Handheld meter robustness testing

[joeqsmith]

It's in the computers hands now.  I just packed it into one 40 minute video that I will upload later today along with a new spreadsheet.   

On the plus side, I doubled down and used both the modified UT181A and the Brymen as a side-by-side reference.  Having used the 181A for a few weeks now, I hope that UNI-T releases a B model that addresses some of my concerns because I rather like the meter in some sick, demented way...

Looking down the road, there are only a few meters that I would like to add to the list.  One would be a Gossen, the other Dave's fat boy custom meter.   

[SeanB]

You think Dave's meter has a chance of surviving at least half way through? Should at least meet the CAT ratings though.

[joeqsmith]

I'm sure it will meet the standard the way they are interpreted.   Maybe once things get closer, Dave will fill us in on his take about some of these details.  He has certainly reviewed enough meters and done several videos where he mentions fuses, MOVs, GDTs, creepage, clearances, layout......   Should be some what of a multimeter expert.   If his goal is to have a product that is robust as a $20 pocket meter, I'm sure he will make it happen.   


As far as my guess on how it would do against the meters I have tested, we would have to run one.  My track record is not very good.   A few examples:

Will the Fluke 87V survive the same transients the Fluke 101, 107, 115, 17B+ survive.  My guess would have been a big yes. 

Will the AMPROBE AM530 which costs about double the AM510 be able to survive the same levels. My guess was the 530 would far exceed the 510.     

Will the UNI-T UT181A, their top of the line meter,  survive an low energy ESD event.   My guess was all day long. 

You think Dave's meter has a chance of surviving at least half way through? Should at least meet the CAT ratings though.

[joeqsmith]

If you want to see some low cost pocket meter carnage, here you go..



The updated spreadsheet is also available. 

[tautech]

Gave one of those Victors to a mate years ago and saw him get it out of his "good stuff" drawer the other day for some checks on auto wiring. That it was still going strong did surprise me at the time but not so much now after noticing that it measured up as reasonably accurate against the others you tested.

That they all didn't fail at 1.5kV was astounding. 

Great stuff as always Joe. 

[joeqsmith]

Thanks.   I was thinking the two UNI-Ts would not survive the grill starter.   Just shows a meter with no MOVs can take a fair amount of abuse.  Maybe by design, but my guess is they got lucky this time.   Sad the 181A with all of it's fancy protection could not take the abuse.   But the layout has to be done right.....

That they all didn't fail at 1.5kV was astounding. 

[Fungus]

Interesting that some of them actually blew the case open when they failed even though it's still in the "low energy" range.

I wouldn't like to have one in my hand when that happens.

(More proof that the magnetic hanger supplied with high-end meters is really a safety feature, not just a convenience)

[joeqsmith]

I keep waiting to see one actually open up all the way but my setup just does not pack that big of a punch. 

Interesting that some of them actually blew the case open when they failed even though it's still in the "low energy" range.

I wouldn't like to have one in my hand when that happens.

(More proof that the magnetic hanger supplied with high-end meters is really a safety feature, not just a convenience)

[Muttley Snickers]

How would this thing below go ?, I had a play with one of these recently only for a few minutes and for a couple of awkward applications I can see a spot for one in my toolbox, they are around thirty odd bucks I think and sold under many names including Mastech.

 

[smithnerd]

Any chance of some hi-res pictures of what's left of the VC921 PCB?

That model has been discussed here before, but the PCB looks to have been redesigned:

https://www.eevblog.com/forum/testgear/victor-vc921-pocket-multimeter-worth-a-bash/

[joeqsmith]

Have a look at these.

Any chance of some hi-res pictures of what's left of the VC921 PCB?

That model has been discussed here before, but the PCB looks to have been redesigned:

https://www.eevblog.com/forum/testgear/victor-vc921-pocket-multimeter-worth-a-bash/

[joeqsmith]

There's lots of different ones that look identical outside the color.  Like the MS8211 and the 8211D.  Some show CAT II, some CAT III.  $22 to $50 depending who sells it.   

How would this thing below go ?, I had a play with one of these recently only for a few minutes and for a couple of awkward applications I can see a spot for one in my toolbox, they are around thirty odd bucks I think and sold under many names including Mastech.

[saturation]

Great video Joe, thanks again.  Given how less robust these DMMs are designed it becomes like a teaching video of where and how they fail, and mitigating one failure can just expose another.

[joeqsmith]

Thanks. 

Just an FYI.  Notice that I had marked the UT10 failing at 4KV when it broke down but when I ran the HIOKI I marked it as a pass.  I'm sure this will be a source of confusion.   When I ran the UT10, I took it to 5KV and knew it was not going to become non functional with the energy levels I test with.  Like the HIOKI, I wanted to know if I prevented it from breaking down, could it be ran at a higher level.   To be sure, I reset the generator to 4KV while I was making the changes as the meter did not appear to breakdown at 3K.  Once I had reworked the meter three times, it because permanently damaged.   The HIOKI on the other hand actually continued to survive after I had increased the creepage with the added plastic.   

I think in the start of these tests I was looking for any sort of a breakdown as a failure.  Of course, I also changed the matrix several times as I was sorting out how much an average meter could handle and how much time I wanted to spend finding their limit.  Using the scope is how I was judging if a meter started to breakdown but the scope and visual does not insure it.  I started looking for permanent damage to judge a pass/fail.   This is why so many meters early on were hit over and over when I knew they were arcing internally.   The UNI-T fan boys did not like their 139C getting treated like this.        And of course, safety was never a concern....

I did get a comment about rotating the selector switch of the UT20 while it was obviously arcing.  No gloves and such.  If I were plugging them into a wall socket here, I would have something over my hands but the max current I can get out of the generator is only a few hundred mA and I set the trip point to 40 or so.  They also noticed that when I run the 1/2 cycle simulator, I only point to the meters with that long discharge probe.   When I play with that setup, I stand off to the side and back on the other side of the room just in case.  So far, the really has not been a case where I was concerned.  I think the worst was blowing up the glass fuse in the VICI.  I knew that was going to make a mess and closed the door. 

Great video Joe, thanks again.  Given how less robust these DMMs are designed it becomes like a teaching video of where and how they fail, and mitigating one failure can just expose another.

[rsjsouza]

Joe, one more great video, thank you.

My cynical side tells me the short creepage in the UT-10 is in fact the DMM protection circuit... 

[joeqsmith]

My dark side says they are not that smart but it's possible.  If I ran that half cycle line simulator on a stock one, it would have cut loose.

Someone just asked me about the AM510s capacitance range.  I looked it up and see they show 4000uF on their site.
http://www.amprobe.com/amprobe/usen/digital-multimeters/am-500-digital-multimeter-series/amp-am-510.htm?pid=74033

Attempted to read a 1000uF with mine and it over ranged.  Tried it on 5ky's AM530, no problem.   I tried a few tests and it over ranges at 100uF.   I looked in the manual that was shipped with the 510 and it states 100uF max.   So their website is wrong! 

They may offer a newer version of the AM510 than the one I bought.   The manual for the 530 shows 4000uF.  I ran 5Ky's against the Brymen at 2200uF and it's close but I had done a complete alignment on that meter after I damaged and repaired it.   

[rsjsouza]

Joe, the online version of the manual also says 100uF, thus I suspect the website suffered a Ctrl+C Ctrl+V issue...

[joeqsmith]

Interesting link showing a commercial generator.   Follow their other link where they modify it.    Imagine some of these low end meters on that setup!   

http://www.electricstuff.co.uk/surge.html

[joeqsmith]

After testing so many meters, I have a fair number of test leads.  The meters get recycled but I hung onto the leads thinking I may do something with them.   

One way to test them would be to put some DC current through them.  The HF meter for example can measure up to 10A.  I would think the leads could handle at least 12.   

Another may be a pull test on the connections.   

Maybe I apply that half cycle transient through them without a fuse.

[saturation]

Hi Joe, IIRC the IEC safety requires a sustained 10-15kV for test leads for working voltages of 1kV with potential breakdown test points in various locations and mechanical stress configurations.

Here's the full suite as applied to Amprobe test leads. 

After testing so many meters, I have a fair number of test leads.  The meters get recycled but I hung onto the leads thinking I may do something with them.   

One way to test them would be to put some DC current through them.  The HF meter for example can measure up to 10A.  I would think the leads could handle at least 12.   

Another may be a pull test on the connections.   

Maybe I apply that half cycle transient through them without a fuse.

[joeqsmith]

   They drop test probes???!!!    Pretty sure I could perform that test at home.   

I am looking at page 19.  Wow!!   So am I reading this right.  For the cable they show 29.0mm for both the creepage and clearance.  Then we see a test voltage of up to 14880 for the cable.   The sub-clause they refer to is 6.4 (for example) which refers to section 6.8 

"Solid insulation between ACCESSIBLE parts and HAZARDOUS LIVE parts shall pass the voltage test of 6.8 for BASIC INSULATION"   

In 6.8.4
"Voltage tests are applied, using the values specified in table 9. No breakdown or repeated flashover shall occur. Corona effects and similar phenomena are disregarded."

Then going to table 9, they step from 20 to 25 to 30mm.   For 29mm, I would have guessed they would use 30mm which calls for a AC peak 50/60 Hz or DC or 17.9KV.  For AC RMS, they show 12.6KV.   

I have no idea where the 14880 came from.        

Hi Joe, IIRC the IEC safety requires a sustained 10-15kV for test leads for working voltages of 1kV with potential breakdown test points in various locations and mechanical stress configurations.

Here's the full suite as applied to Amprobe test leads. 

[saturation]

Yes, I think a bottom line is test leads are pretty robust for DMM applications at 1kV given how overrated the requirements are.  I guess this takes into account user abuse [ e.g. yanking, pulling], and field wear and tear.

[joeqsmith]

Some of the leads from the free Harbor Freight meters, I don't think would handle the 10A.   I have a few sets of these I could try.   

In the report you linked, strange that they show rms/peak/dc but the standards call out a different number for rms and dc.   Strange.   For 30mm using the 1.5/50 impulse, they call out 23.3KV for 30mm.    I could run them on that generator and if they fail, that would be a good indicator there is a problem with them.   

[joeqsmith]

Made up some adapters that will allow me to place the probes in line with some sort of current source. 

My plan is to twist the probes together and hit them with the older generator with it set to the 15KV.  The probe tips will not be connected to anything.   Not even close to the voltage levels called out in the standards require but none of my testing ever is.

Then I plan to have some sort of computer programmable current source that will step the current and hold it for some time.  It will increase the current until the probes open up or I run out of current.   
 
Any probes that I can't damage that way, I plan to just connect across the half cycle generator.

May change the order. 

[joeqsmith]

Finished with the current source test setup.   I ran some tests with it using a few different probes and the differences are pretty dramatic.   I already melted one lead (in the range it was rated for!).