Handheld meter robustness testing

[Kosmic]

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

Even though YT don't promote your channel you end up popping up if you search specific subjects or watch really niche videos, they are going to propose your videos on the side. I guess they want the viewer to stay on the platform even if they are not directly making any money with your videos.

For some reason I found your videos on YT before this thread.

[joeqsmith]

The UNI-T UT61E+   

[bdunham7]

Yikes!  How many volt-Hz is that--3 billion or so? 

On the plus side, perhaps this could be a useful desoldering technique.

[joeqsmith]

Sort of a Metcal iron.   

Obviously I wasn't thinking too much about damaging the meter.  Looks like minimum, 4 X PTCs and 4 X transistors would need to be replaced but the PCB seems in  good shape.     

Maybe our friends from China will decode that box for me.

[rsjsouza]

On the video you nailed the interesting bits of the UT61 series: the lightning fast capacitance meter (the original UT61E is the king on this, even on capacitors with tens of thousands of µF) and its wide input frequency (which cooked the frequency range unfortunately). I did not improve the safety on mine since, as long as I keep it on the bench, I don't see it being cooked anytime soon. Also, the absence of an auto-power off is great.

Back to the frequency range, the UT61E+ manual indicates that frequencies above 40MHz have unspecified accuracy, thus the marketing department was very optimistic on this feature.

The manual also says the maximum input voltage at that range is derated to 20V_RMS maximum, although it says it has a 1000V overvoltage protection. How many volts did you apply?

[joeqsmith]

On the video you nailed the interesting bits of the UT61 series: the lightning fast capacitance meter (the original UT61E is the king on this, even on capacitors with tens of thousands of µF) and its wide input frequency (which cooked the frequency range unfortunately). I did not improve the safety on mine since, as long as I keep it on the bench, I don't see it being cooked anytime soon. Also, the absence of an auto-power off is great.

Back to the frequency range, the UT61E+ manual indicates that frequencies above 40MHz have unspecified accuracy, thus the marketing department was very optimistic on this feature.

The manual also says the maximum input voltage at that range is derated to 20V_RMS maximum, although it says it has a 1000V overvoltage protection. How many volts did you apply?

Right, they don't spec the voltage required for it to work above 40MHz, so I was creeping up on it.  I assume that 20V is the maximum required for it to read the proper value.   I searched for "derate" and it doesn't appear to be mentioned in the manual I have.  I certainly talks about the 1kV overload protection.     

I have no idea how much voltage I was putting into the meter.  There was nothing monitoring it.   I wonder what the impedance of the PTCs would be at 200MHz.   Consider most meters have a surge rated resistor in them, their impedance may be more inductive where the PTCs may be more capacitive.  I doubt the MOVs would have much of an effect as I had the low voltage clamps engaged.   Anyway, what I am getting at is the voltage may not be all that high but the impedance may be low causing excessive power to be dissipated in the PTCs.     

Easiest thing is not to put 200MHz on the box and I won't play at 200MHz.   But with it on the box I want to see it and I'll smoke the meter trying to get it there....   That should have been expected based on my previous years of videos..   

****

Plotting the impedance of a 1.5K surge rated resistor and a PTC.  These are not the parts used by the UT61E+ but they are parts I have shown and used in other meters.   You can see how the PTC is a lot less stable than the resistor.   I suspect this is one weakness in using two PTCs in series like the 61E+ in place of the more typical resistor+PTC combo. 

Again, the UT181A also uses the two series PTCs, but the manual shows the upper range is 60MHz.  In this case, we can verify this really is the upper limit.

https://youtu.be/PjNXbKlr3MI?t=1921

[Fungus]

The UNI-T UT61E+   

(Looks at length of video, very short)

Prediction: The grill starter killed it... 

[bdunham7]

Prediction: The grill starter killed it... 

Nope!  RF burns....

[joeqsmith]

Looks like the majority would like to see it repaired and continue the testing.   Of course we will not use any of the data from these tests and it will purely be for entertainment.    The UNI-T fanboys can talk about how great the meter is if it survives or how I fucked it up during my repairs if it fails.      


Fungus,
Good catch on the 17B+.  I am not sure why it was removed.  The one below it was also missing.   Guessing a user error.  I have added both meters back into the spreadsheet.

[Fungus]

I'm a bit confused: Shouldn't a hot PTC have saved the meter? Isn't that their job?

[bdunham7]

I'm a bit confused: Shouldn't a hot PTC have saved the meter? Isn't that their job?

It looks like they tried, but when dV/dt gets up to 10¹⁰V/s, things happen.  I'd be interested in knowing exactly how much power was applied to the meter.  Expecting a cheapo DMM to measure 220MHz on banana-jack inputs seems a bit outlandish, but that's what UNI-T puts on the box. They probably aren't expecting too many people to be able to actually give it 200MHz @ 20V.  I have a really nice frequency counter that only  goes to 180MHz on the high-impedance channel. 

[2N3055]

I'm a bit confused: Shouldn't a hot PTC have saved the meter? Isn't that their job?

At those frequencies, he basically made a Metcal soldering iron from a meter.
All the RF energy was simply dissipated inside meter as heat..

[rsjsouza]

I'm a bit confused: Shouldn't a hot PTC have saved the meter? Isn't that their job?

At those frequencies, he basically made a Metcal soldering iron from a meter.
All the RF energy was simply dissipated inside meter as heat..

Well, despite the PTC was cooked, the rest of the meter is intact (or should be).

Protection parts can be sacrifical parts.

That is a whole debacle over the role that CAT ratings have in cheaper meters that supposedly are approved (UT139C, UT61E for the EU market and so on): should they survive a transient unscathed or should they only guarantee the operator survives regardless of their own functional state? The second option should be expected, but the first one is a bit in the air.

[bd139]

Personally if it blows the meter up and I survive then it’s a win. Meters can be replaced or possibly repaired.

Sometimes you can’t win.  people blow the meter up and then go for a second round…

And yes the fuse is soldered in as well.

[Fungus]

I'm a bit confused: Shouldn't a hot PTC have saved the meter? Isn't that their job?

At those frequencies, he basically made a Metcal soldering iron from a meter.
All the RF energy was simply dissipated inside meter as heat..

OK, let's see if I've got this straight:
a) Very few actual electrons were getting through the PTC (it was hot!)
b) The PTC melted because of the capacitance of its own leg on the input side and the massive numbers of electrons rushing in/out heating up that leg.
c) The screen grayed out because of the huge EMPs created inside the meter.

Well, despite the PTC was cooked, the rest of the meter is intact (or should be).

If my understanding above is correct then I'd agree.

PS: I think Evil Joe did this on purpose, knowing what would happen.

(...in which case he needs to do it to some more meters in the name of "fairness to Uni-T". How about one of the new Brymens, eg. the one that isn't calibrated? The dodgy Fluke 87V...? )

[joeqsmith]

Part 2 is up. 

[EEVblog]

That is a whole debacle over the role that CAT ratings have in cheaper meters that supposedly are approved (UT139C, UT61E for the EU market and so on): should they survive a transient unscathed or should they only guarantee the operator survives regardless of their own functional state? The second option should be expected, but the first one is a bit in the air.

The CAT ratings are safety standards, it just has to fail safe.
I don'y recall if that's for every test in the standard, but it's the basic gist of the high energy testing.

[Fungus]

That is a whole debacle over the role that CAT ratings have in cheaper meters that supposedly are approved (UT139C, UT61E for the EU market and so on): should they survive a transient unscathed or should they only guarantee the operator survives regardless of their own functional state? The second option should be expected, but the first one is a bit in the air.

AFAIK the CAT ratings have never even mentioned the first. All they say is that the device should fail safely.

[Fungus]

And yes the fuse is soldered in as well.

This is both:
a) Unbelievably common.
b) Why you never, ever borrow a meter in a life-or-death situation.

Still, it ought to take a major protocol failure for that 'fuse' to kill anybody. You shouldn't be holding the meter in your hand when you do any current measurements in an industrial environment - that's what magnetic holders are for - and you shouldn't be doing any current measurements at all where there's a danger of overload. Get a clamp, FFS

If I were an "Industrial Electrician" I wouldn't take a meter with any current measurement functions to work. I'd own a current clamp and the simplest possible multimeter, eg. a Fluke 113 with only two input connectors and two positions on the dial.

[bd139]

Makes sense to me. I rarely even use current measurement in electronics work. There’s usually a resistor somewhere I can take a voltage sample off or a calculation I can make. Breaking circuits for measurements is a pain in the ass.

When debugging some car issues recently I grabbed a cheap UT210E clamp meter.

[joeqsmith]

I've provided you with quotes from the standards and how at least one of the top handheld manufacture and one of the lower tier interprets them.   

[bdunham7]

The CAT ratings are safety standards, it just has to fail safe.
I don'y recall if that's for every test in the standard, but it's the basic gist of the high energy testing.

IIRC, the meter must continue to indicate the presence of hazardous voltages, although it doesn't have to be accurate.  This makes sense, or alternatively (not in the standards AFAIR, just my opinion) at least it should fail completely (no display) rather than indicate 0.000V.  Imagine I hang my meter up at a machine I want to test and hook it up, it says 240VAC.  I walk over to the service panel and turn off the wrong breaker, but there is a big transient that blows open the inputs on the meter--but doesn't cause any physical damage.  I come back and my meter reads 0VAC--so I go to work.  This is an argument against fused leads and also is the basis for the 'test before touch' protocol that requires that the meter be proved immediately before and after testing.

[joeqsmith]

And don't forget, there is actually a sticky thread for multimeters and safety which has nothing really to do with this thread. 

[rsjsouza]

I've provided you with quotes from the standards and how at least one of the top handheld manufacture and one of the lower tier interprets them.

Precisely. It hinges on interpretation. It is not crystal clear nor set in stone.

As I also mentioned before, this is understandable as standards are comprised of various players in the industry with different goals and aiming different target prices. There is always a give and take going on at standards' commitees.

[joeqsmith]

I've provided you with quotes from the standards and how at least one of the top handheld manufacture and one of the lower tier interprets them.

Precisely. It hinges on interpretation. It is not crystal clear nor set in stone.

As I also mentioned before, this is understandable as standards are comprised of various players in the industry with different goals and aiming different target prices. There is always a give and take going on at standards' commitees.

IMO, besides the vague verbage I don't personally feel they have kept up with the times.  Who would ever have thought a meter with latching relays would every make it to market.  There's no test for it.      Have they lost their way?   Seems so.