Multimeter Uni-T UT61E Problem (Solved)

[ptluis]

Solved, see my last post

Hi guys
This problem started to happen a few days ago. I have a problem with the UT61e multimeter making measurements in the volts mode. The measurement of any DC or AC voltage (auto mode is slower or manual mode) takes about 3-5 minutes to stabilize and display the real voltage value. For example measuring a 9v battery, measurement starts at 9.53v and goes down slowly (3-5 minutes depending on room temperature) until it reaches the stable value of 9.32v which is the real value compared to other multimeters. This problem occurs when I start the multimeter for the first time and only after 10-15 minutes of keeping the multimeter on, the measurements are fast and normal within 1-2 seconds. The rest of the multimeter's functions work perfectly by measuring correctly and within 1-2 seconds. This is annoying because if I want to make a quick measurement from off position, I cannot get a stable value right away. This multimeter is only used for low voltage readings.

Things I've done without success:
- I turned off RS232 because the readings were even slower, about 2 minutes more
- Cleaning the entire flux on the pcb
- Switch cleaning
- I cleaned flux from the main chip ES51922 (improved a little)
- I tested the related components according to the schematics and did not detect any problems.
- New battery
- Check/tested voltage trimpot
- (update) Resolder all pins of ES51922
- (update) Try to find overheating capacitors or other components, none detected
- (update) Inspect all pcb soldering with optical stereo microscope

I suspect that the ES51922 chip will be defective because after 10-15 minutes on, the measurements work perfectly. I noticed that the ES51922 chip is warm after 10-15 minutes on. when warm works well, cold does not.

If it is the ES51922 chip, I start thinking about buying another multimeter, but certainly won't be Uni-T, maybe an owon B41T + but I have to read reviews first. The price of an ES51922 chip costs almost as much as the multimeter.

Does anyone have an idea of ​​what it might be?

[rsjsouza]

Indeed it seems you did a lot already. One last thing I would do is retouch the solder on its pads - maybe some cold joint is getting in the way of proper work.

I would perhaps do this only after buying another meter, so I wouldn't be stuck.

I don't know what is available in Portugal, but some cheaper and interesting options to the Owon are the Zotek ZT-219 (also Aneng AN870, Richmeters RM219) or the newer Zotek ZTY (also Bside ZT-Y, Richmeters RM777, Aneng SZ18)

https://zotektools.com/products/

I did some tests on my channel (in portuguese) of both the RM219 and the first impressions of the ZT-Y.

[wraper]

I noticed that the ES51922 chip is warm after 10-15 minutes on

It should not become warm since its current consumption should be only 2mA. Not necessarily failure of IC itself, might be some surrounding component, like shorted MLCC.

[PKTKS]

true..   try the alcohol  method to spot the hottest component..
or petrol or IPA...

Careful inspection of the evaporation will tell you the most likely culprit

If the main chip...   it really costs almost a brand new one

Paul

[ptluis]

Indeed it seems you did a lot already. One last thing I would do is retouch the solder on its pads - maybe some cold joint is getting in the way of proper work.

I would perhaps do this only after buying another meter, so I wouldn't be stuck.

I don't know what is available in Portugal, but some cheaper and interesting options to the Owon are the Zotek ZT-219 (also Aneng AN870, Richmeters RM219) or the newer Zotek ZTY (also Bside ZT-Y, Richmeters RM777, Aneng SZ18)

https://zotektools.com/products/

I did some tests on my channel (in portuguese) of both the RM219 and the first impressions of the ZT-Y.

Hi thank you for your answer. I indeed resolder all chip pins, forgot to mention that above. Didn't solve. Then inspect them and all other components on the pcb with a stereo optical microscope.

 
I'm looking for a 22000 count or better. I think the models you mention are 19999 count but I will consider them if price is good. Normally I won't buy equipment in Portugal due to inflated prices, prefer to order from other EU coutries or China even if they came with 1 year warranty.

[wraper]

You first need to check its current consumption. When you see IC becoming warm, blindly reordering all of its pins is not the smartest idea.

[ptluis]

You first need to check its current consumption. When you see IC becoming warm, blindly reordering all of its pins is not the smartest idea.

I updated list of tasks I've already done. The only component heating is the main chip. I resolder all pins of the ES51922 using a stereo microscope and 45x zoom to make sure they are perfect soldered. I'm thinking of removing the chip and inspect the pcb under it, but only when I get my hands stable again.

[wraper]

You first need to check its current consumption. When you see IC becoming warm, blindly reordering all of its pins is not the smartest idea.

I updated list of tasks I've already done. The only component heating is the main chip. I resolder all pins of the ES51922 using a stereo microscope and 45x zoom to make sure they are perfect soldered. I'm thinking of removing the chip and inspect the pcb under it, but only when I get my hands stable again.

Yet you don't have any actual measurement figures, only subjective "feel". IC heating does not mean that IC is necessarily problematic. Some part may be shorting one of IC pins, thus drawing current and causing it to heat up.

[ptluis]

You first need to check its current consumption. When you see IC becoming warm, blindly reordering all of its pins is not the smartest idea.

I updated list of tasks I've already done. The only component heating is the main chip. I resolder all pins of the ES51922 using a stereo microscope and 45x zoom to make sure they are perfect soldered. I'm thinking of removing the chip and inspect the pcb under it, but only when I get my hands stable again.

Yet you don't have any actual measurement figures, only subjective "feel". IC heating does not mean that IC is necessarily problematic. Some part may be shorting one of IC pins, thus drawing current and causing it to heat up.

Hi thank you for your answer. what would be the proper way of testing this chip without lifting the pins or cutting the tracks? btw do you know what is the normal consumption of the multimeter turned on? The chip you already said it was about 2ma in your previous post.

[wraper]

I don't think the rest consumes much more than this IC. Likely less than 5mA. Even 10mA won't make IC noticeably warm, so current should be quite significant.

[ptluis]

I don't think the rest consumes much more than this IC. Likely less than 5mA. Even 10mA won't make IC noticeably warm, so current should be quite significant.

Today the meter works just fine 
I measured the consumption of the multimeter and it's 4.7ma
I thought about what changed between yesterday and today and it was the outside temperature
Today the ambient temperature is +22º Celsius and the meter works well. Yesterday it was +17º and measure poorly.

So I decided to do the experiment by placing the meter in a box and then lowered the interior temperature below + 18º, I waited a few minutes to stabilize the temperature and ... the meter started to measure poorly again 

Then I increased the temperature to + 22º and let it stabilize for a few minutes and the meter works perfectly!

So seems temperature related??? When the meter measure ok the ES51922 chip stay cold, when measure poorly the chip stay warm 

I really don't understand what's going on here 


edit: I forgot to mention that the multimeter consumption when measure poorly is 18-19ma

[rsjsouza]

This is probably on the board - either a track, an inner layer or a via that is acting up.

It is definitely temperature related - the minute difference is enough to trigger this issue.

I would cool the DMM and warm specific parts using a hot air station that can blow a very narrow hot air stream.

[wraper]

Might be bad MLCC. When they become shorted, often short disappears at higher temperature. Or something may be oscillating at lower temperature.

[wraper]

Check voltage regulator W1, something might be going funny with its 3V output voltage. https://ibb.co/Wvn5myS

[ptluis]

I measured the resistance between the ground pins or ground connections on the chip side, pins 3, 4, 5, 28, 116 and the results are not zero, some with 0.02 ohms (pins 3-5), others with 0.04 ohms (pins 3-4). I'm using thin gold-plated multimeter wires. Is it possible that these small differences could somehow cause the multimeter to measure poorly? Another thing I'm thinking about is a pcb defect like rsjsouza  mention, but it looks like it could be the tracks below the chip, so I need to remove it to check, and this is an operation I can't do at the moment.  A defect in the pcb tracks could explain the increase in current and the heating of the main chip or am I thinking wrong?

Edit: between COM and pins also non zero, 0.05 ohms for pin 3, 0.08 ohms for pin 4 and 0.06 ohms pin 5

[ptluis]

Might be bad MLCC. When they become shorted, often short disappears at higher temperature. Or something may be oscillating at lower temperature.

I Will check that maybe apply some heat and cold to better test them, haven't do that yet

Check voltage regulator W1, something might be going funny with its 3V output voltage. https://ibb.co/Wvn5myS

measured on output pin = 3.26 V (bellow or above ambient temp of +22º )

[wraper]

It should be 3.0V if Vreg is as on the schematic. It could be normal only is 3.3V Vreg is used instead.

[ptluis]

It should be 3.0V if Vreg is as on the schematic. It could be normal only is 3.3V Vreg is used instead.

Can't say for sure what is this chip, there is no markings

[ptluis]

Problem solved for now.

I don't know exactly what was the problem, but I start thinking that since it was malfunctioning below a certain temperature it could be something related with moisture, so I place the PCB in the oven for 1 hour at 50º celsius. Now the multimeter work perfect down to 0º celsius    Note: first I remove all parts that could melt.

will it last? I don't know but it's working. I didn't notice any deviation in calibration. All functions work fine.

I also did a "mod" that made the multimeter to be less affected by external temperature and readings are now faster. I cover the "resistor bridge" (don't remember what's the English name), it's the black rectangle, with aluminium tape, and this makes the multimeter much more stable and unaffected by external temperature. On some models this "bridge" is shielded, on mine there's no shield.

Thank you all for you help!