EEVblog 121GW Multimeter Issues

[J-R]

The DMM will drift somewhat for many minutes after power-on, and for some seconds after changing modes and starting measurements, and yes, some drift over many minutes or hours depending on the measurement.  I would interpret the accuracy specs as simply as they are presented, which is that at any given moment (unless otherwise specified in the manual), the readings on the display are within specifications compared to the reference.  Drift isn't really covered.

I can tell you from experience that certain measurements do wander for what seems like forever on this DMM, and so do other DMMs.  But how much and under what function and under what conditions is what needs to be discussed further I think:

- mV more details please?  AC/DC? How much drift?  What are you using to short the jacks?
- Resistance, use a direct short between the jacks and step back.  What does that look like?
- Capacitance, are you holding the capacitor at all?  But that drift value is very low and so I would not be concerned.  (Something like the DE-5000 might be a good purchase to have as something to verify against.)
- The 270mA going through the 400mA jack is probably to be expected since the fuse and shunt are going to heat up and the resistance will change.  This changes the circuit it is part of as well.  Try the 10A jack for a comparison.
- Capacitance short, solved.
- That 0.0041mA and similar readings are quite normal from what I'm seeing on my two units.  Tinkering around with a short lead I would say random readings under 0.0050mA is going to be normal.  Use the uA range for low values instead.

[vyruz]

Hi everyone.
It seems my 121GW just broke, and I have no idea how or why :S
I wanted to measure a simple 230V AC mains socket, so plugged my leads in to the COM and V sockets, set the DMM to V, and the mode to AC, but when I plugged the leads into the mains socket, there was  loud POOF and scorch marks on one of the lead tips. Strangly the DMM display kept working, but displaying no voltage. I tried another DMM (Uni-T UT61E) and that one just correctly displayed 230V AC.

So I opened up the 121GW, and to my horror noticed the typicall smell, strangly, both fuses were still intact. Further disassembly showed burned/exploded components at the U17,  C61, R131, C69, C68, C80 and PTC3 positions.

So 2 questions here: What have I done wrong? I'm 100% sure I connected the leads to COM and V, the DMM was set to correct mode, I can't think of a reason why something like this would happen.
Secondly, have I got any change at repairing this DMM? Replacing those 7 components at the earlier listed positions (given I can find the correct parts/values somewhere) is something I'd feel confident enough to do, but how high is the chance other components will have failed invisibly, and would the DMM still be thrusworthy calibration-wise afterwards?

I feel really sad about this, It's the most expensive DMM I ever bought, and worked really well for most things I do

[joeqsmith]

Hi everyone.
It seems my 121GW just broke, and I have no idea how or why :S
I wanted to measure a simple 230V AC mains socket, so plugged my leads in to the COM and V sockets, set the DMM to V, and the mode to AC, but when I plugged the leads into the mains socket, there was  loud POOF and scorch marks on one of the lead tips. Strangly the DMM display kept working, but displaying no voltage. I tried another DMM (Uni-T UT61E) and that one just correctly displayed 230V AC.

So I opened up the 121GW, and to my horror noticed the typicall smell, strangly, both fuses were still intact. Further disassembly showed burned/exploded components at the U17,  C61, R131, C69, C68, C80 and PTC3 positions.

So 2 questions here: What have I done wrong? I'm 100% sure I connected the leads to COM and V, the DMM was set to correct mode, I can't think of a reason why something like this would happen.
Secondly, have I got any change at repairing this DMM? Replacing those 7 components at the earlier listed positions (given I can find the correct parts/values somewhere) is something I'd feel confident enough to do, but how high is the chance other components will have failed invisibly, and would the DMM still be thrusworthy calibration-wise afterwards?

I feel really sad about this, It's the most expensive DMM I ever bought, and worked really well for most things I do

Wow.   Looks like the path was the PTC to the shield down the grounding spring.   With that PTC connected right to the input, there's not much to limit the current.    I doubt this was any fault of yours.   Thanks for posting.     

[PA0PBZ]

Looks like the marked part of the shielding should not be there - manufacturing error?

[J-R]

It appears the edge of the shield was cutting into the heat shrink around PTC3.  Assembly or re-assembly error?

[Andrew McNamara]

Looks like the marked part of the shielding should not be there - manufacturing error?

I'd say possible assembly error, but also a design shortcoming - there is a layer of plastic over the shielding, but it's too tight in that area and it appears the shielding hit the top of the PTC, nicked it and crumpled up slightly instead of going down beside it - and there's no way to tell if this has happened as the meter is (re)assembled.

[J-R]

Just checked the ones I have and the shield is firmly attached to the back cover and quite a bit away from PTC3.

[EEVblog]

Hi everyone.
It seems my 121GW just broke, and I have no idea how or why :S
I wanted to measure a simple 230V AC mains socket, so plugged my leads in to the COM and V sockets, set the DMM to V, and the mode to AC, but when I plugged the leads into the mains socket, there was  loud POOF and scorch marks on one of the lead tips. Strangly the DMM display kept working, but displaying no voltage. I tried another DMM (Uni-T UT61E) and that one just correctly displayed 230V AC.

So I opened up the 121GW, and to my horror noticed the typicall smell, strangly, both fuses were still intact. Further disassembly showed burned/exploded components at the U17,  C61, R131, C69, C68, C80 and PTC3 positions.

So 2 questions here: What have I done wrong? I'm 100% sure I connected the leads to COM and V, the DMM was set to correct mode, I can't think of a reason why something like this would happen.
Secondly, have I got any change at repairing this DMM? Replacing those 7 components at the earlier listed positions (given I can find the correct parts/values somewhere) is something I'd feel confident enough to do, but how high is the chance other components will have failed invisibly, and would the DMM still be thrusworthy calibration-wise afterwards?

When did you buy this meter?
Serial number?

[EEVblog]

It appears the edge of the shield was cutting into the heat shrink around PTC3.  Assembly or re-assembly error?

The metal shield side would have to have been folded over somehow and touching or very close to touching a terminal of the PTC.
Like the outer PTC3 was actually behind the plastic shield and now on the metal side.
Can't see any other way this could have happened. Just cutting into the top of the heatshrink wouldn't have been enough for something as basic as 240V mains. Unless fo course there was a combination of this plus a sizeable tranisent to breakdown the clearance/creepage.

[J-R]

Looks like PTC3 was pulled away from PTC4.  On mine they are still stuck together, which keeps PTC3 out of the way. So it seems plausible the metal shield had been doing some cutting into PTC3.

Maybe vyruz can tell us if the DMM had been previously disassembled?  Or they may not want to admit that...

[J-R]

Yes, I can see the plastic pulled away from the metal shield...

[EEVblog]

Looks like PTC3 was pulled away from PTC4.  On mine they are still stuck together, which keeps PTC3 out of the way. So it seems plausible the metal shield had been doing some cutting into PTC3.

Yes, but even if it "cut" into the heatshrink of the PTC you still have the device encapsulent, so that's not enough for a 240V on it's own to cause an issue.

[J-R]

Looks like PTC3 was pulled away from PTC4.  On mine they are still stuck together, which keeps PTC3 out of the way. So it seems plausible the metal shield had been doing some cutting into PTC3.

Yes, but even if it "cut" into the heatshrink of the PTC you still have the device encapsulent, so that's not enough for a 240V on it's own to cause an issue.

Dunno, that metal seems pretty sharp to me...  How much would it need to cut into the PTC before it causes a short?

Could it be a combination of the proximity of the metal and a catastrophic failure of the PTC?

[Monkeh]

If the edge of that shield is sharp enough to cut through a PTC it shouldn't be allowed to flap in the breeze.

[EEVblog]

Dunno, that metal seems pretty sharp to me...  How much would it need to cut into the PTC before it causes a short?

Just tried to get my exacto knife into one and didn't get very far.

Could it be a combination of the proximity of the metal and a catastrophic failure of the PTC?

Possible. Although the OP said it happened imediately when touching the mains, so that seems to rule out a surge as a contributing factor.
Maybe a combination of variability in the PTC encapsulent thickness on the top and a small cut into it?

[J-R]

Could it have split the PTC open a bit?  Or maybe that happened when it failed?

[vyruz]

Thanks for all the replies!
Thinking more about this, yesterday might have been the first time I ever measured mains current with this DMM, I mostly bought it (besides sponsoring Dave of course ) for working on low power electronics (5-12V, <5A). I normally use a simply Fluke 113 when I work on mains, it just happened yesterday that this DMM was laying on the table right beside the mains socket I wanted to measure...
I was thinking that I should have noticed 'issues' (not sure how they would manifest exactly) with low power circuits as well, if there was a persistent short across the shield, but as others have suggested, it may have been a freak case of the shield having nicked the PTC just enough so only mains power would cause a short.

I got this DMM through the initial Kickstarter campaign, the serial number is 180701285, and there's an EEVBlog authenticity label with the number 001882 beside it.
I found a shipping notification of Welectron on July 18th, 2018, so probably received it a few days after.

I never dis-/re-assembled the DMM before this happened yesterday, but of course can't offer more than my word on this. I am also aware the 1 year warranty is well expired. The intention of my post was not to get free replacement (though I would be happy to accept one ), I just though I had done something wrong. at this point I'd be happy to get some advice on  the feasability on a possible repair, and of course preventing this from happening to other users if it really is a manufacturing defect.

I snapped some more pics of the PTC ('vanilla' and with the heat shrink removed) and shield in the specific location, the shield shows a hole a couple mm's down from the edge of it, so that would mean the shield was not actively cutting into the PTC at the time this happened. It looks like the short went out of the side of PTC into the flat area of the shield (through the plastic film).

Yes, I can see the plastic pulled away from the metal shield...

The hole I mentioned exists in the plastic as well as the metal shield, so would mean the plastic was adhered to the shield when the fault happened. I fumbled a bit when disassembling the DMM yesterday, so might have been me that separated the plastic from the shield.

[crispus]

It seems my 121GW just broke, and I have no idea how or why :S

Unfortunately yours is looking way worse than mine, but still, this shouldn't happen. I am in the same situation (Kickstarter, never used it on mains until the incident, just a matter a fact, I used it very rarely)

We must have been lucky enough to catch some corner case... For some reason 121GW doesn't like mains 

[J-R]

The cut marks in the heat shrink around the PTC still really make me suspect the metal shield was cutting into it.  PTC fails and splits open, internals make contact with the metal shield.

[jancumps]

For reference, I've opened my meter, and took a photo of that area of the shielding.
SN 180701835
The plastic is starting to shed from the alu.

[joeqsmith]

I'm interested in knowing how it's handled.  Granted, there's no warranty but it sure seems like a manufacturing issue and should be covered.   

[e0ne199]

I'm interested in knowing how it's handled.  Granted, there's no warranty but it sure seems like a manufacturing issue and should be covered.

looks like you really hate everything about 121GW...hope that you don't hate its designers as well 

[joeqsmith]

I'm interested in knowing how it's handled.  Granted, there's no warranty but it sure seems like a manufacturing issue and should be covered.

looks like you really hate everything about 121GW...hope that you don't hate its designers as well 

That's just your perception.  I haven't met the designers.  While the 121GW certainly falls short in much of my testing, I don't take it personal.  Some meters are just far more robust than others is all.  In this particular case, no one was hurt.   It did remind me of one of the first videos I saw of Dave's with what he called a near death experience.  Good stuff.

[Brumby]

Not sure about anywhere else, but here there are consumer protections which extend outside of any warranty.  If a product is offered with a 1 year warranty and a fault existed in that product at the time of purchase, it does not matter that the fault was not discovered until 2 or 3 years later.  The product was not fit for purpose at the time of purchase and (here in NSW at least) the consumer has the right, by law, to choose either a refund or replacement of a unit with a major fault - which is what the above example would seem to be.  (They cannot be forced to accept repair of a product that has a major fault.)

[vyruz]

I think something along those lines is the case here in Europe/Belgium as well, however, I didn't really 'purchase' the DMM, I just backed Dave's kickstarter campaign and got a DMM in return as thanks, at least that's my crude understanding of how kickstarters work. Dave did seem intrested in the serial number and purchase date a couple of posts ago, so I'll see if he gets back to me about a possible solution. I'd also be happy to settle for an affordable PCB replacement, if that would be an option. Anything really since now I've just got a very expensive paper weight.