesn't mean it's displaying the correct voltage. It's like your car speedo - if it says 100kph one day, it will say 100kph the next day if you drive at the same speed. However, your car may actually be doing only 95kph, since most car speedos are not accurate.

              Also, using a CR coin cell to test your multimeter for accuracy isn't a good idea because CR batteries do not have a constant known voltage like the SR cells do.

              You can't measure a 3v battery on the 2v range of your meter. I believe that your meter is showing an over-range condition by displaying 1. It's like trying to weigh a 5kg item on a 2kg kitchen scale.

  • +1

    Is like a person with good potential CV but crack under pressure and gets no work done. CV still looks good — what gives

    • Damn this is savage LOL! I know few people like that at work and I'm glad I am no longer apart of that.

  • BTW, if you're using one of those really old smoke detectors… they are get triggered when humidity is very high. I used to have one in my old apartment and it will go off during a very long downpour.

    • :O

      That absolutely checks out, as we were under Qld's last week's heavy rains.

      They don't look old, though. The construction is less than a couple of years old.

  • +1

    (I used one of them [possibly discharged battery] to power the multimeter, for instance, and that worked.)

    If you put a bad battery in a multimeter (especially a $5 model) don't expect reliable operation.

    At the end of the day, even a cheap and nasty multimeter can be a useful tool. Just don't go hooking it up to mains, no matter how many labels they print on the case that suggest a CAT III 600V rating.

    If nothing else your cheap multimeter should allow you to compare one battery against another and see which one is "better". To get a meaningful comparison, you should be taking a measurement when the battery is in circuit. Install the battery under test in a device that uses 9V batteries, turn the device on and then measure the voltage at the battery terminals. Repeat with a new battery. The resulting readings should give you some idea. Please be aware that you need to compare batteries of the same chemistry. i.e. alkaline with alkaline, zinc-carbon with zinc-carbon, etc. Comparing batteries with different chemistries will not be very useful without an accurate meter and some knowledge of the underlying characteristics of those batteries.

    • Comparing batteries with different chemistries will not be very useful

      Why not?

      • +1

        Because different chemistries have different discharge characteristics. Some battery chemistries will have a gradual discharge curve where the voltage steadily drops as the battery loses capacity. With such batteries and a good multimeter, you can make a reasonable guess as to where they are in terms of their remaining capacity. Other chemistries will have steep voltage fall off at the start and at the end of their useful life, but the voltage will remain relatively constant for most of the service life. The voltage of a (discharged) battery will also vary with the chemistry. Those differences can be very small (so a good meter is required) or blisteringly obvious. Some battery chemistries recover their no-load voltage very rapidly, others not so much.

        As a rule of thumb, the typical cell in a traditional primary household battery has a nominal voltage of 1.5V. The battery is considered completely discharged when the voltage of any cell in the battery under load drops to 0.9V. Nine volt batteries are constructed by stacking 6 cells in series. Cells do not discharge uniformly so a completely flat battery will not read 5.4V, it will most likely be somewhere in the 6V to 8V range. Of course, the ability of a device to operate to any specific low voltage depends on the circuit design. Some devices are capable of draining a cell to 0.2V, while others will stop working with cells at 1.2V.

        Suffice to say, batteries are complicated. Even a good multimeter is not an adequate tool to test batteries. You really need a decent battery analyser and a big chunk of knowledge to get definite answers. For most people, the best way to proceed is to swap old batteries for brand new ones and see if it resolves a problem with the device. If it does, discard the old batteries, because measuring their voltage out of circuit will not help much.

        • I see. Perhaps an even better solution is using rechargeable batteries, and rotate them whenever there's a problem. That's what I do. I guess I was attempting to fly to high with my $5 multimeter 😅

          • +1

            @wisdomtooth: I use rechargeables whenever I can and would recommend them for most uses, but smoke/fire alarms are one application where zinc-carbon primary cells are the better option. I am not aware of any LSD (Low Self-Discharge) 9V rechargeable batteries or good quality chargers/analysers that support 9V batteries. A rechargeable 9V battery will be expensive and is unlikely to last longer than a couple of months.

            I often see alkaline batteries being suggested for smoke/fire alarms. Alkaline batteries will work, however you run some risk of leaking batteries damaging the alarm. Alkaline batteries are easy to get and are inexpensive, but make sure to check them once or twice a year. If you live in a state with daylight savings, checking the smoke/fire alarm batteries when it's time to change clocks (twice a year) is a good idea.

            • @peteru: Just found a good use for the cheapo multimeter: it does tell me when more than one battery is involved, whether they're all expired, or it's one of them holding back the rest. Two remote controls stopped working, each using 2 AAA batteries (these weren't rechargeable for some reason). I tested them 4 and found only two were spent. I put the other 2 on one control, and it saved me throwing all 4 in the bin.

Login or Join to leave a comment