Handheld meter robustness testing

[HKJ]

I think I get it.  So you keep changing the pulse time until you find the spot where the meter is stable, triggering each time.   To get any sort of precision it must take a very long time to run.   If you had a way to monitor the meter (HS camera) one long pulse and you would have it.

It takes a few minutes, precision is not really an issue (2.5s or 2.53s do not matter).

30fps or 33ms would do it.  Basically any cell phone should handle it.  One cycle would get you there.  Maybe a counter in the background for the time reference.

I am not sure it would be faster that way.
And as I have already said: The measure time varies (At least for some DMM's).

[joeqsmith]

I am not sure it would be faster that way.
And as I have already said: The measure time varies (At least for some DMM's).

Does seem like with your eyes being used for the  feedback, 30ms error would be conservative.   Maybe not.   Camera would let you run more than one meter at a time plus better precision?    What meters do you have that vary? 

[joeqsmith]

Using a 10uF, even at 250fps, it's a bit difficult to tell what you would call settling with the BM789.     I think you would need to define how you measure it.  Is it when the display first shows the correct answer even though it may not be readable with the human eye, or when the display is actually fully turned on.  It takes about 100mS just for the display to settle after showing the correct number.   I guess it depends what precise means to you.   

Personally, just seeing how a few of them compare side by side is gives a good enough indicator.     

[HKJ]

I am not sure it would be faster that way.
And as I have already said: The measure time varies (At least for some DMM's).

Does seem like with your eyes being used for the  feedback, 30ms error would be conservative.   Maybe not.   Camera would let you run more than one meter at a time plus better precision?    What meters do you have that vary?

The timing is the pulse width before the meter can read the value each time, the time the meter needs to display the value is not included.

I have not noted what meters varies, I simply adjust the time until I get a reading each cycle. With ohm the variance is usually small, but when doing continuity it easily triple (or more) the reaction time.

[joeqsmith]

I am not sure it would be faster that way.
And as I have already said: The measure time varies (At least for some DMM's).

Does seem like with your eyes being used for the  feedback, 30ms error would be conservative.   Maybe not.   Camera would let you run more than one meter at a time plus better precision?    What meters do you have that vary?

The timing is the pulse width before the meter can read the value each time, the time the meter needs to display the value is not included.

I have not noted what meters varies, I simply adjust the time until I get a reading each cycle. With ohm the variance is usually small, but when doing continuity it easily triple (or more) the reaction time.

Your using the display for visual feedback so of course the time for the meter needs to display the value is included.  But the time that your eyes will detect it seems like it can vary depending on you.  The time for the 789 to show the value dimly on the display vs fully on is about 100ms.   

I assume what you are calling reaction time is the time from when the load is applied across the meter to when the meter displays the value and you are able to see it.  With the continuity being so much faster then when measuring resistance, even if the variance is the exact same time, we would expect the percentage to be much higher for continuity. 

I looked over your blog and couldn't make sense of it. The video is nice as the viewers can easily see how the meter compares side by side.   

[HKJ]

Your using the display for visual feedback so of course the time for the meter needs to display the value is included.  But the time that your eyes will detect it seems like it can vary depending on you.

I have explained how I do it and that explanation do not need to include my reaction time.

[joeqsmith]

I certainly believe you feel you have explained it well enough,  but details mater.    No matter though. 

Video of the 121GW prototype and the Gossen with the counter taken at 1000fps.   
https://youtu.be/13nv-NsQXDs?t=118

I tried one with the BM789 and can pulse it to where I can clearly see with the high speed camera that the display is showing the correct value but only for a few frames.  It's fast enough that the LCD does not reach maximum contrast.  Some human eyes may miss it.   I would imagine that some meters may hold the data up longer once the meter has settled.   As we saw with the 121GW, you may get the first few digits shortly after applying the load but it can take a very long time to settle.   IMO, the text should be written such to handle all of these cases and explain it in detail what the criteria is.     

[joeqsmith]

The spreadsheet has been updated to include data from the BM789.   

[joeqsmith]

I've shown the open circuit voltage waveform I apply to test the meters many times but I have never shown the current.   If you were curious and afraid to ask, this video is for you.  For you vintage TE collectors, you can see my 1960's Tektronix in operation.  Not a bad probe for it's age.     

[joeqsmith]

It's been a while since anyone has asked me about looking at the UNI-T UT61e+.   It's been available long enough that I assume the interest in it has died down but the one comment this last video had was once again asking about this meter.   

While I have vowed never to purchase another UNI-T (unless it were an improved UT181A) because of their poor track record, if you want to see this particular meter ran, let me know.   Maybe there are still some UNI-T fan boys out there who still believe??

Dave's review showing the internals: 

[joeqsmith]

Dave did a pretty good job with the overall review and made it pretty clear that the value wise, it's not a good choice.  It may have curbed some of the interest. 

For the few of you who still want to see it, we could treat it as a special case and deviate from the normal tests to try and keep it alive (don't expose it to any ESD).   I have never life cycled any of the UNI-T products.   It may be interesting to see how it holds up.   

[joeqsmith]

Another option,  start with the grill starter, kill it in a few seconds and leave it at that....  Nice short video, right to the point.   

[Fungus]

Another option,  start with the grill starter, kill it in a few seconds and leave it at that....  Nice short video, right to the point.   

Correct.

I mean, seriously, what is the point of testing the UT61E for "robustness"? We all know it isn't.

[rsjsouza]

Another option,  start with the grill starter, kill it in a few seconds and leave it at that....  Nice short video, right to the point.   

Correct.

I mean, seriously, what is the point of testing the UT61E for "robustness"? We all know it isn't.

At first glance yes. The UT61E (and probably the UT61E+) are finicky with transients but why not put it through its paces? Who knows? Perhaps it could be an additional advantage of the "plus" model...

Obviously that the "real" test would be with the "third party plus-certified and mega-accredited and über-listed" UT161E, but that is in another price league...

At any rate, I don't think there is much to be lost, especially after you already beaten to death the UT61E original, with excellent suggestions to increase its robustness.

(edit) Kerry Wong also did a teardown of the UT61E+
http://www.kerrywong.com/2021/04/04/teardown-of-a-uni-t-ut61e-true-rms-multimeter/

(edit2) Tenma has some discounts on beefed up UT61E (non plus)
https://www.eevblog.com/forum/buysellwanted/newark-sale-on-many-tenma-brand-meters/msg3655997/

[gnavigator1007]

I realized some time ago that if Joe runs it, I'll watch it. There's been quite a few meters that I really had no interest in ever owning, but I still find their point of failure and Joe's analysis interesting. I really appreciate his systematic approach. There have been a few surprises along the way too.

[joeqsmith]

Another option,  start with the grill starter, kill it in a few seconds and leave it at that....  Nice short video, right to the point.   

Correct.

I mean, seriously, what is the point of testing the UT61E for "robustness"? We all know it isn't.

At first glance yes. The UT61E (and probably the UT61E+) are finicky with transients but why not put it through its paces? Who knows? Perhaps it could be an additional advantage of the "plus" model...

Looking at the data I've collected,  of the eight UNI-T products I looked at, half never made it past the ESD, AC line tests.    During the EEVBLOG review for the 61E+, Dave focuses on the lack of a surge rated resistor.   The UT181A is the same and survived some decent hits after a few small changes.    Dave talks about the new ground path.   It could help but my guess is it won't.   

Obviously that the "real" test would be with the "third party plus-certified and mega-accredited and über-listed" UT161E, but that is in another price league...

That meter would fail ESD just like the stripped down version.   

At any rate, I don't think there is much to be lost, especially after you already beaten to death the UT61E original, with excellent suggestions to increase its robustness.

For you, skim the video for a few seconds and your done.    Flip side, it takes fair amount of my time to run the tests and edit the videos.  For this meter, assuming the grill starter kills it, maybe six days.   Half of that would be cycling the function switch.   
 

(edit) Kerry Wong also did a teardown of the UT61E+
http://www.kerrywong.com/2021/04/04/teardown-of-a-uni-t-ut61e-true-rms-multimeter/

(edit2) Tenma has some discounts on beefed up UT61E (non plus)
https://www.eevblog.com/forum/buysellwanted/newark-sale-on-many-tenma-brand-meters/msg3655997/

Skimmed the Kerry Wong video but didn't notice anything beyond what Dave had gone over. 

We could reorder the tests for this special meter.  Start with the 100us transients.  If it survives  that (which it won't) then run the AC line test and then the ESD.  Not sure we would learn anything more doing this.  The end results the same. 

UNI-T fanboys are already thinking I am biased against this meter.  I can see the mass of down votes. lol.  I have more than enough data now to know what to expect. 

[joeqsmith]

The green was a bit odd but now blue as well??    Is blue better? 

https://youtu.be/bxluPKhyOUU?t=1738

[Nixfried]

Hey Joe,

i was one of the few who voted against doing these kind of videos. (Seems like years ago...)
If i could vote again, i would definitely change my answer.
Just wanted to say that your videos are awesome, please keep them coming.

Just a little question regarding the BM869s:
I am using my BM869s to discharge capacitors before working on them(Through capacitance mode). In the manual brymen only states that
large value capacitors should be discharged through an appropriate resistive load. What is your opinion on this?

Best regards,
Tim

[rsjsouza]

Another option,  start with the grill starter, kill it in a few seconds and leave it at that....  Nice short video, right to the point.   

Correct.

I mean, seriously, what is the point of testing the UT61E for "robustness"? We all know it isn't.

At first glance yes. The UT61E (and probably the UT61E+) are finicky with transients but why not put it through its paces? Who knows? Perhaps it could be an additional advantage of the "plus" model...

Looking at the data I've collected,  of the eight UNI-T products I looked at, half never made it past the ESD, AC line tests.    During the EEVBLOG review for the 61E+, Dave focuses on the lack of a surge rated resistor.   The UT181A is the same and survived some decent hits after a few small changes.    Dave talks about the new ground path.   It could help but my guess is it won't.   

The impact of a transient would be highly dependent on many factors - a certain Joeqsmith even said once that it is very hard to evaluate the robustness by looking at a PCB...

Obviously that the "real" test would be with the "third party plus-certified and mega-accredited and über-listed" UT161E, but that is in another price league...

That meter would fail ESD just like the stripped down version.   

Indeed ESD is a different kind of beast.

At any rate, I don't think there is much to be lost, especially after you already beaten to death the UT61E original, with excellent suggestions to increase its robustness.

For you, skim the video for a few seconds and your done.    Flip side, it takes fair amount of my time to run the tests and edit the videos.  For this meter, assuming the grill starter kills it, maybe six days.   Half of that would be cycling the function switch.   

And don't I know that? It seems the more effort put on a video, the lesser the audience watches thoroughly. Anything related to multimeters or consumer products goes a long way on my channel, but everything else does not go very far w.r.t. audience and retention. Or perhaps it is me. (oh, the mysteries of life...)

At the end of the day, do whatever you want to do with regards to testing. I think that, even if the audience/return is small, someone will use the knowledge to something better (at least that is why I convince myself to keep doing the videos I like and not only the ones that have the further reach).

We could reorder the tests for this special meter.  Start with the 100us transients.  If it survives  that (which it won't) then run the AC line test and then the ESD.  Not sure we would learn anything more doing this.  The end results the same. 

Well, every test exposes a level of weakness - the ESD is quite violent but less energy. IIRC the cert agencies are mostly concerned about the transients (or not, don't quote me on that), I would imagine more meters would survive further in your round of tests.

UNI-T fanboys are already thinking I am biased against this meter.  I can see the mass of down votes. lol.  I have more than enough data now to know what to expect.

[joeqsmith]

Hey Joe,

i was one of the few who voted against doing these kind of videos. (Seems like years ago...)
If i could vote again, i would definitely change my answer.
Just wanted to say that your videos are awesome, please keep them coming.

I don't take offense to it and being from Germany, I'm sure you didn't like me showing off your countries top engineering efforts at Gossen. 

It's never been a real popular subject anyway but I still find it interesting.

Just a little question regarding the BM869s:
I am using my BM869s to discharge capacitors before working on them(Through capacitance mode). In the manual brymen only states that
large value capacitors should be discharged through an appropriate resistive load. What is your opinion on this?

Best regards,
Tim

My opinion is that your question is very generic.   With the lack of constraints as a general rule I suggest you get in the habit of always discharging any capacitor before connecting it to any meter.   While I've had some very ignorant people suggest I was directly discharging capacitors into these meters, it's never been the case.   Still the tests do seem to indicate that some meters may handle a direct discharge better than others. 

For a worse case, you could try it with a DC supply.   Say for example your caps are charged to 1kV worse case.   You need about an ampere.   Roughly 2kohms (PTC + surge rated resistor) with 2X clamps, or 1kV/1kohms.  Maybe cycle the meter through all the modes (assuming you will as some point be in the wrong mode when you connect them).    If the meter survives this, you should be fine (until you have one charged to 2kV).    Keep in mind, 1kV @ 1A could prove lethal.     1kV may be enough to damage the contacts if you do it live.   Dave made a video about this after someone had damaged a 121GW.  I made a couple of counter videos to make the point.  To get some idea, see the attached video. 

Say your worse case was instead 100V.   You know I test every function of the meters with a full wave rectified 220V.  That's with every function selected.   It's killed a few meters but rare.   No Brymens I have looked at were ever damaged from this test.

[joeqsmith]

The impact of a transient would be highly dependent on many factors - a certain Joeqsmith even said once that it is very hard to evaluate the robustness by looking at a PCB...

True but in this case I am drawing from my years of looking at UNI-T products in detail.   Not everyone's educated guesses will carry the same weight.   We can certainly run it as I normally would.  The meter will fail or I will once again eat crow.     

And don't I know that? It seems the more effort put on a video, the lesser the audience watches thoroughly. Anything related to multimeters or consumer products goes a long way on my channel, but everything else does not go very far w.r.t. audience and retention. Or perhaps it is me. (oh, the mysteries of life...)

My two most popular videos show an old FPGA based CPU prototyping board I had wired by hand and a low cost handheld VNA.   

I am scratching my head why the channel has so many subscribers. Ads are turned off, so YT doesn't promote it.  I don't advertise outside of EEVBLOG.  My videos are not released on any sort of schedule and months can go by without making one.   It's maybe the worse way to run a channel.     Combine all that with I am no showman.   You're never going to hear:  Please give it a thumbs up and join my Patreon so I can continue to bring you these high quality videos.       

***
SP

[rsjsouza]

I am scratching my head why the channel has so many subscribers. Ads are turned off, so YT doesn't promote it.  I don't advertise outside of EEVBLOG.  My videos are not released on any sort of schedule and months can go by without making one.   It's maybe the worse way to run a channel.     Combine all that with I am no showman.   Your never going to hear:  Please give it a thumbs up and join my Patreon so I can continue to bring you these high quality videos.       

You did something that nobody did before, so there's that. Also, EEVBlog forums are primed for test gear, therefore solely promoting this here is already a tremendous leg up. The fact you don't give a hoot for brand loyalty is another factor. At last, a bit of controversy (the subject of this poll) stirs the audience, including the fanboys that will throw tomatoes at the screen.

Although I am monetizing my channel, I don't have any hopes of it becoming a true source of income (the first US$100 only after 13 months - yay!) and could easily gather more audience speaking a language less uncommon than Portuguese. On the other hand, I do this for the love of the trade and to spread some of this knowledge to my fellow countrymen (or languagemen, as Portugal, Angola and others also watch).

[joeqsmith]

Although I am monetizing my channel, I don't have any hopes of it becoming a true source of income (the first US$100 only after 13 months - yay!) and could easily gather more audience speaking a language less uncommon than Portuguese.

You owe me a coffee.  Mine just came out my nose.   

The question now is how much was spent during that time?      

[rsjsouza]

Although I am monetizing my channel, I don't have any hopes of it becoming a true source of income (the first US$100 only after 13 months - yay!) and could easily gather more audience speaking a language less uncommon than Portuguese.

You owe me a coffee.  Mine just came out my nose.   

The question now is how much was spent during that time?   

Pssst! I can't say it out loud...

Seriously, in the year-and-a-half of the virus not much was spent - mostly freight here and there. Throughout the history of the channel I have been buying dirt cheap gear from and selling it back to the local market after doing the videos and some restoration, so the losses are very low. Also, those dirt cheap chinese multimeters are sold for almost the same price (after all they are practically new), so no significant loss is found there as well.

Overall the drive is not the money, but it surely helps and can materialize in more content to be shown.

[joeqsmith]

Since the virus, I've slowed my spending.  Some forced by lack of available parts.   This is the first new meter I have purchased.   Sadly, there isn't much of a market for damaged DMMs.  If I had no ethics I would peddle them.   Some of them still look new, in the box.  Sold as non-working, looks new,  I don't have the ability to test them,  no returns.       The one I had salvaged for parts to give to another forum member, they couldn't bother to say thank you.  After that, all the damaged ones head to the recycle bins, which is where this meter will soon be heading as well.