Precision in different multimeter readings...

[Externet]

Greetings.
Readings of a battery showed 14.4V, 13.9V, 15.5V, 14.7V with four different high impedance 'modern' digital voltmeters for general use (not professional bench instruments).
Probably other scales and measurements are also imprecise.

How to discern which reading - instrument is more reliable as I think cannot be user calibrated -never seen a manual with calibration instructions-
Any way to make/obtain a home-made decent voltage reference ? Would a fresh-new carbon-zinc cell be decent ? What for higher voltage scales ?
How do you and how often check accuracy of your daily use multimeters ?

Checking against a decent oscilloscope falls also to the same uncertainty of such reference being precise. What % of error is reasonable to accept ? Sometimes I charge lithium cells and prefer to avoid errors as their maximum voltage is not very forgiving.

For resistance measurements, perhaps there is a reference kit in the market you prefer to use ? What would you get without breaking the wallet ?

 

[D.A.(Tony)Stewart]

It depends on your test method. Is it static DC only, AC ripple will add error.

the scope has a calibration ref. no? And specs... read them

If the DMM's have no specs or manual, then not very reliable.
I would expect <1% error

 

[danadakk]

Precision V references :

https://www.analog.com/media/en/technical-documentation/data-sheets/1021fc.pdf

https://www.ti.com/lit/pdf/SLYT183

https://www.youtube.com/results?search_query=diy+voltage+reference

Regards, Dana.

 

[Easy peasy]

You don't state how many seconds between readings ! also any ripple will affect a non true rms meter.

 

[D.A.(Tony)Stewart]

If there is radiated RF, twist the leads to see if it changes.
Show your DMM's and what did you measure?
car?

 

[BradtheRad]

Certain IC instructions told me about built-in reference voltages. Example, 3914 (popular bargraph IC) makes a default volt separation between each level. Example, TL431 produces its own 2.495V reference.

I used the above when I wished to calibrate my meters. A year or two ago I picked up a Craftsman DMM on sale at Sears. I compared its new readings with my other meters. The differences were slight, so I'm happy.

For as long as I remember a 12V car battery (according to electronics articles) in good health is really 12.6V after resting half a warm day. Supposedly your meter can read 12.4-12.8 and still be within two percent.

 

[KlausST]

Hi,

Readings of a battery showed 14.4V, 13.9V, 15.5V, 14.7V with four different high impedance 'modern' digital voltmeters for general use (not professional bench instruments).
Probably other scales and measurements are also imprecise.

Terminology: this is not a precision problem, this is an accuracy problem.
Accuracy error: is how much you are away from the true value. (usually given in "% FS, plus offset")
Precision error: = repeatability error. Do the same measurement multiple times with the same equippment and see how much one value differs from the other.
(usually given in "+/- some mV")

But here - as I understand - you want to know the real, absolute voltage of a battery.

--> For this you need to be sure the voltage of the battery is stable, and not influenced by other devices. Best: do not connected to anything else than your voltmeter.
(no charger, no load ....)

You don't state how many seconds between readings ! also any ripple will affect a non true rms meter.

The batteries are DC .. so use the voltmeter in DC mode. (= average mode, or low pass filtered)
Don´t use AC mode.
Don´t use true RMS mode, since this is for AC measurements. Many DVMs in trueRMS mode only show the AC components (no DC at all), so only values in the mV region on a (let´s say 12V) battery.

***

Time between measuremets: Batteries (alone) don´t have ripple. (Only ignorable noise caused by the chemistry)
But the voltage may drift:
* with temperature (very slow)
* after a charging process (for some minutes) --> wait 5 minutes after charging
* after a discharging process (for some minutes) --> wait 5 minutes after discharging
* during and a bit after mechanical movement (moving fluid inside, bubbles..) --> wait 5 minutes after moving

***
I´m currently running a measurement during charging/discharging batteries. Big truck batteries, for currents up to 2000A. Load and charging current is only 16A and 6A respectively. My measurement equippment gives an "average value over 20ms" .. every 20ms. Then every second the min and max of the past 50 values are calculated.
Even during charging/discharging the (50) values are precise to +/- 2mV (on a 24V nominal battery voltage).

Any way to make/obtain a home-made decent voltage reference ? Would a fresh-new carbon-zinc cell be decent ? What for higher voltage scales ?
How do you and how often check accuracy of your daily use multimeters ?

REF: I have a reference voltage IC for 5.000V used to check my meters. (For sure

Meters:
* I have an expensive one, reliable, just for the lab.
* And a FLUKE handheld.
* I have a cheap one for coarse measurement on a construction site.
* And I have my own designed ones (Especially designed for fast measuremnts on AC loads. With USB interface). They are as stable as the expensive one .. and much better than the Fluke one - especially at low AC and DC voltages.

(I think about designing my own AC reference. 3 phase one with let´s say 5.000V RMS pure sine 50/60Hz. With fixed 120° phase shift and an adjustable phase shift. A unique design ensures very accurate and precise voltage as well as phase. Anyone interested: please contact me).

Checking against a decent oscilloscope

Most scopes are not accurate ... at least not in the meaning of a reference.

For resistance measurements,

You need to provide informations. There are good circuits for high resistance values ... and there are other good circuis for low resistance.

--> Expected measurement range. Expected accuracy? 2Wire, 3W, 4W? Battery powered? Display or with interface?

*****
But back to your batteries:
Absolute voltage measurement does not give much information about a battery. The absolutoe volatge depends on a alot of parameters, like production, chemistry, age (health), charging state, temperature,
--> if you get an accurate voltage of a battery - let´s say 12.34V ..it tells about nothing.

Accurate voltage measurement is only useful for special situations: Like "end of charging" detection on LiIon batteries. During constant_current charging at room temperature.

Or to see the the difference before and after a high current pulse (which can be healthy for some chemistry) regarding V to charge dependency.

Klaus

 

[Externet]

Thanks.
Unloaded battery measured is a series of 12cell NiMH presenting different readings on a variety of multimeters, prompting this thread.

Perhaps a simple kit as this can improve certainty to a reasonable level, unless you suggest other; but is of no help for voltages around 100VDC.

 

[barry]

You don't state how many seconds between readings ! also any ripple will affect a non true rms meter.

Ripple on a battery?

 

[D.A.(Tony)Stewart]

I have an old HP3748A 6 digit multimeter if I ever need accuracy, precision, A handheld Fluke and a couple other DMMs and an old analog Simpon meter good for 200 mV FSD and collecting dust. But I like my Chinese 65W digital power supply with 3 displays V, A, W that can measure external V with 4 digits and CC digital tuning, so I can charge any battery I use and compute ESR, kF easily.

 

[dick_freebird]

You can order sub-% resistors from distributors.
So too voltage references. Some, but fewer I.
Major brands are trustworthy.

Much more trustworthy than pile of crusty
you've got there, and methods?

Number one, hook them all up at the same
time to the same source and same return
points. Observe readings for variation
over time. Put data in order and report.

 

[Easy peasy]

Ripple on a battery?

Yup - for all we know the battery was in circuit and being charged ( 100/120 Hz ) or discharged into a pulsy load - one never knows with random posts from non educated posters.

 

[Externet]

Yup - for all we know the battery was in circuit and being charged ( 100/120 Hz ) or discharged into a pulsy load - one never knows with random posts from non educated posters.

Agreed. Non-educated posters that cannot even read should be banned, leaving the forum only for geniuses.

 

[D.A.(Tony)Stewart]

Pulsed battery charge ripple is the best way to measure ESR while charging at the same time with a trickle + power pulse charger. It's is proven to improve rapid charge with lower heat rise. It is also called Load regulation error= dV/V max *100% and dV/dI=ESR nom. which varies with T, and SOC. Leave at least 10 % SOC margin to increase longevity, preferably 20% margin at each end.

--- Updated Yesterday at 3:56 PM ---

Agreed. Non-educated posters that cannot even read should be banned, leaving the forum only for geniuses.

Users ought to show experience history in profile to rank answers. Yet everyone makes misteaks.