UNI-T UT61D AC reading issue

[hozone]

Hello,

I've a UT61D. I few times ago I've found it has AC reading issues.
As example
- 12V AC out from the sec of a transformer.Autorange reads 3.3V. Selecting the right range reads 12.48V (my other meter reads 11.8V)
- 220V AC mains. Autorange reads 5.5V. Selecting the right range reads 160V (my other meter reads 218V)

I've opened it last week, and resolder three components on the input side, things seems solved. Today I've to check the 12V transformer, and I've found that it does not work.

You can find the images of the input side for this meter on my last week post: https://www.eevblog.com/forum/testgear/uni-t-ut61d-genuine-bodge-diode/msg4654879/#msg4654879

Do you have any clue on what's happening?
Thank you!

[shakalnokturn]

Your other post doesn't give any pictures of the front side of PCB.
I think this multimeter is based on Fortune Semiconductor FS9922 IC (usually in black blob style), datasheet will give you something quite close to actual schematic.
Being a true RMS model there should be an extra IC for that. I've had DMM's giving low AC readings because of electrolytics in the RMS section.

[hozone]

Thank you!
I'll check the datasheet. At the moment find attached pictures of the meter.

[shakalnokturn]

I'm uncertain from your photo, could be worth cleaning the switch contacts (including PCB side) it can't do any harm as long as you don't scrub too hard.
Check solders on trimmers, measure 4 capacitors between VR1 & VR2.

[floobydust]

U2 (AD737J) does AC RMS and has poor soldering on one row it looks like, weird the other side is resoldered though.
Spend some time with a magnifier and look for bad soldering. Cracked resistors are also a possibility.

[hozone]

Thanks for suggestions!

Quick update.
Checked for crack, resoldered AD737 and surrounding resistor and through hole. Not solved.
Checked components value inline, seems all ok.
Checked Ca and Cc AD737 caps, also I've replaced them. Not solved.
Checked all signals in and out from AD737. Given the same AC source (12V AC transformer output), the only thing changing when the conversion is not working is the Vout of the AD737.
-- What's strange --
When I resolder component all seems to works (read almost 12V AC), then it slowly (10 min or more) keep to read wrong values again (almost 3V).
What I've notice: if I heat the components in the AD737 zone, even using an air gun on the other side of the board (i mean when it's mounted in place), the AC keeps reading right values... when it cools down reading keeps going wrong. I though it was the Cc or Ca cap of AD737, replaced them but I've not solved.
I'm starting thinking it's the AD737. What do you think?

[shakalnokturn]

Could be the AD737 or intermittent cut track / via.
Have you checked supply voltage is right at the AD737 GND / Vcc pins when not working correctly ?
If you have a soldering iron with variable temperature you could try pin-pointing AD737 applying (150°C?) directly to the top of the IC when readings are wrong see if that restores normal operation.

[hozone]

I could exclude 99% AD737 intermittent track, cause I've removed and resolvedered it.
Supply voltage are stable, the only PIN i can see change is the OUTPUT.
It's hard to try this cause there's the LCD on top of the AD737. I've heat all the area... what I can try is to heat AD737 from behind using the via that goes to Ca capacitor, that way I could heat up just the cap and the AD737, the cap is new, so I don't thing it's not subjected to temperature shift to induce the errors, that way the only variable is the AD737... I'll be home in a few, and I'll check.
Thanks.

[hozone]

I confirm you, If i warm the AD737 through the anode of the 10uF CC cap (pin 1 of AD737), the AC reading restore to the correct value. Then they slowly went down. Any tought?

[shakalnokturn]

At this point I'd change the AD737...

[hozone]

Thanks!
I've ordered a couple... let's wait for the delivery.

[hozone]

Update: I've replaced AD737.
Long story cause they takes a lot to be delivered. First time the seller send me wrong IC, refund and send me again... IC was fake Third time I've ordered from a reputable seller, I've replaced it and same story.
BUT i've notice that breathe on the back of the board makes reading ok again. So it could be a resistor... I think one on the input amplifier gain could be the guilty. I'll check it as soon as possible and let you know.

[Kleinstein]

The describtion of the error sounds a bit like am amplifier problem with the DC biasing. So something like a resistor used for DC biasing of an AC coupled amplifier is open or has a bad solder joint. As the readings gets low, this should be close to the input, with still another AC coupling capacitor in front of the AD737. So the prime suspect would be the biasing of the first AC coupling, directly towards the input.

[hozone]

Thank you,
ideed I want to take a look at the biasing resistor of the AD737.
Based on the FS9922 datasheet (https://www.ic-fortune.com/upload/Download/FS9922-DMM4-DS-13_EN.pdf), diagram 20, I want to check the 90k + 50k + 200k input circuitry, which in my image on the post above it seems to me R6 and related, but I'll take a look later when I'll be back home. Finger crossed...

[Kleinstein]

The suspect parts would be the 200 K and the lower end of the 50 K trimmer.
Another possible weak point could be the negative supply.

[hozone]

Sorry for my delay and thank you for your help!

Think I've fixed it.

I've check the resisor as you suggested, they are all ok. Something else comes in to my mind. Because the heat of a 3 times breathe on the back of the multimeter is enough to keep readings go, and the reading stays right for a couple of hours my suspect gos to some electrolitics. There must be something chemical cause if it was a resistor the heat will dissipate in a few minutes.
I've replaced the one of the input stage, no luck. Then the both of the power supply rails. And now it works. I have left it unused for a day, and now when i turn it on AC mode it reads good.Damn electrolitics!

[Aurgelme]

Hi, it would have been nice if you could tell us which capacitors, number marking and possibly the value, were the cause of the problem.

I have two pcs. UT61D, where one of them stopped measuring ac yesterday, and number two stopped today. Only measured ac below 30V.


EDIT

One of the meters only had a low voltage battery.

I had to open the second meter and found some of the ugliest soldering I have ever seen, as well as that the pcb was not cleaned and had to pick away dozens of tin balls scattered all over the pcb and even on the edge strip/connector to the lcd display there was a tin ball .
Cleaned the pcb and resoldered the AD737 and most of the surrounding components, and the meter works fine.

[hozone]

Hello Aurgelme,

forget to mention here, I've made a post about this repair, you can find it here: https://davidegironi.blogspot.com/2023/04/uni-t-ut61d-ac-mode-repair.html
To me it was the power rail electrolitics of the AD737, but I forgot to mark them on the PCB, sorry about that.

Good to hear your now works. Yep soldereing it's not that god on that PCB.