Uni-t ut61e ohms/capacitance

[algorithm]

meter bit the dust. Everything works but resistance and capacitance.
Been struggling to get anywhere fixing this. Ive tested lots and found nothing that made me sure.
So on resistance I get resistance with the leads separate. I think around 15k but it varied. Its not self zeroing anymore to OL.

On capacitance mode I get a capacitance reading with leads separate. Iirc 0.230 µF. Well my meter used to not have that reading. I think it would show OL until testing capacitance. But I read many people saying their meter had Nywhere between 0.50-0.300 µF but if anybody could clarify on that? Guess its irrelevant if testing is accurate but it appears like it wouldnt be due to the stray µF?

So I have a spare meter of ut61e. Comparing between boards the following differences. But the boards themselves arent identical. Working ut61e ommits r36a-d and r39 at least. Also different placement of transistors footprint and some side waffling. Or pancaking rather cause theyre flipped.like r29a-r32.
Bad --------------- good
D5 and D6 give short beep. But removing thrm they test good. Q5 and q6 also made me suspect them with beeps but they were good off board. On good they dont beep even on board lol.

On ca+ it reads 47k for bad. On good 220 ohms. (I dont have a meter to test pF but I thinks thats supposed to be 1pf?.)



I get 3.6 M ohm resistance between common and r6 and that yellow line wiper track on bad
5.16M on good.

On q1 emitter that has continuity to common, testing from collector and base to common post was 3.5M ohm on bad. But now after lots of testing that resistance is missing from q1 and q8 on both boards. Weird.

Do any of those things give any clues? I again am limited by my meter only reading so much capacitance. (Fluke 12).

I tried reading resistance on the small capacitors to do with ohms. Like c7, c13, c14, c15 and c16.
But I saw such random results I couldnt trust it I felt. But testinv now my fluke does show a difference on c13-c16. On bad its 7.5 M, on "good" its 3.65M. Weirder I saw that but opposite earlier. The bad had 3.65M. And the "good" 7.65M. So I have no confidence or trust in that. But I figure I might have triggered something via transistors? To have seen it both ways both boards... I put good in quotations because Ive not confirmed the newer ut61 is calibrated perfectly. But its a unit out of the box.

thanks for any advice or ideas.

[Per Hansson]

Did you actually test it when you removed the protection devices Q1/Q8 and Q5/Q6?
I suspect the fault will be there, if so removing them should make it behave again untill you can buy some new replacements, take a look at this thread:
https://www.eevblog.com/forum/repair/uni-t-ut61e-diode-mode-repair/

[algorithm]

Did you actually test it when you removed the protection devices Q1/Q8 and Q5/Q6?
I suspect the fault will be there, if so removing them should make it behave again untill you can buy some new replacements, take a look at this thread:
https://www.eevblog.com/forum/repair/uni-t-ut61e-diode-mode-repair/

Well before reading this I only took q5/6 out since they connected to the 2 diodes I got short beeps across. But yeah now have taken q1/8 off as well and they both have 2 diode junctions. 

Took d1 off, but it tests good too... crikey

Yeah I had read your topic. Was a pleasant surprise to see a response from you. Reax every topic I think lol. But yeah one thing though I dont have a smd tester like yours so im probing by hand. Might be missing somethjng my way. Well guess if I take enough parts off eventually will fjnd it.

[Per Hansson]

But yeah now have taken q1/8 off as well and they both have 2 diode junctions. 

Isn't it working ok now then after you removed these faulty transistors, surely something must have improved?

[algorithm]

But yeah now have taken q1/8 off as well and they both have 2 diode junctions. 

Isn't it working ok now then after you removed these faulty transistors, surely something must have improved?

So are you saying that there shouldnt be 2 diodes?
Like in your linked topic tests. A junction from b to e and b to c? It should be like your original q8 test?
My tests match your final photos where base has a diode both ways. If those were the failures pictured thenhopefully thats solved. I dont have smd 8050 but I think I have some through hole. Probably should use the right thing though. Might need to make another part order.

[Per Hansson]

But yeah now have taken q1/8 off as well and they both have 2 diode junctions. 

Isn't it working ok now then after you removed these faulty transistors, surely something must have improved?

So are you saying that there shouldnt be 2 diodes?
Like in your linked topic tests. A junction from b to e and b to c? It should be like your original q8 test?
My tests match your final photos where base has a diode both ways. If those were the failures pictured thenhopefully thats solved. I dont have smd 8050 but I think I have some through hole. Probably should use the right thing though. Might need to make another part order.

Yes to the bolded part!
Q1 in my reply #1 with the extra diode from E to C is failed, this is why the meter always reports a diode in testing, wasn't this clear in my text?
You can test your meter without these components, just don't hook it up to any high voltage things because there will not be any input protection any more!
And in all honesty the meter is pure crap so even when you have replaced those components don't hook it up to any high energy circuits as you can see in my last reply to the thread you will just end up repairing the meter more than the equipment you are actually working on!

[algorithm]

Well bad news tested a new throughhole 8050 and those tested the same as what came off the board, both q1 and q8. So still no idea.

[algorithm]

Anybody know or able to find the value of ra+?
Both versions of pcb I have have it but that schematic seems to be missing it. Also r39 isnt there but its on the older of 2 pcbs...

So far thats the biggest difference I could find. Is ra+. Was hoping to get lucky with transistors or diodes, but those all seem good. Also now on d5/6 on both boards, I get a short beep both ways despite the diode being good. Can anybody confirm or deny if those diodes traces have 9 ohms across both pads?
Because Im pretty sure I tested the newer and older, initially newer didnt have a short beep. But at this point ive tested every resistor and everything else I knew how to a few times across both. Also both boards are slightly different in layout so might have been my confusion in the chaos of desperately trying to fix this.

anyway thanks

[algorithm]

My dumbass finally got the bright idea a few minutes after posting to read the resistor via magnifier.

Those values were correct for the part. Both ra* were different. Older pcb ver with r39 has 47k 473. Newer has 220 ohm.

Hmm so I started trying to test around the relevant fc track wiper vias. That led me to c35 and c36. Well the 1a fuse first. On said fuse I had from 1 whole ohm up to 10 dancing all around. C35 tested solidly, c36 though was also dancing a lot. From the expected range of 17 microfarads up to 60. The tc1 tester says its a diode. So hopefully this will fix it.

On c35 I get 34µF. On c36 get 16µF. But on bad board c36 sometimes shows open or random values.

Could w1 be causing this? Sometimes the capacitor value is correct and I have to wonder if it will be correct off circuit.

[algorithm]

Took those caps off. Still no good... still a shit load of resistance on the 1A fuse. I think I saw even more ohms after cleaning those caps up.

But did notice c35 had resistance that steadily decreased, when it reached a certain amount of ohms it went back up to like 47k and started counting down again.

I wish I had the ut61e to test the ut61es caps. But 1 is missing a lcd so...

[Per Hansson]

The tc1 tester says its a diode. So hopefully this will fix it.

Just make an adapter for your TC1 to be able to test SMD components?
A PCB prototyping board + 3 pin headers would not take 5 minutes to solder up, or just solder some legs on the SMD transistors?

[algorithm]

On resistance I get 65M but it starts at like 120M and counts down.
Strangely after I replaced the 1A fuse it TRIES to OL but after displaying OL it shows resistance even without probes and counts down to 65M.

Milivolt range does the same but it claims the numbers are mv. Also saw it show ridiculous opposites like -60 from 100. Jumping around but initially counting down.

Capacitance mode shows 0.50 so I guess thats alright since others said thats normal and nobody corrected it.

Still dont know what to blame... any of what I described point to anything in the eyes of someone who knows? This is really bothering me and Im past my wits end.

[algorithm]

Got the idea to test the thing powered by clipping a battery on.

Got 0.190v across com plug hole and volt ohms hz plug hole.
Also was seeing about that much on 3 r36s. 1 only on 1 side, the other 2 on both sides.

I had already halfway desoldered that 200k resistor to test. It tested good. Should I fully take it off? Or is it going to be explainable by the circuit why it had that on both sides of resistor?   

R36c had 0.180 on half. R36d had it both sides. And r36A had it both sides but sometimes didnt.
Actually r36a was getting 0.230v.

And I saw that 0.230 going into q6 3rd pin if the single pin side is 1
R36a tested 202k, would that be enough to cause problems?

The tc1 tester says its a diode. So hopefully this will fix it.

Just make an adapter for your TC1 to be able to test SMD components?
A PCB prototyping board + 3 pin headers would not take 5 minutes to solder up, or just solder some legs on the SMD transistors?

What voltage should I see on q1 and q8?
Or q5 q6?

I did try that but failed at the 3rd wire. Than trying to geg 3rd wire reattached 2nd came off. But it still tested the same on dmm. Of course oblivious of vbe or hfe. But yeah.

[joeqsmith]

From what I have seen, UNI-T in particular are more sensitive to ESD.  Normally with the 61E, it damages the controller IC.   I did a whole series of videos on the 61E explaining the problems with it.  Hopefully your IC was not damaged.

***
Also, I have a few friends who purchased 10 or so UNI-T meters around the same time.  I think we have seen four switch failures so far.   

[algorithm]

So c21 measures OL. I thoughf haha finally. But took it out and replaced it with a new good esr cap. Also OL. Damnit!!

Does that mean the ad737 is to blame? U2. Bono got me again. Itz funny, cause I still havent found what im lookjng for.

[Per Hansson]

What voltage should I see on q1 and q8?
Or q5 q6?

Please watch EEVblog #1157 that I linked from my UT61E diode mode repair thread.
The circuit formed by Q1/Q8 and Q5/Q6 is only there to provide a dead short to allow the PTC's to increase in resistance and actually protect the meter.
As you can see from the schematic they are connected ONLY to the emitter: so no voltage should ever be present on the other side of the junction.
That is unless you do something "horrible" and hook the meter in Ohms, Capacitance, Frequency/Duty cycle or Millivolt mode up to a voltage source higher than the VEBO rating of the transistors.
In that situation their purpose is to draw some current so that the PTC's in the schematic heat up and actually limit the current by increasing their resistance.

I hope it is clear now that there should be no voltage on their collector side in normal situations, but also that if they fail the meter will always see a diode for example in these test modes, referring to my reply quoted here:

Q1 in my reply #1 with the extra diode from E to C is failed, this is why the meter always reports a diode in testing, wasn't this clear in my text?
You can test your meter without these components, just don't hook it up to any high voltage things because there will not be any input protection any more!
And in all honesty the meter is pure crap so even when you have replaced those components don't hook it up to any high energy circuits as you can see in my last reply to the thread you will just end up repairing the meter more than the equipment you are actually working on!

[algorithm]

I see like -0.080v on q8 and q1 b/c.
-0.060v on q5 and q6 b/c pins.

Also saw the same amount I saw on r36s going into emitters of q5 & q6. Also q8 and q1 iirc.

Weird thing is my diode mode doesnt read a diode. It zhows OL   
So I still have to wonder if the faults the same.

That 1a fuse I replaced was dancing all over upwards of 50 ohms even out of circuit. So something fried that. Bad thjng is I dont know because I let someone borrow it for work.

Well thanks for your help and image per hansson. Hopefully you are right and those transistors will fix it.


Gonna hVe to order some though. so more time wasted.
I have to say I hooked my meter up in diode or resistance to ac and dc/charged caps like 7 times before it fjnally died over the years of having it. That was 120v mains stuff but still a testimony of how strong these meters can be. Also I cant afford a better meter than the ut61 so I wouldnt say its crap. But if it doesnt work its crap. Lol

Thanks again. And hopefully when I report back the meter is just dandy.

[Per Hansson]

Weird thing is my diode mode doesnt read a diode. It zhows OL   
So I still have to wonder if the faults the same.

What the meter shows will depend on the type of failure of the transistor, see my last reply in the other thread.
Notice how the transistor tester reports a capacitor inside the transistor? (referring to file: Q1-2021.jpg)

Well thanks for your help and image per hansson. Hopefully you are right and those transistors will fix it.

Gonna hVe to order some though. so more time wasted.

I've said this before: you don't need to have Q1/Q8 and Q5/Q6 in circuit for the meter to work, just remove them.
If it works after that you found the fault, and know what parts you need to order to restore the protection.
If not you just need to keep looking.

I have to say I hooked my meter up in diode or resistance to ac and dc/charged caps like 7 times before it fjnally died over the years of having it. That was 120v mains stuff but still a testimony of how strong these meters can be. Also I cant afford a better meter than the ut61 so I wouldnt say its crap. But if it doesnt work its crap. Lol

Thanks again. And hopefully when I report back the meter is just dandy.

In my opinion the meter should survive this easy to make mistake, I never saw a Fluke fail due to such mishaps!
The Brymen BM235 is a pretty cheap and decent meter if you want an actual recommendation for something that doesn't break the bank.

[algorithm]

Another weird thing I noticed while testing virtually every part imaginable.

I get no resistance / continuity on r28.
On the pcbs pads specifically. If I move dmm probes up on the wire theres resistance, but more than 0.01.

Could q1-q8 cause that?
Could r28 being open across the pads cause the resistance issue?
Or possibly 1 of the 2 adjacent diodes?

Ive ordered the transistors.
Even if those arent the problem good to have sot23 variety in the parts stock.

I havent had a chance to remove transistors but I left the meter on accidentally after my last test. Like 30 min.
I tested the voltage on q8 collector which ties to q1 collector but q8 had -0.107v
Looked at the screen and 107Meg ohm was there.

Not sure if that definitively says anythjng but I figured it was promising. Not a coincidence at least. I hope. Lol

I watchdd thd eevblog 1157 on clamps with bjt. Interesting. Some of it went over my head or left me a tad puzzled, but the whiteboard shows 0.60v so I was wkndering since I saw -0.060 on q5/q6 would that entail they are good. Or am I just sadly mistaken there by assumption.

Sorry for my ignorance

[Per Hansson]

Another weird thing I noticed while testing virtually every part imaginable.

I get no resistance / continuity on r28.
On the pcbs pads specifically. If I move dmm probes up on the wire theres resistance, but more than 0.01.

Could q1-q8 cause that?
Could r28 being open across the pads cause the resistance issue?
Or possibly 1 of the 2 adjacent diodes?

No, the Amps and Milliamps input has nothing to do with the Ohms/Voltage/Capacitance modes.
Please watch this EEVBlog video: EEVblog #373 - Multimeter Input Protection Tutorial:

Ive ordered the transistors.
Even if those arent the problem good to have sot23 variety in the parts stock.

I havent had a chance to remove transistors but I left the meter on accidentally after my last test. Like 30 min.
I tested the voltage on q8 collector which ties to q1 collector but q8 had -0.107v
Looked at the screen and 107Meg ohm was there.

Not sure if that definitively says anythjng but I figured it was promising. Not a coincidence at least. I hope. Lol

As I said in my previous post: "As you can see from the schematic they are connected ONLY to the emitter: so no voltage should ever be present on the other side of the junction."

I watchdd thd eevblog 1157 on clamps with bjt. Interesting. Some of it went over my head or left me a tad puzzled, but the whiteboard shows 0.60v so I was wkndering since I saw -0.060 on q5/q6 would that entail they are good. Or am I just sadly mistaken there by assumption.

Sorry for my ignorance

The only time there will be voltage on the other side of the transistors is if you have an active overload condition, or if the transistor is leaky (broken).
Your time would be better spent making a testing PCB for SMD components for your TC1 transistor tester, here is one good example:
DGW Homemade SMD to DIP adapter:

[algorithm]

Took off q1, q2, q5, q6, d1 (because if was connected to blown 1a fuse with high resistance undulating)

But still shows a value of megaohms in resistance. It flashes like 190 or something then goes like 60M generally.

I get like 7M ohm across com and v posts.

God damn it.

Well haVe to keep looking.

[Per Hansson]

Yea that sounds wrong, I posted this in another thread.
You can follow the link in the quote if you want more details.
You can compare the values to yours, just make sure you are in DCV mode for the tests like mentioned:

The meter must be switched to the DCV mode to measure the input impedance.
Here an example of two meters:

Code: [Select]

	UTE61E:	AN8009:
OFF:	25.60M	36.4M
1V:	11.15M	10.982M
10V:	10.13M	10.085M
100V:	10.04M	9.995M
1000V:	10.03M	9.985M

[algorithm]

Took all r36s off. R36a was 202.x and 1 seemed to be open on the lower half of pad.
But replacing all 4 with new 1% 1206s fixed nothing.

With r36s removed the resistance was constantly jumping in ohms. It just flashed values up or down.

If I probed parts of the trace for voltage with fluke and looked at uni-t display certain parts brought the resistance to 7M. Like r33 iirc. But r33 is what its supposed to be, at least in circuit (iirc 100k but its been a few days)

Also a few timewhen I seemed to probe 1 of the main electrolytics it showed 0L. But returned shortly.
I replaced those caps so dont think it was the capz. Thatz just what I tested voltage on to look at uni-t

[algorithm]

Think I found it. Luckily thought to check the cyrustek datasheet for es51922(i think?)

Ohmc1-4 pins has filter caps respectivy. C15 was weird where in one polarity it just flashed 0.00µf.
Reversing polarity zhowed.. yaagh I dont remember... like .168.
C7 also seemed possibly suspect. Although tested identical to other pcb.
Which for that matter in circuit tests all seemed kosher between the 2. C15 even.
After removing c15 and c7. (Which c7 tested .5nf . But should be .3 apparently. Is that acceptable or could that be problematic?)

C15 seems to test .100 nf over or so. But it has resistance across it. About 70M. Same as I last zaw on meter readout on ohms.

I dont have any smd caps at 330nf. I do have ceramic and mylar throughhole. Could I use that? Or should I not?

Thanks in advance for advice on those queries and much thanks to hansson. I appreciate your help.

[Per Hansson]

It would not hurt to bodge in a through hole ceramic capacitor for testing, it might not be as accurate but fine to rule things out...
I'm guessing that the OHMC capacitors should be of type C0G also known as NP0 for good temperature stability.
Capacitor C7 sits directly on the input, protected by the transistors we tested earlier, it could be that the value in the schematic is wrong too.
I would replace C15 and put back the original C7 and see how the meter behaves.