Poll

How many cycles will the KeySight U1281A's detent spring last?

0-2000
7 (17.9%)
2k-4k
5 (12.8%)
4k-8k
14 (35.9%)
8k-16k
7 (17.9%)
>16k (most rubust meter ever made)
6 (15.4%)

Total Members Voted: 36

Author Topic: Handheld meter robustness testing  (Read 1149440 times)

0 Members and 4 Guests are viewing this topic.

Online tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Fluke 15B is the meter I'd like to see play with the evil lady.
I can get one to Joe for USD75 and happily pay half.
Any takers?

15B, 17B same internals.

I was thinking the same thing (but one of the newer 15B+ and 17B+ versions...)

Also ... one of those old-school Fluke 27s. They're supposed to be pretty tough but I wonder if they'd hold up or not.
Nobody but Fungus seems to want to play.  :--

But we must ask Joe first if he wants to be inundated with meters to fry.  >:D
 :popcorn:

Joe?


This is the one I had in mind and Fungus, it is the latest plus model:
http://www.aliexpress.com/store/product/Fluke-15B-F15B-Digital-Multimeter-Auto-Manual-Range-AC-DC-1000V-10A-40M-ohm-Capacitance-100uF/917544_1187715988.html
I know this seller and he may want to play along,  :-\ especially if he can link this thread to his products.  :-DMM

If people are really interested in testing more meters, maybe we could do a Paypal account or kick start for each meter people want to run.
Paypal would be OK

We need to know how many people are interested though. Let's have a show of hands...
Paypal fine with me and I'll put some money where my mouth is.  :D

Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Nobody but Fungus seems to want to play.  :--

Give it a chance...it's only been a few hours.

As a reference: The last video in this series got 64 views.
« Last Edit: October 20, 2015, 01:33:13 pm by Fungus »
 

Offline Lightages

  • Supporter
  • ****
  • Posts: 4313
  • Country: ca
  • Canadian po
I said I was interested, I kind of started the idea no?

A Fluke 15B+ would be good, as would a UT71D, UT171B, a Keysight, Hioki, Gossen, Yokogawa, and Ideal.
 

Offline Wytnucls

  • Super Contributor
  • ***
  • Posts: 3045
  • Country: be
Planning to willfully damage a whole bunch of mainly very safe and expensive meters. And this to prove what again? You guys have gone mental.  ::)
Thankfully, not too many takers.

It would be far more useful (and cheaper) for everybody, if meters with suspected bogus CAT ratings, were subjected to transients in accordance with the IEC regulations, to separate acceptable meters from the dangerous ones.
« Last Edit: October 20, 2015, 05:55:06 pm by Wytnucls »
 

Offline Lightages

  • Supporter
  • ****
  • Posts: 4313
  • Country: ca
  • Canadian po
Mental?

I think that Joe is not mental, nor was 5KY. I might be.

The point is more data on where a multimeter fails? Maybe to educate and illuminate the differences between protection schemes and how a specific multimeter might be a better buy in the case of error in use? How a bottom of the line multimeter might be better suited to a newbie who could make some mistakes instead of a higher end one? Have we learned nothing from Joe's tests?

Wouldn't it be interesting if a Gossen couldn't take 1000V on the ohms range for example?
 

Offline Wytnucls

  • Super Contributor
  • ***
  • Posts: 3045
  • Country: be
You stand more chances of getting pregnant than experiencing high transients on the Ohms range. That's why meters aren't built to withstand that abuse. 1000V is the maximum required and, like all Gossen DMMs, have already been tested to comply with their CAT rating.
Multimeter protection is not rocket science. Most high-end multimeters have the same kind of layout. PTC, MOV and crowbar circuit. As expected, they will fail safely eventually, if pushed above their safety rating.
The message to beginners should be to only acquire meters which have been tested by approved independent agencies.
There is no need to muddy the waters with indiscriminate destruction on all ranges.


« Last Edit: October 20, 2015, 11:32:41 pm by Wytnucls »
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
If you haven't seen it yet, DextersLab2013 posted a video using his new generator.    Looks like it does a nice job on them. 



Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Planning to willfully damage a whole bunch of mainly very safe and expensive meters. And this to prove what again? You guys have gone mental.  ::)
Maybe...but if it's only a few $$$ then it's entertainment. People go to the cinema every day to watch them wreck nice cars, etc. :popcorn:

 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
You stand more chances of getting pregnant than experiencing high transients on the Ohms range.
The protection isn't just there for transients, it's there for when operators make mistakes.

Would you lend somebody your Fluke 87 at an electronics club meeting if you could lend them a 107 instead?
 

Offline Wytnucls

  • Super Contributor
  • ***
  • Posts: 3045
  • Country: be
Well, the mistake is to test resistance on a live circuit, hence the 1000V protection. You may have to wait a few months before you get any transients.
If you play with higher voltages without a high voltage probe, you deserve that Darwin award.
As for the entertainment value, watching grass grow comes to mind. Maybe if you happen to live on the Atacama plateau:

« Last Edit: October 21, 2015, 09:13:28 am by Wytnucls »
 

Offline Lightages

  • Supporter
  • ****
  • Posts: 4313
  • Country: ca
  • Canadian po
OK, Wytnucls, keep baiting. I am going to ignore you from now on. I just thought perhaps we had a different point of view, but you have now proven you are trying to troll me.
 

Offline Wytnucls

  • Super Contributor
  • ***
  • Posts: 3045
  • Country: be
Stop taking everything so seriously Personal information of another member removed by moderator, please "don't do that"!, that was meant as a teasing joke. I'm surprised you reacted that way because my Canadian friends usually have a good sense of humor.
I'm really sad you're going to ignore me. I enjoyed the few multimeter skirmishes we've had so far.
I'll get so bored and lonely now that I might watch a few Gossen DMMs getting blown up.
« Last Edit: October 21, 2015, 05:36:03 pm by Simon »
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Wytnucls,  I have no problems with you trolling this thread. 

It would be far more useful (and cheaper) for everybody, if meters with suspected bogus CAT ratings, were subjected to transients in accordance with the IEC regulations, to separate acceptable meters from the dangerous ones.

But again, while I had thought about using a standard transient generator to run the meters using the standard IEC waveforms,  it was never my intent to validate CAT ratings.   Forget that for now and lets talk about the "and cheaper" comment.   I have a fair amount of experience using outside labs for testing although it has been several years.   What does lab time cost there?   Do you pay by the day, hour, job?    I would also be very interested in how much time you feel it would take to validate one meter to the IEC standards?    It sounds like you have a lot of experience in this area, so please fill us in.


The protection isn't just there for transients, it's there for when operators make mistakes.
...
Would you lend somebody your Fluke 87 at an electronics club meeting if you could lend them a 107 instead?

That 107 is very hardened and I would have no problems at all about giving one to someone as a basic meter.   

I can see from Wytnucls perspective why none of this makes any sense.   If all you are doing with your time besides trolling is looking up specs for meters (paperwork) and putting all of this information into a spreadsheet (more paperwork) , you will never damage a meter.  Safety is still a huge concern for these people as I understand some paper cuts can get infected if not treated.    These people will never make any mistakes in the lab other than maybe forgetting to change their meter's battery.   


I said I was interested, I kind of started the idea no?

A Fluke 15B+ would be good, as would a UT71D, UT171B, a Keysight, Hioki, Gossen, Yokogawa, and Ideal.

Sorry, but I see no point in testing another UNI-T.   None of the four tested survived any where near what the majority of meters tested did.   The only good use I have found for one was this and really, a cheap set of ear buds did much better.  :-DD   





Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
The protection isn't just there for transients, it's there for when operators make mistakes.
...
Would you lend somebody your Fluke 87 at an electronics club meeting if you could lend them a 107 instead?

That 107 is very hardened and I would have no problems at all about giving one to someone as a basic meter.   

I also like the fact that there's no 250mA fuse that costs $12+shipping to replace.

Not one person had a working fuse on the 250mA range last time I did a 'bring your multimeter' day at the club.

If they blow the 10A fuse, well ... something else was probably going to blow anyway. The fuse might pay for itself by saving that thing.

« Last Edit: October 21, 2015, 11:53:53 am by Fungus »
 

Offline Wytnucls

  • Super Contributor
  • ***
  • Posts: 3045
  • Country: be
Joe, you're not interested in IEC regulations testing. I get that. Not much fun as meters don't fail so spectacularly under those. If you change your mind, we can discuss it constructively, like I did for a couple of years with Lightages, Dave and others.
By cheaper, I meant acquiring cheaper meters as the expensive ones have a proper CAT rating already. There is no point retesting a Brymen CAT IV 1000V meter which has a UL listing. Your testing wouldn't validate the CAT rating by any means, but offer some valid guidelines for people who can't afford meters with recognized CAT ratings.
I have damaged a couple of meters actually, but that wasn't by playing carelessly with high voltages. I don't pretend to have your level of knowledge and skills, but paper cuts still hurt!
As for trolling, if it means not agreeing with someone publicly, then I guess I am trolling. Live with it.
« Last Edit: October 21, 2015, 11:14:23 am by Wytnucls »
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Actually, I am very interested in IEC regulations testing but not when it comes to handheld meters.     Under those standards the meters would fail much more spectacularly than what I show.    But that's ok,  I understand, between playing with spreadsheets, reading meter specs and trolling all day gives you little time to do anything else.  And that's fine. 

Yea, I figured you had no idea what it took to test a meter under IEC or what the lab costs would be.  Actually, that's too bad as I was hoping to see how they would compare around the world.   Maybe it's something you could research rather than trolling?  This way you can still spend time reading specs and playing with spreadsheets.   

Quote
There is no point retesting a Brymen CAT IV 1000V meter which has a UL listing. Your testing wouldn't validate the CAT rating by any means, but offer some valid guidelines for people who can't afford meters with recognized CAT ratings.

I don't see how anyone would ever draw a conclusion about a meters CAT rating from my testing.   Maybe you could explain this?    Most of my testing is under 10J.  The video from Dexterslab was 250ish?    Again, because you are special, I will repeat myself yet again,  I am just looking for robust meters.     If you have other test interests, you could always step to the plate and take a swing.   Keep in mind that playing with real hardware can be addictive!   :-DD

As for trolling, if it means not agreeing with someone publicly, then I guess I am trolling. Live with it.

I'm surprised you reacted that way because my (fill in the blank) friends usually have a good sense of humor. :-DD  Hang in there.  I can't afford to damage too many more Flukes.

Offline Wytnucls

  • Super Contributor
  • ***
  • Posts: 3045
  • Country: be
OK, Joe, keep baiting. I am going to ignore you from now on. I just thought perhaps we had a different point of view, but you have now proven you are trying to troll me.  :-DD




 

Offline Wytnucls

  • Super Contributor
  • ***
  • Posts: 3045
  • Country: be
The protection isn't just there for transients, it's there for when operators make mistakes.
...
Would you lend somebody your Fluke 87 at an electronics club meeting if you could lend them a 107 instead?

That 107 is very hardened and I would have no problems at all about giving one to someone as a basic meter.   

I also like the fact that there's no 250mA fuse that costs $12+shipping to replace.

Not one person had a working fuse on the 250mA range last time I did a 'bring your multimeter' day at the club.

If they blow the 10A fuse, well ... something else was probably going to blow anyway. The fuse might pay for itself by saving that thing.

It is useful. Some meters have a polyswitch instead to protect the mA range. The 107 doesn't seem to have one. Just as well, as they tend to fail if subjected to high voltages.
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
OK, Joe, keep baiting. I am going to ignore you from now on. I just thought perhaps we had a different point of view, but you have now proven you are trying to troll me.  :-DD
:-DD

Lightages, no more Mastechs as well.  I'm not sure how many we have checked now with the re-branding BS but like the UNI-Ts they never seem to hold up very well.   

Offline Lightages

  • Supporter
  • ****
  • Posts: 4313
  • Country: ca
  • Canadian po
Yeah, I have yet to see a Mastech that I would buy at any price,let alone recommend even they would pass any kind of test. They all seem to be very poorly made. The Uni-Ts were just ideas to show the difference, if any between the old and new high end models.

I am still up for donating towards the testing of other meters.
 

Offline saturation

  • Super Contributor
  • ***
  • Posts: 4787
  • Country: us
  • Doveryai, no proveryai
    • NIST
Great job Joe. :clap:
Best Wishes,

 Saturation
 

Offline dexters_lab

  • Supporter
  • ****
  • Posts: 1890
  • Country: gb
If you haven't seen it yet, DextersLab2013 posted a video using his new generator.    Looks like it does a nice job on them. 

 :-+ >:D

thats just shy of 1700v, 185uF

it's a bit of a sledgehammer in comparison to your generator Joe, but it's good to show how well the blast protection might work  :-DD

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
That 107 is very hardened and I would have no problems at all about giving one to someone as a basic meter.   
I also like the fact that there's no 250mA fuse that costs $12+shipping to replace.
It is useful.

Sure, but very delicate.

And in Fluke's case, very expensive.
 

Offline saturation

  • Super Contributor
  • ***
  • Posts: 4787
  • Country: us
  • Doveryai, no proveryai
    • NIST
Of concern to this thread is the sample size of n =1 for each meter and any generalizations to the rest of the production runs of each meter.

If a manufacturer does tight quality control, and Fluke makes public it follows a form of six sigma manufacturing, then the probability of product defect hover about 3 ppb. 

Thus, a single test device can be representative of the entire population, as any product in this batch is expected to be effectively defect free and made to perform as designed.  You could say the chance the single device doesn't represent the population its from is ~ 3 ppb.

Note, this holds only for a manufacturer that has a strict QC method like six sigma.  So for the other meters where the quality control is unknown, one cannot not say they failed or survived due to sample bias, or chance, a 'fluke'  :palm: ;).
« Last Edit: October 21, 2015, 10:12:00 pm by saturation »
Best Wishes,

 Saturation
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Saturation,

I believe what you wrote to be true as well and had posted on this subject a few pages back in an attempt to address mtdoc's comments.  See below.   Thanks for posting.   


Interesting and entertaining videos.  Well done.

That said however, I would hardly call a one-off test of this sort any definitive test of a meter's robustness. 

Sure, if the Fluke 87 had failed spectacularly at a relatively low voltage that might say something, but that's not what happened.

If 100 fluke 87s had been compared to 100 101s across a variety of controlled tests then you might be able to make a valid conclusion about their relative "robustness".

I do think the Fluke 87V is a bit overpriced at $400 compared to its competitors but that's a different issue.

In the end, any test with an n=1 is just anecdotal in nature and not evidence. But it does make for fun viewing an interesting discussion.

Any test engineer or working scientist knows that you can't draw any meaningful conclusion from testing with an n=1 (or 2) - so all this teeth gnashing about exactly what these tests ultimately prove is misplaced IMHO.  They are certainly interesting and entertaining and make for a good discussion. I say well done Joe! (even though I know he has not tolerated my critique in the past).  Joe's biases come through but that is not a criticism - we all have our biases.  Any strongly held opinions about the  87V based on these tests are unfounded IMO but that's ok - we all have opinions..


In the first series of meters I mentioned that I had only tested one of each meter and that this was not much of a sample size.  Obviously, we are never going to be looking at large same sizes for any of these reviews.    Many benchmarks are performed with minimal sample sizes and in some cases, even one sample  but it does not mean they are completely invalid or that we can not learn anything from them.   Statisticians are rolling there eyes now..... 

Let me start by saying that these products were obtained through normal channels.  In most cases they were procured through Amazon.   Why does this matter?  Well, if say the manufacture sent me meters directly for these tests, how would we know that these products were not special in some way in order to bias the tests?  So to be clear, in no case did Fluke or Brymen supply me with product for these reviews.   This is what I don't like about regulatory groups.  The companies ship the products to be evaluated.  Many times these may not be the final production parts.   They may need to make some changes and go back and forth a few times before they are certified.    In the end, do we really know the product that was certified is what will be supplied to you and me?  Or, will a MOV be removed to increase profits by some accountant.   Maybe a part was changed out for a cheaper part of what they think is the same quality and the product was not re-certified because of cost, time, etc.     So, because I obtain the meters my some means that the average person could,  I am making some assumption that the meters I test represent the average meter.     

Now this makes for another assumption.  I assume that the manufacturer has their process under control.    They may not and then my first assumption that my one meter represent the mean goes out the window.      So we could say for example, the 87V I tested does not represent the average 87V.  It was some outlier.  Then we also say that Fluke does not have control of their process.    Now I doubt they have a process control problem, but I don't know.  I would more guess that the meter just does not handle the transient by design in these other modes besides voltage.   It's a pretty old design when compared with the 101.  I am sure they have learned a few things and have improved their designs.   I have no data to back that up and am just giving them the benefit of the doubt. 

In the case of the 101 being the only meter to survive my first round of testing, we had a member repeat these tests using a different meter and commercially available transient generator.    The results were the same.   Even at 12KV they could not damage the 101.   I went further and increased the FWHH and added 1KV and still could not damage it.    I have some level of confidence that the 101 is very robust.   That said, I was not too surprised that the 107 survived this same test.     

Now had say the Fluke 87V failed at 2KV and the Brymen BM869s at 2.5KV, I would say we are well withing the margin of error of my tests and the meters them selves (what brand of components, date codes, etc).  But this is not at all what happened.   We have one meter living at 6KV and one failing at 1.5KV.    That's a pretty big window.   I bet if I tested 100 pcs of each meter we could find some 87Vs that would live to 1.7KV and some Brymens that fail at 5.5KV.      If the windows were much wider, I would really question their process and quality control. 

The problem in gaining confidence in the tests is that we are no longer talking about low cost meters.   I doubt that our members are going to run out and buy an 87V and BM869s knowing the 87V may be damaged at 1.5KV just to repeat the test and see which is more robust.     

I don't believe I have skewed the results or biased the test towards one brand or another.     Again, take it for what it is worth or feel free to step up to the plate and take a swing.   I am open to what ever tests the group can come up with to help determine which meters are more electrically robust than others.   





Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf