Which Meter. Extech EX330 or Amprobe AM240 ?

If you have a 4000 count meter and want to check the accuracy on the lower voltage scales you DO need another reference. 5V will cause the meter to switch to the next higher range.

My meters are only 2000 counts.

I have been thinking of another way to get more accurate readings … My reference is known to be 5.0002V … If I connect my reference positive to my battery positive and measure between the reference negative and the battery negative with a high impedance meter , the reading with a fully charged 18650 would be about 0.800V … If I subtracted the 0.800V (or whatever) from the 5.000V then that should give me a more accurate reading than I get at the moment … I realise that this is a bit long-winded , but it would cost nothing extra to do.

I might not bother … I think that two places of decimals will be OK for checking my Li-Ion voltages.

Maybe one day I will justify a better meter.
.

I recommend the HP3458A. And for a better reference a Fluke 732 or 734. Add a Fluke 720A Kelvin-Varley divider to it to get any voltage you want (< 10V).

But then, that brings up the question of how do you know it is all working properly? A man with one meter never knows what volt it is. A man with two meters can know something is wrong. Better get three of each… I did.

Unless this is necessary for some other enterprise, a super-accurate DMM to measure battery voltage is entirely pointless. The Vout is variable depending on load, so any crapo-meter with a “battery” mode (like the $3 harborfreight ones, which btw can be calibrated inside) are going to be more “accurate” than a fluke.

It’s also why getting thick leads are pointless for measuring current, too. The thin wire spring inside a tailswitch is going to offer far more resistance than any DMM probe.

Nope, the thin wire spring is very short and has a low resistance compared to long cheap meter leads (which are almost always made of some thin stranded non-copper mystery-metal wire). But the real killer in measuring current is usually the current shunt resistor in the meter.

The leads on the $3 meter have too low resistance to meaningfully measure with a meter. What’s the shunt resistor value on DMMs?

On a LiIon you uses the unloaded voltage, that makes a DMM perfect for the job.

For everyday use I do agree that a 4.xx reading is enough, but the tolerance on that reading is fairly high on some DMM's.

Not all tail springs are thing and long, on the more powerful lights they are usual with thicker wire.

Yes, it requires another method, you can see one in my DMM guide.

That varies with range, usual you can expect between 0.3 and 1 volt burden voltage when the current is maximum for that range. Unfused cheap DMM's can be than the more expensive DMM's.

With auto ranging DMM, you might get a lover voltage drop in the low uA and low mA range, because they uses the same resistor as the high uA and mA range.

I.e. the high uA/mA has 0.3 to 1 volt burden, the low uA/mA has 1/10 of that.

It’s certainly not on the cheapest possible DMM available on the mass market today, so I can’t imagine it would be on a more expensive one.

Neither is going to be that meaningful for the 1A or some measurements we do.

[quote=agenthex]
It’s certainly not on the cheapest possible DMM available on the mass market today, so I can’t imagine it would be on a more expensive one.

[quote]

I am going on the specifications, not on a random test with a reference.

A 0.5%+3 specification means that 4.20 volt can be shown as anything from 4.15 to 4.25 with a extra digit the same specification will show 4.176 to 4.224

[quote=HKJ]

[quote=agenthex]
It’s certainly not on the cheapest possible DMM available on the mass market today, so I can’t imagine it would be on a more expensive one.

That as much as spec of accuracy as precision. I’ve never seen any meter be that imprecise.

That is a fairly typical specification for a low end DMM, most meters are better indoor at 25 degrees.

A fairly good DMM (Fluke 179) has 0.1%+3, for 4.2 volt that gives 4.194 to 4.206 (This specification is between 18 and 28 degrees).

[quote=HKJ]

[quote=agenthex]
It’s certainly not on the cheapest possible DMM available on the mass market today, so I can’t imagine it would be on a more expensive one.

My meters are very cheap ones (Sinometer M-830B) and are not auto-ranging … I have to use the 20V range for my Li-Ion battery measurements … The “accuracy” on the 20V range is quoted as + /- 0.5% of the reading +/- 2 digits … My voltage reference is 5.0002V at 72F and my meters actually read 5.00V … On a 2000 count meter , it isn’t possible to get any more accurate than this … I realise that the the actual voltage could be between 4.995V and 5.005V which would show on my meter as 5.00V … Being up to 5mV out , in the measurement of an off-load Li-Ion isn’t too bad really.

I was hoping that for up to £50 I would be able to get a meter that would read exactly 5.000V , but it seems that this is not possible in my low price range without tweaking an adjustment potentiometer (on a VC99) … It seems that perhaps I should just stay with my cheap meters as I can’t justify the expenditure needed to get the 5.000V accuracy when measuring my 5.0002V reference.

Do we need this degree accuracy (4.200V) for just checking Li-Ions ? … Perhaps not.
.

Yes, I know what the spec is, but context for why/how that’s meaningful is also relevant. 0.5% or whatever here doesn’t mean that for each arbitrary measurement of 4.20v DUT, the result will sometimes be 4.18 and then 4.22 a sec later.

In any, this isn’t particularly meaningful given that far more expensive instruments in far more professional settings are far less precise+inaccurate (often inherently, eg high freq sampling).

Similarly, I can’t imagine that the cheapo spring + various shitty contact points inside those typical 5cent switches in budget lights are as decent as the 18awg wire + stiff jacks on even the cheapest of meters.

Maybe if we’re working on the mars rover the last decimal point of accuracy would matter, but this is hardly it. To provide more context, not even the voltages inside intel chips running at >>1ghz are controlled to anywhere near that type of precision.

Getting a DMM that is guaranteed to show 5.000 is very expensive. My Fluke 289 (One of the best handheld DMM's) has 0.025%+2, with 5.0000 volt that would show 4.9988 to 5.0012 (Out of the box it will be better than that).

Going all the way to a 6½ digit (Fluke 8846A) meter you would get: 4.99985 to 5.00015 (With one year calibration).

These are not in my £50 budget unfortunately … However , a (tweaked) VC99 would be.

It all depends on how much accuracy we really need in measuring a Li-Ion battery !

I’m beginning to think that my existing meters are OK for my purpose.
.

They are a bit expensive!

There has never been a reason to go for 4.200 volt, the specification typical says 4.15 to 4.25 volt charge voltage (Not battery voltage!).

I can quite easily measure the resistance of under 1 inch of meter lead… it just takes the proper equipment.

Most cheap meters have at least 0.1 ohm shunts on the high amp ranges. The problem with measuring tail cap current with them is the voltage drop across that shunt (ohms * volts) is a very significant percentage of the battery voltage. Couple that voltage drop with the extreme variation of LED current with changes in voltage and you can easily get a huge (like 50%) error in LED current without the meter burden resistance. The total shunt resistance in my LED analyzer (Any interest in a LED/Battery analyzer device?) is around 2 milliohms (50 times lower) than a typical meter.

Is that 0.1ohm for the 10A range?

Do you measure significantly lower than 0.1ohm for budget light tailcaps?

I would just like to have a meter that would read 4.???V … I already have one that is great for 4.??V … I realise that it is not essential to go for an extra digit when measuring charger or battery volts, but I would like one … The Flukes etc are a lot more expense than I can justify for taking low voltage battery voltages … It would be nice (but not essential) to have a better meter.

Maybe one day ……………