4-1/2 Green LED DC 19.999V Digital Volt Meter - JOKE ?

[bilko]

I saw this on Ebay, quoted accuracy 0.2%, 4 1/2 digit voltmeter and 57 units sold already

    * It is suitable for the precision measurement .
    * Good linearity.
    * Small Temperature Drift.
    * Extremely high input impedance.
    * Work stable.
    * Adopts ultra-bright LED display.
    * Low Power consumption.
    * Over-range indication.
    * Decimal place is adjustable.
    * Common ground or Not common ground are available.
    * Over-range indication: "EEEE" or -EEE.
    * Easy install.

Product Name
   
EL8145A-G20V 4-1/2 Red LED Volt Meter
Power Supply
   
DC5V
Measuring range
   

DC -19.999V?+19.999V

Display Type
   

7 segment, 0.56 inch
Running current
   
<80mA
Sampling rate
   
5 readings per second
Operational temperature
   
0?+50?
Relative humidity
   
?85%RH
Accuracy
   

± 0.2% FS
Zero setting
   
Auto
Dimensions
   

8.2mm×26mm×22mm
Size of mounting hole
   

69.5mm×24mm
Meter Weight
   
30g

    * For use in

It is suitable for the precision measurement .

http://www.ebay.co.uk/itm/4-1-2-Green-LED-DC-19-999V-Digital-Volt-Meter-Excellent-/180712727326?pt=LH_DefaultDomain_0&hash=item2a1351631e

So why 4 1/2 digits when its only 0.2% quoted accuracy 

[Paul Price]

Accuracy, precision, linearity and resolution all have different meanings.

Sometimes having the ability to resolve so as to see values or small changes even in the slippery domain of an absolutely inaccurate reading, can be very helpful at times.

[rolycat]

So why 4 1/2 digits when its only 0.2% quoted accuracy 

Well, despite the Chinglish the vendor has pretty much nailed it:

It is suitable for the precision measurement

High precision can be useful even if the accuracy isn't great. The vendor also claims good linearity, low drift and stability, which suggest that the reliability of the readings will be good.

There's a big difference between precision, accuracy and reliability - i believe Dave covered this in an early video, and there are numerous explanations on the web.

[Paul Price]

Rolycat,

Not really.

Precision firstly means accuracy, the correct and  non-misleading words for the vendor to use would only be "Suitable for use for high resolution."

[c4757p]

I always learned that precision = resolution. I distinctly remember this image from high school chemistry:



Not that my high school education was worth much, but that's the meaning I still see quite often to this day.

[pickle9000]

Stability is the real issue at that resolution, I doubt it be stable over time (days weeks months) for that price.

[Paul Price]

Stability can be definitely not the issue in most cases where resolution is more important than accuracy, where some fast or tiny change might need to be noticed in a short period of time of observation.
 
I have on many occasions had the need to measure the drift or level of the tiny millivolt or microvolt offset of an opamp or adj. something to an exact setting relative to a reference.  If I was to check with high accuracy a power supply that put out 5.000000 volts and I needed to compare it to a 5-V standard, I would require only an instrument that would reveal the difference, not at all concerned with the accurate magnitude in making this galvanomic measurement.

 I would be reasonably certain, at a fixed temperature, that the readings would not drift more than a last digit or so over a few minutes or possibly even hours of observational time, even though the absolute reading maybe lacking in accuracy, but not in resolution nor short or long term stability. I only say this with some confidence because at one time in my illustrious career I was a Senior Metrologist and have spent many an hour qualifying and calibrating all types measuring equipment according to their specs.  But only hooking the instrument up to a stable, but not necessarily accurate input for some time could really verify my guess. I have not used this instrument.

[Paul Price]

C4757

What I see in your diagrams seems a little confused between precise and what could better be called repeatability of readings.
The scatter of the slightly off center hits show good repeatability and this is an important measure by itself of an instrument.

The problem here is that some dictionaries equate precision with accuracy by definition.  These two concepts should not be confused.

[rolycat]

Paul,

First you state that precision and accuracy are the same:

Precision firstly means accuracy, the correct and  non-misleading words for the vendor to use would only be "Suitable for use for high resolution."

Then you say that they are not:

The problem here is that some dictionaries equate precision with accuracy by definition.  These two concepts should not be confused.

I think you may be the one who is confused 

Part of the problem may be that there seem to be several slightly different uses of the term 'precision'.

In computer science it means the detail with which a quantity is expressed - the resolution.

In engineering or physics it means the reproducibility or repeatability of a measurement - as neatly illustrated by c4757p's diagrams.

And in common usage it is often equated with accuracy...

Just to confuse matters further, ISO5725-1 defines accuracy as being composed of both 'precision' and 'trueness', but personally I think they are full of it 

[jucole]

So why 4 1/2 digits when its only 0.2% quoted accuracy

Perhaps because it has an adjustable decimal place you could use it with a shunt;  then move the decimal place to measure currents larger than 19.999.
Or maybe you could add some resistors and attenuate the input voltage to measure higher than 19.999.


[edit]
Because it says: "It is suitable for the precision measurement ."  but doesn't give any real facts about the accuracy or chipset etc.  I'd be inclined to perhaps print-out some of c4757p targets,  then just throw these modules at the targets to see what precision or accuracy I could achieve ;-)

[T4P]

F87V has a accuracy rating of 0.05% +/-10counts on 4-1/2" mode
Does that make it a joke? OF COURSE NOT.
If it's precision you need that's what it will do, accuracy over time probably not.

[Paul Price]

RolyCat,

I am not confused, I saw this definition in a dictionary, to point out how these words become used and create confusion when using them to describe the accuracy of measurement.

[Fsck]

This is the idea I think of when I compare accuracy and precision, but I'm more scientist than engineer.

[amyk]

I wonder what type of voltage reference it uses... looks like an all-in-one ADC chip but I can't see the part number.

[bilko]

F87V has a accuracy rating of 0.05% +/-10counts on 4-1/2" mode
Does that make it a joke? OF COURSE NOT.
If it's precision you need that's what it will do, accuracy over time probably not.

Quoting 4 1/2 digits at 0.2% is a JOKE because

19.999V X.02% = .0399V

This means that the 2 LSD are nonsense, that's the joke !!

The absolute voltage accuracy of the module can be calibrated out at one particular voltage level.
No figures are given for linearity, stability or resolution.

The module could use an 8 or 10 bit AD Converter, this is again nonsense with a 4 1/2 digit meter

[bilko]

I wonder what type of voltage reference it uses... looks like an all-in-one ADC chip but I can't see the part number.

Some of the cheap meters use the A/D converter of a PIC chip or similar, its nonsense to derive 4 1/2 digit accuracy from these devices.

I think to establish any credibility for these meters, the manufacturer or seller should quote the chip used.

[rolycat]

Some of the cheap meters use the A/D converter of a PIC chip or similar, its nonsense to derive 4 1/2 digit accuracy from these devices.

As many people have pointed out, 4 1/2 digit resolution has nothing to do with accuracy, and there may be perfectly valid reasons for using such a meter.

Whether you would trust any of the specifications claimed by this seller is another question.

I think to establish any credibility for these meters, the manufacturer or seller should quote the chip used.

Credibility? From a Chinese eBay widget-shifter? Not their strongest suit, I'm afraid.

[jucole]

From the Ebay site;  if you hit the image host site with the url without the image size params, you can get the natural size of the image (slow image server!)
hk3.image.pushauction.com/ViewPicture.aspx?Key=c2f5261f-1ad6-4802-a127-60f55de3097b_f53dfb35-da9f-4e8f-8f8f-c86da27c7de1

I think it's a near all in one solution, like this one; if you do a parametric search on:  28 pin, 4 1/2 digit, 20000 count etc.
http://www.gme.cz/dokumentace/320/320-038/dsh.320-038.1.pdf


"widget-shifter" LOL- I like that! ;-)

[madires]

From the Ebay site;  if you hit the image host site with the url without the image size params, you can get the natural size of the image (slow image server!)
hk3.image.pushauction.com/ViewPicture.aspx?Key=c2f5261f-1ad6-4802-a127-60f55de3097b_f53dfb35-da9f-4e8f-8f8f-c86da27c7de1

At the top right between the orange cap and the precision trimmer there's a nice TL431 :-)