121GW strange range/socket problem

[Morg]

Hey,

my 121GW has done splendidly for some time now (190505xxx serial), but today I got irritated -

while measuring current (expected ~3.5 mA), it consistently showed somewhere between 300 and 400 mA. A second analogue multimeter confirmed the current being around 3-3.5 mA.

Later I tested the 121 by connecting to my PSU (current limited to 50 mA), and independent of selected range the displayed current always was at around 5 A. (Mind you, connected to the socket with a 400 mA fuse...)

After using the leftmost socket, it hat the same reading as the PSU at 48 mA over all ranges.  A quick crosscheck again with the second socket showed the error again.

I did accidentally enter CAL mode by holding MODE on switching on, but I didn't press REL to enter calibration but immediately shut down. Just to be sure, I restored an old calibration backup with no effect.

Any ideas?

[electr_peter]

Just an idea - did DMM showed wrong current reading when mode selector switch was set to different range than current input socket? Say, uA range on selector switch and "A" current socket or A range and "uA/mA" current socket

[J-R]

The 121GW uses jack detection to switch between current modes.  Maybe the jacks need to be cleaned?

[Morg]

What irritates me is that it displays almost exactly the 100-fold value, so 5000 mA instead of 50 mA. Led me to believe this could be a range problem as Peter mentioned, but there is no 100x range jump between sockets, only 1000x.

Maybe I'll try reflashing the software...

[J-R]

Due to a lack of other ideas, I still would suggest investigating the jack detection.  I was able to generate some results adjacent to your issue by having leads connected to both current jacks at the same time.   I put 50mA through the A/500mA jack then inserted a lead into the mAuA jack and it displayed 5mA instead.
Here is the 121GW schematic: https://www.eevblog.com/wp-content/plugins/download-attachments/includes/download.php?id=11618

[Morg]

Jack detection works fine; plugging an additional jack (not connected) while measuring on either jack doesn't change anything.

Not surprisingly, flashing firmware again didn't change anything.

Concerning the schematics - I couldn't find an obvious defect causing this behaviour. Any voltage divider would have to be way out of spec to result in "proper" 100-fold voltages (as in 50 mA -> display 5 A, 25 mA -> display 2.5 A etc).

[J-R]

I reproduced the issue exactly by sending 50mA through only the upper half of the mA/uA jack while leaving the lower half open. 5A displayed.  So it's definitely the "Input Warning Circuit" not working correctly on the mA/uA jack causing it to think you're connected to A/500mA.  Inspect the jack itself and then expand to the surrounding PCB if necessary?

[Morg]

You're absolutely right - connecting only the upper half yields 100x the value; connecting only the lower half doesn't yield anything. Not sure what I'm achieving by inserting a 4mm banana plug and bending the top end towards the 121GWs lower end, but with enough - soft - force I get correct readings.

So this sounds a bit like a bad contact, having problems connecting both sides at once? Any experience which (mechanical?) work might remedy this? Bending the "half tongues" (or the lower one) - but how? It's quite tight in there, and I'm not sure how to remove the main pcb without damaging the central knob fastener.

[J-R]

The PCB has a tendency to hang on to the selector shaft, so just wiggle it while pulling and it should come free.  Good teardown photos here: https://lygte-info.dk/review/DMMEEVBlog%20121GW%20UK.html
Essentially, you remove the battery cover, back cover, the black PCB screws (not the LCD ones), then the PCB.
And to clarify, you do not remove the black portion of the selector from the PCB.

[Morg]

Did some disassembly, but found no apparent error. Tried resoldering the jack and the black thing next to it. Could not find a way to bend the jacks inside contacts, especially as I didn't find a tool slim and strong enough to get it between the metal and casing.

Symptom persists, but can - partly - be remedied by bending the jack upward.

Any further hints? Knowing source and symptom, it will surely be possible to just work around the issue, but for equipment of this class and price, this seems unsatisfactory...

[electr_peter]

Have you tried different banana plugs? Some 4mm plugs are stiffer and wider, maybe it will help.
Also, try inserting unshielded 4mm plug while DMM is disassembled - cracks could be visible, if there are any.

[J-R]

Which test leads are you using, the Brymen ones with the lantern style or the UEi ones with the 4-way split?

[Morg]

I have only test leads with these jacks -> https://media.rs-online.com/image/upload/bo_1.5px_solid_white,b_auto,c_pad,dpr_2,f_auto,h_399,q_auto,w_710/c_pad,h_399,w_710/F3469064-01?pgw=1

At the moment I don't have alternatives to test with, but up to now I never had contact problems (also not with the 121GW...)

I'll try the hint with the cracks...

[J-R]

Mechanical issue, should be easy to resolve with some more tinkering.  If it's a cracked jack Dave has replacements available.

[Morg]

I can't see cracks on the jack in question for sure. I can see the whole jack fixture is cracked right in the middle, though I'm not sure if this is relevant to my problem.

Is there a special way to ask for replacements or should I just write to Dave?

[tautech]

I can't see cracks on the jack in question for sure. I can see the whole jack fixture is cracked right in the middle, though I'm not sure if this is relevant to my problem.

Is there a special way to ask for replacements or should I just write to Dave?

Report yourself to the moderators with a @ Dave, which will get his attention.

[J-R]

The replacement jacks were for units where the cylindrical portion split apart due to a plastics contamination.

If the jacks themselves are not cracked, then I would focus on finding a loose connection such as where each jack is soldered to the PCB.  There are two solder points on each jack, 180 degrees from each other.

Can you post a couple photos?  Front and back of the bottom half of the DMM PCB.

[Morg]

You can see the crack on the backside; I can't see any other (obvious) cracks.

The jack for mA/µA has already been resoldered, but to no avail.

[robert.rozee]

using a second multimeter, check for continuity between each of the two solder joints for the offending jack and the associated half-round inside of the 4mm hole. i'm picking you will find continuity on one, and not on the other.

the break will be within the molded plastic, where the half-round narrows down to the 1mm wide strip that emerges and goes to the solder pad. apart from replacement, the only way of fixing will be to dig down into the plastic from the back until you get to the edge of the half-round and carefully solder on a wire. since it is the 'sense' contact it will be carrying next to no current.


cheers,
rob   :-)

[J-R]

Those jacks appear to be in excellent shape.  I don't see how they could be the issue.  But still agreed, start by verifying continuity on each jack, and then follow the circuit, checking for loose solder joints or some other type of crack.

[electr_peter]

As mentioned above, measure continuity on all connections. All terminals are exposed on the back side, almost no plastic cutting should be required.

[Morg]

Measured continuity between inner contacts and outer terminals on the back of the jack, outer terminals and solder joints - nothing (means all are connected...).

Guess I will dig into the terminals anyway, as mechanical stress on the terminals could trigger the (mal)function in the first place. Will also check further into the pcb wiring, though I don't expect anything there...

[J-R]

If the jacks test OK, then for sure look at the PCB and components for a while instead.

Some common techniques are pressing/flexing the PCB using a plastic poker of some sort, and heating/cooling the PCB & components to see if that helps to narrow down the fault.  Also, SMD parts can crack themselves, not just the solder joint.

Also, try to reproduce the issue with the leads connected to the jacks, and perform the continuity tests at that point.

[Morg]

On the jack detection side everything seems fine with/without connected lead up to the OP-Amp (black brick next to the jacks, present next to uA and A jack). I'm neither sure what it really is, nor where and how it's soldered (I see two solder joints at the sides, but they look as if they are really soldered to the PCB below the part?).
I removed part of the back of the jack's plastic, but didn't dare "dig deep"; anyway, measuring from inside the jack and alternative from a plugged lead every contact point leading from the jack to the "jack detection" side measures okay so far. As soon as I plug a lead, both jack contacts are well connected.

[J-R]

If flexing the jack a certain way solves the issue, then just keep doing that while you hunt?  Also, it looks like the sense connection for the mA/uA jack goes into an inner layer of the PCB then goes under the selector?  Does pressing on the selector make any difference?