Multimeter Shootout

[EEVblog]

For those interested, the meters on their way soon are (hopefully):

$50 price category:
Extech EX320
Amprobe AM220
Global Specialties PRO-50
Elenco M-2625
VC99 ebay cheapie

$100 price category (all TRMS):
BK Precision BK2709B
Amprobe 34XR
Extech EX505
Ideal 61-342
UEi DM391

Dave.

[Simon]

would be very interesting to see the comparison of the AM220 to the VC99,  have both as main meters and paid the same for them (both ebay), I think the VC99 is cheap garbage with as many counts and features thrown in as possible, the AM220 seems to have the same capabilites but slightly less for example 4000 count versus 6000 count and the max ranges that go with that but much better quality and maybe accuracy, there is a 20 mV difference between the 2 and I think I'd trust the AM220 over the VC999

[stevestk]

I guess the one problem is trying to reach all the audience tuning in. When your're new to a hobby you don't know what you don't know. For someone getting started like me, I'd like to know what the minimum expense is that I can get away with while having a usable instrument. If I buy a $30-$50 dollar instrument what won't I be able to accomplish as apposed to a $100 meter. Or is the issue safety or longevity? Or can you tell me why I'll regret my purchase and will have to spend more money in a few months for a replacement. These same questions apply to most test equipment purchases for novices.

[Simon]

anything around £30/$50 should make a decent meter to start with just don't go lower, my £10 meters actually failed to survive a flight in a hold (in my suitcase). Personally although I've hardly used it as of yet I'd recomend the amprobe AM220, it seems to be a reliable machine and costs the same as the micky taking meters on ebay (althoug you can get the AM220 on ebay too). a £50+ fluke seems to make a good baseline meter and over 100 pounds you should get a nice durable machine

[saturation]

You did a wonderful job on the comparison sheet, but there are 3 things in it that are difficult to test.  The UT71C DC performance on the U-Tube video was impressive, but so was this:



How durable is the unit ?  How long will it stay in spec, under what conditions such as varying environmental temperatures?

How protected is it really from transients, and is it truly rated for whatever CAT level its specified for?

If you use it purely for electronics design work at home it should be fine, and CAT ratings may not be important.  In that case, I'd be more concerned with how accurate it will continue to be within the 1st year you've owned it.  Also, with these no-name meters, its hard to say if their published specs are true compared to the reputable meters or with what an independent reviewer will find out  ... like Dave will do soon.

I'm still looking for a very accurate DC meter, and thanks to you, the Uni-T models is something I'll look into ... I'd previously reviewed its portable oscilloscope but decided against it.

Dave, super idea, this multimeter shootout!

Because I was planning to buy a second, more accurate DMM, I did some "market research" and put  the specs of the products I liked in this document: Quality Autoranging TrueRMS multimeters compared.

The document is far from complete: a lot of multimeters could be added and in most cases for the accuracy of a given function (e.g. DCV) I took the best one I could find for the different DCV ranges. Also, I "calculated" the total accuracy for every function by taking the measuring accuracy and then adding the "counts" accuracy divided by the total number of counts. Dunno if this formula is any good, but I wanted something to compare the bare facts.

The Cineese Uni-T UT71C seems to be a very good multimeter (on paper anyhow). Compared to the Fluke 87 V and the Gossen Metrahit PRO, it almost beats them at every spec at a third of the price! But then again, my comparison is probably very inaccurte, incomplete, or Uni-T could be lying or incomplete? And the dial is doubtfully as sexy als the Gossen's

Anyhow, I did find a nice video that seems to suggest the Uni-T multimeters have an easy calibration feature:

Uni-T also has the UT71D and UT71E models. Apart from the number of data logging records or a power meter option, they have the same specs.

If you guys think this comparison table is any good or you have suggestions, I can give you write permission to add stuff.

[EEVblog]

The meters will be on their way on Monday, here they are at tequipment.net who are providing them!

Dave.

[PetrosA]

The meters will be on their way on Monday, here they are at tequipment.net who are providing them!

Dave.

Bravo!

[Kiriakos-GR]

I am a bit skeptic , because the 50$ limit , excludes many models , that some people all ready own.

But the positive part is , that the upcome would be an comparison of apples with apples.
And it will be quite educational ..     

For something like that , that needs time , its worth the waiting, and there is no need anyone to stress Dave about  acting faster ...

Detail and speed , does not coexist .. We have always to choose our priority s.
So I vote for detail = patience .     

[EEVblog]

I am a bit skeptic , because the 50$ limit , excludes many models , that some people all ready own.

But the positive part is , that the upcome would be an comparison of apples with apples.
And it will be quite educational ..     

For something like that , that needs time , its worth the waiting, and there is no need anyone to stress Dave about  acting faster ...

Detail and speed , does not coexist .. We have always to choose our priority s.
So I vote for detail = patience .     

Yes, the hard part was picking suitably comparable models. And no matter what gets picked, someone will complain I didn't include XXXXX
But I was essentially limited to what tequipment.net offered, as they have generously donated the gear.
I hope the $50 mark and $100 was a reasonable compromise, and it seemed to work out quite, as for example all the $100 units have TRMS, but the $50 units don't. And one of them is waterproof and shock proof! 

It'll sure take some time to review and do a tear-down of them all, so I'll have to deliberately keep it very brief with comparison tables etc.

I'll also have to do it in two separate blogs, the $50 ones and then the $100 ones.

Dave.

[saturation]

You could take your time, and do a detailed comprehensive eval in stages.

The accuracy in DMM functions versus printed specs is something we all want to know which you can give us quickly in a rundown.  But long term stability will need time, and accelerated testing will be hard to validate i.e., operation in higher ambient temps and humidity.

Then there's drop tests, waterproof etc.,

Then there's the ergonomics of actual use.

Then safety, CAT I-IV level testing.

Lots of fun for you and us!

I am a bit skeptic , because the 50$ limit , excludes many models , that some people all ready own.

But the positive part is , that the upcome would be an comparison of apples with apples.
And it will be quite educational ..     

For something like that , that needs time , its worth the waiting, and there is no need anyone to stress Dave about  acting faster ...

Detail and speed , does not coexist .. We have always to choose our priority s.
So I vote for detail = patience .     

Yes, the hard part was picking suitably comparable models. And no matter what gets picked, someone will complain I didn't include XXXXX
But I was essentially limited to what tequipment.net offered, as they have generously donated the gear.
I hope the $50 mark and $100 was a reasonable compromise, and it seemed to work out quite, as for example all the $100 units have TRMS, but the $50 units don't. And one of them is waterproof and shock proof! 

It'll sure take some time to review and do a tear-down of them all, so I'll have to deliberately keep it very brief with comparison tables etc.

I'll also have to do it in two separate blogs, the $50 ones and then the $100 ones.

Dave.

[Kiriakos-GR]

The accuracy in DMM functions versus printed specs is something we all want to know which you can give us quickly in a rundown.  But long term stability will need time, and accelerated testing will be hard to validate i.e., operation in higher ambient temps and humidity.

Then there's drop tests, waterproof etc.,

Then there's the ergonomics of actual use.

Then safety, CAT I-IV level testing.

Lots of fun for you and us!

   As independent " observer " ,

I would describe this message as a "Hit under the belt"  

Or like an side way movement,  to influence the judgment of the reviewer ..  

Everything are understandable    
This upcoming review  had throw sky high , the expectations of this community ..
soon we will need the help, of the Australian Firefighters .  

[EEVblog]

Does anyone have an Amprobe 34XR?
Just took it out of the pack and tried it and it gives a very sickly sounding beep every time the range switch is moved. Not only is it incredibly annoying, it actually sounds like the unit is faulty in some way. Almost like the battery is out of juice and it's trying to beep a last gasp (Yes, I checked that battery)
Can anyone confirm that their unit does this?

Thanks
Dave.

[Simon]

You could take your time, and do a detailed comprehensive eval in stages.

The accuracy in DMM functions versus printed specs is something we all want to know which you can give us quickly in a rundown.  But long term stability will need time, and accelerated testing will be hard to validate i.e., operation in higher ambient temps and humidity.

Then there's drop tests, waterproof etc.,

Then there's the ergonomics of actual use.

Then safety, CAT I-IV level testing.

Lots of fun for you and us!

I'm particularly interested in accuracy (I am after all a QC inspector by day) as i have two meters, the amprobe AM220 and the VC99 off ebay and they gave a slightly different voltage reading for the same voltage, naturally i trust the amprobe more

[alm]

I'm particularly interested in accuracy (I am after all a QC inspector by day) as i have two meters, the amprobe AM220 and the VC99 off ebay and they gave a slightly different voltage reading for the same voltage, naturally i trust the amprobe more

The problem with testing accuracy is that it's a statistical quantity measured over a period of time. To get a good idea of accuracy, you need to track multiple samples over a few years, not something we can expect Dave to do. The initial accuracy does say something about QC, but since I don't expect a <$100 DMM to ever be calibrated, the long term specs are the important ones.

[Simon]

as far as i know you can't "callibrate" a multimeter ? I know you can send them away for "calibration" but they do not get callibrated as such they are merely checked for accuracy.

On the contrary you can take a measurement and work out if it is withing ther tollerance stated, of course 1 year down the line you should recheck the unit to verify it is still accurate. I'd not be surprised though if many cheap multimeters are sold with poor accuracy to start with.

[Kiriakos-GR]

as far as i know you can't "callibrate" a multimeter ? I know you can send them away for "calibration" but they do not get callibrated as such they are merely checked for accuracy.

Well this Damn Fluke 87V , made me to believe that they do get calibrated ... by software.
I have read 4 times the manual , and check the service manual of the old 87 III .
Even every message in the LCD are programed, with the proper tools, you can calibrate it , or even make it to speak Russian ..  

Instead of  the { LEAD } warning , it will say {??}  

Text taken from the service manual:

After pressing {Auto Hold}, wait until the step number advances before changing
the calibrator source or turning the Meter rotary knob.
If the Meter rotary knob is not in the correct position, or if the measured
value is not within the anticipated range of the input value, the Meter emits
a double beep and will not continue to the next step.
Some adjustment steps take longer to execute than others (10 to 15
seconds). For these steps, the Meter will beep when the step is complete.
Not all steps have this feature.
8.  After the final step, the display shows "End" to indicate that the calibration
adjustment is complete. Press {Auto Hold} to go to meter mode.

[EEVblog]

I'm particularly interested in accuracy (I am after all a QC inspector by day) as i have two meters, the amprobe AM220 and the VC99 off ebay and they gave a slightly different voltage reading for the same voltage, naturally i trust the amprobe more

The problem with testing accuracy is that it's a statistical quantity measured over a period of time. To get a good idea of accuracy, you need to track multiple samples over a few years, not something we can expect Dave to do. The initial accuracy does say something about QC, but since I don't expect a <$100 DMM to ever be calibrated, the long term specs are the important ones.

Correct, absolute accuracy is not something I can do with any sense of authority with only one sample, limited time, and no thermal chamber.
Remember, most meters are only "accurate" to the 2nd LSD, so being up to 10 counts out between meters can be expected. But in practice it's usually better than that, even on the cheapies.
I'd be surprised if any of these cheapies are more than say 5 counts out from my reference Fluke out of the box.

Sadly, a lot of beginners think that if it reads the same as the Fluke then it's obviously "just as good". The truth of course is nothing like that.

Dave.

[EEVblog]

as far as i know you can't "callibrate" a multimeter ? I know you can send them away for "calibration" but they do not get callibrated as such they are merely checked for accuracy.

Essentially correct.
I've talked about this in my meter counts blog.

"adjustment" is usually a separate process that is only done if required or requested.

Modern good quality meters will have software calibration adjustment instead of old style trimpots.

Dave.

[alm]

Technically, calibration is just checking if it's within specs, although in the test & measurement industry, this is often called 'performance verification', and calibration as performance verification + adjustments. Any multimeter can be checked, and most can be adjusted. Even my father's cheap Voltcraft DMM has two internal trimmers. The more recent and expensive multimeters usually allow closed-case calibration (from the front panel or an external interface). In many cases, a cal lab will adjust anything outside specs (or something outside 50% of specs), although you can request them not to if you want to track the aging yourself.

Accuracy specs are usually valid for at most a year (assuming they have any grounding in reality to begin with), so if you care about accuracy, you should send it out for calibration/adjustment at least once a year. If you don't, you're essentially in unknown territory, accuracy-wise. Although most high-quality meters rarely go out of spec and are likely to be close to factory specs, even after more than ten years. I wouldn't expect cheap meters to have the same stability, but you never know.

[Simon]

there probably is a method of actully calibrating a meter but you must assertain first what the trst company mean by callibration, you must make sure they mean that they will recallibrate the meter, this sounds obvious but with the thousands of models around they will have to know how to work on each model and possibly have the equipment neccesary. Although I know it is not impossible I think most companies by "callibration" will mean check to see if it is in tollerance

[Rhythmtech]

i think manual re-calibration will be too tedious as the device named "multi" meter. i guess you have to calibrate everything, V,A,mA,Ohm etc. And even my cheapo one, for years still not far off enough to alert me for a re-calibration (i dont even noticed it!). Thats why they simply left that out and do it properly in the software. Just my 2cents. 

If you do 5 points for each range, you could do the calibration check in under and hour. On many modern digital meters some of the functions are in software and may not require calibration as the same hardware that has already been checked is used, although I personally would consider it bad practice to not check every range of every function. Usually when meters are "out" of calibration it is not out in every range so you only have to make one adjustment for that range.  If the meter is out across all ranges, I would suspect more than just adjustment is necessary. More like the meter is broke and needs to be replaced.

[alm]

Usually when meters are "out" of calibration it is not out in every range so you only have to make one adjustment for that range.  If the meter is out across all ranges, I would suspect more than just adjustment is necessary. More like the meter is broke and needs to be replaced.

Or it's an offset error which is equal on every range. I have an HP DMM that was off on all resistance ranges (even four wire) by something like -0.4 ohm. A dead short read -0.4 (4-wire), a 100 ohm resistor read something like 99.6 ohm (and 99.9 something on other meters), a 1k resistor was 999.6ohm, and so on. After re-calibrating the zero ohm measurement on all resistance ranges, they were all within spec again.

[Rhythmtech]

Usually when meters are "out" of calibration it is not out in every range so you only have to make one adjustment for that range.  If the meter is out across all ranges, I would suspect more than just adjustment is necessary. More like the meter is broke and needs to be replaced.

Or it's an offset error which is equal on every range. I have an HP DMM that was off on all resistance ranges (even four wire) by something like -0.4 ohm. A dead short read -0.4 (4-wire), a 100 ohm resistor read something like 99.6 ohm (and 99.9 something on other meters), a 1k resistor was 999.6ohm, and so on. After re-calibrating the zero ohm measurement on all resistance ranges, they were all within spec again.

Yeah, I agree that makes sense. I should have been more explicit.  For example, if volts, amps, and resistance all required adjustment then I would suspect something happened to the meter. In which case, you could shorten the cal interval and verify that the adjustment took or the meter needs to be replaced.

[RayJones]

there probably is a method of actully calibrating a meter but you must assertain first what the trst company mean by callibration, you must make sure they mean that they will recallibrate the meter, this sounds obvious but with the thousands of models around they will have to know how to work on each model and possibly have the equipment neccesary. Although I know it is not impossible I think most companies by "callibration" will mean check to see if it is in tollerance

This very true, especially for expensive signal generators.

For example, a calibrated unit usually means the power / frequency was within the published specs for the unit. ie 0dBm may actually really be 0.5dBm

Adjustment requires an even greater financial outlay, and you will be contacted if you desire this if your unit is found to be out of spec.

[saturation]

An alternative to calibration for DIY without access to a lab with calibrated gear is to buy a voltage reference or DMM calibrator.  It should calibrate the most critical ranges.  alm on this thread summarizes the limitations.

https://www.eevblog.com/forum/index.php?topic=601.0

Essentially you buy a lab's voltage, current and resistance reference board.  This is calibrated by the seller.  Every 6-12 mo you can return the reference board to them for recalibration, and they mail it back to you, ~ cost $5 US after 1 year, free prior.  I estimate total cost would run $10-15, mostly from the postage.

I think Dave said once you are aware of the drift of a particular DMM, calibration may show that its rock stable year after year, and annual checkups become 'less' or unnecessary, unless you are in a professional setting.  If you own several DMM, you can calibrate them at different times of the year, so as they drift, you can cross reference them against each other to check accuracy.

On the side, Fluke DMM owners know them be quite stable, something not immediately viewable by its appearance or mentioned in its manual or sales literature; a measurement of some voltage references show my 85 still reads true to spec 20 years since its last recalibration.

Here's a fellow describing his uncalibrated 10 year old Flukes:

http://www.cappels.org/dproj/1.024voltreference/portvoltref.html