New UNI-T UT892 2000V AC/DC High Voltage Multimeter

[PartialDischarge]

I don't really see the point, 1) it is not like anyone is going to modify the board and 2) these voltages are 3x..5x times higher than anyone would use this multimeter for.
Also there is a chance that the board is getting punctured since the ground trace runs directly below the input trace and the board is only 1mm thick or so. If this is the case adding silicone or varnish would not help.

I still think that removing the PTC and adding some 100kOhms in there is a nice improvement.

[Fungus]

I still think that removing the PTC and adding some 100kOhms in there is a nice improvement.

I still want to know why the PTC is there in the first place. Is it supposed to protect against any arcing that happens before the resistors?

[NoMoreMagicSmoke]

I still think that removing the PTC and adding some 100kOhms in there is a nice improvement.

I still want to know why the PTC is there in the first place. Is it supposed to protect against any arcing that happens before the resistors?

To give reviewers something to point at and say "look at this great input protection. 🤷

PTC will only come into play with enough current to heat them up. If it's arcing before the resistors someone definitely messed up the design. Then again it's Uni-T... It's possible they did mess up the design.

[joeqsmith]

I still want to know why the PTC is there in the first place. Is it supposed to protect against any arcing that happens before the resistors?

To give reviewers something to point at and say "look at this great input protection. 🤷

PTC will only come into play with enough current to heat them up. If it's arcing before the resistors someone definitely messed up the design. Then again it's Uni-T... It's possible they did mess up the design.

They may have a spark gap after the PTC on the backside (intentional or otherwise).   Would be interesting to see it all apart.  Maybe they swapped the parts without removing the board. 

[Mechatrommer]

I still think that removing the PTC and adding some 100kOhms in there is a nice improvement.

I still want to know why the PTC is there in the first place. Is it supposed to protect against any arcing that happens before the resistors?

To give reviewers something to point at and say "look at this great input protection. 🤷
PTC will only come into play with enough current to heat them up....

and you will only see that when the dmm breaks down, such as i've mentioned when high energy arcing occured AFTER the PTC. the reason you think its useless being there is because you havent seen its failing in such a manner yet. i dont think UNI-T is fool enough to put $1-5 part just as pure show off. you may argue that it is slow acting in ms range, but few ms second cutoff is better than few seconds of mass slaughter.

[NoMoreMagicSmoke]

the reason you think its useless being there is because you havent seen its failing in such a manner yet. i dont think UNI-T is fool enough to put $1-5 part just as pure show off. you may argue that it is slow acting in ms range, but few ms second cutoff is better than few seconds of mass slaughter.

The reason I think it is useless is based on previous Uni-T meters. A quick peruse of Joe's video's will show multiple Uni-T meters with PTC's in them that breakdown at surprisingly low voltages.

[joeqsmith]

Typically the low end meters I have looked at use a 5mm body.  The small PTCs still have a lot of thermal mass (compared with how fast things would change with an arc).   I would imagine if they did have a spark gap after the PTC, if it did come into play you may fine the PTC is missing it's coating after. 

[Fungus]

PTC will only come into play with enough current to heat them up.

Which will be "never" when it's in series with 20MOhm, hence my curiosity. 

But without a surge-rated resistor it's going to undergo a rapid uncontrolled disassembly in the case of an arc. Let's hope they go open-circuit when they fail.

i dont think UNI-T is fool enough to put $1-5 part just as pure show off.

The only thing I can think of is they're using them as fuses.

[NoMoreMagicSmoke]

PTC will only come into play with enough current to heat them up.

Which will be "never" when it's in series with 20MOhm, hence my curiosity. 

But without a surge-rated resistor it's going to undergo a rapid uncontrolled disassembly in the case of an arc. Let's hope they go open-circuit when they fail.

i dont think UNI-T is fool enough to put $1-5 part just as pure show off.

The only thing I can think of is they're using them as fuses.

Honestly a lot of Uni-T designs scream what a junior engineer would do if shown a bunch of meters that are "good" and told to make a new meter. You will get a meter that mostly looks good, but will be full of not quites due to the engineer not understanding the why behind the design choices.

[Mechatrommer]

The only thing I can think of is they're using them as fuses.

what do you think it is? Its designed to go open, surge Resistor will go short. It will need something else to go open anyway. which one you want to put in your design?, i can understand the suggestion of more robust input such as surge+fuse etc on a budget device, even though they missed the point of the economy. What i dont understand is removing a better perfectly working protection component in exchange with component that provides no protection at all.

[Fungus]

Maybe there's an excess of cheap PTCs in Shenzhen right now, so ... use what you've got! 

[Mechatrommer]

better cheap than nothing. no? where to get $40 2KV meter verified to 4KV (modded with 2x 0806 resistors) safe?

[joeqsmith]

Its designed to go open, surge Resistor will go short.

Once the PTCs arc, they become a very low impedance.  Without anything to limit the current, the down stream protection will normally get damaged to where it goes open then it's everything after that.   While testing the meters, we have seen the resistors go open.   

The PTCs come into play (will start to heat)  when the low voltage clamps are active and you have applied a high enough voltage to the inputs for a long duration.  For example, you have the meter set to ohms and then proceed to plug the meter into your wall socket.   

[Mechatrommer]

Its designed to go open, surge Resistor will go short.

Once the PTCs arc, they become a very low impedance.

and this happens when we violate its physical law (distance between naked pins) and we cant blame the manufacturer really, if we test it beyond its recommended limit.

[joeqsmith]

Its designed to go open, surge Resistor will go short.

Once the PTCs arc, they become a very low impedance.

and this happens when we violate its physical law (distance between naked pins) and we cant blame the manufacturer really, if we test it beyond its recommended limit.

If designers use a 100V part in a 10kV application, I most certainly will.   There is a reason that the higher class meters use the larger parts and further limit the current.  There also a reason I have never seen a PTC fail in the higher class of meters.   Nothing to do with this particular meters 2kV input. 

[Martin72]

UT196 arrived at work...
For appx 160€ it looks solid, but not the quality a brymen in this range do got.
There are "air gaps" around the input, but this is not the input where you can measure up to 1700V.
First test before giving it away for calibrating on the magna power 0-1500V supply.

Martin

[Kleinstein]

The combined Ohms and high voltage input could be tricky: the ohms part kind of needs a connection to the switch and there tend to be marginal creep paths.  So the seprate higher voltage input with no ohms fuction is what make it relatively easy to provide a safe 2000 V range. It is still not safe from user error - but with 1000 V or more some case is advised anyway.

[floobydust]

The UT196 they at least splurged on MOV's and four pair of bigger SOT-89 clamp transistors. But the spacings for R1 2MEG are hilarious if you follow the pcb trace.
We don't know if it's actually 61010 certified, Uni-T removed their certificates from their website after it revealed their dodgy practices. It's common practice to deceive and say "designed to" or "follows" 61010 but nobody ever tested it. Provide the certificate and file number if the multimeter has passed regulatory.
It's very difficult to find a PTC rated for 1,000V and UL approved. Putting two in series does not work because they have unequal resistance and heating times.

[joeqsmith]

Larger PTCs as well.   Are they using three MOVs in series for the V input side and two in series on the low voltage circuit? 

I have ran my BM869s up to 1.5kV but for a quick demo,  the clamps start to conduct around 1.8kV and they don't blank the display.  The prototype BM789 I have will run up a bit higher before it starts to clamp.    I am uniquely qualified in the art of stupidity and am not suggesting anyone attempt to do this with your meters.   It would be far better to buy a proper probe.