Broken WaveTek DM27XT Multimeter

[corrado33]

Hi all,
I have myself a broken DM27XT. Now, let's get this straight, I have multimeters, I'm not trying to fix this to become my "main" meter. I'm doing it for fun. So, here's what happened.

No, I cannot find the service manual, the manual you find online just says "Only qualified dealers etc should repair it blah blah blah" Yes, the fuses are fine, not that I think that'd matter here.

The person who gave me the meter was trying to measure very high voltage (on the order of a few kV) through a voltage divider to see if the power source was working. Yes yes it's a terrible idea and what not but I wasn't there and she was only trying to troubleshoot something (that turned out to not be broken.)

It worked fine for the lower voltages 1000 kV or so, I believe she had a divider on there that knocked it down by an order of magnitude or so (so she thought). (Plenty high power resistors.) No, she wasn't touching anything, it was connected via alligators apparently.

So she turned up the source to higher voltages, and eventually the meter started giving weird readings (this is all secondhand by the way).

And eventually, she heard a buzz, let the magic smoke out, and her beloved meter stopped working.

Yes, I made her buy a fluke to replace it with and an actual high voltage probe up to 40kV, so no worries there.

The meter does actually still turn on in certain modes. The display pops up just fine  in voltage mode, or resistance mode, but goes blank in capacitance or inductance mode. (Doesn't measure anything in any mode however.) BUT, sometimes in between a working mode and a non working mode, the display will work, but be VERY dim, almost like it's not getting enough current.

I initially thought that something was shorted in those modes, but the current draw is next to nothing. The current draw in the DCV mode is 80ish mA. That draw didn't seem weird to me.

I looked over the entire board, nothing burnt, nothing busted.

So, I started reading input protection stuff. Well, I started measuring components around the volts inputs. After looking VERY closely, it seemed that a few of the solder joints are VERY shiny (and now look like cold solder joints) as compared with the tarnished solder joints on the rest of the board. I have no idea if they looked like that beforehand, but it stood out. So, I started measuring resistors. The first one I measured, a 5 line Yellow-Violet-Black-Gold-White. I'm assuming that's... 47 Ohms. Well, it reads 80ish. Not... too far off (still terrible.) So I read a few more, the next one was fine, but the third one I measured, a 200k (RedBlackBlackOrangeBrown) and... nothing? It didn't read at all, I got absolutely nothing. I don't know if my probes aren't sharp enough (they were for everything else) or if the resistor is just blown. (No burn markings on it or anything.)

I stopped there, for now. (I should be studying haha) Also, those few things I measured were the few solder joints that were weird.

After my test tomorrow, I actually want to try the amps mode (since it uses different input circuitry), then I'll desolder those few resistors (and there's a cap that has a similar weird shiny solder joint). There ARE test points on this board (well, two empty holes labeled TP1 and TP2.) They measure around 3.5 and 3.8 volts. (I have NO idea if that's relevant.) I also can't trace them because they go under switches that I don't want to desolder.

As for the meter itself, I can't honestly find that much info on it or WaveTek. Apparently they merged with some company that I've also never heard of (AeroFlex?). I've also seen online that fluke acquired some of wavetek's product lines? So I'm confused there. The meter LOOKS good (after watching all of Dave's multimeter teardowns.) It's a two board meter, connected by headers (well, three boards if you count the tiny board that the banana jacks are mounted on). The only bad thing I've seen is a cold solder joint on the current shunt. The date codes are around early 1995. Looks like the main chip is a UMC UM7108F. I'm assuming it's the main chip, it's the largest (and only) QFP on the board (and the datasheet says it's a multimeasurement something or other with some ADCs etc.)

So, if this post was a question, what do you guys with much more experience think happened?

[retiredcaps]

Now, let's get this straight, I have multimeters, I'm not trying to fix this to become my "main" meter. I'm doing it for fun.

Like yourself, I try to repair things to learn and for "fun".

I initially thought that something was shorted in those modes, but the current draw is next to nothing. The current draw in the DCV mode is 80ish mA. That draw didn't seem weird to me.

I'm no expert, but most multimeters I have seen and used only draw about 3mA or less when measuring DCV mode.  Even the backlight only draws about 30mA.  So drawing 80ish mA is a lot.

Are any components obviously hot to touch when the multimeter is drawing 80 mA?

PS. A clear focused picture of the pcb might help us help you.

[corrado33]

Ask and you shall receive.

I didn't know 80 mA was a large value for the current in DCV mode, I'll see if anything is getting hot.

Here are some pictures. Please excuse them, I am by no means a photographer. I also don't know how to use flikr apparently. I think the pics are big enough for now.
Full board.

Closeup of the volts inputs. One of the two grey resistors closest to the pin is the 47 ohm resistor (that measures 80) and the second blue resistor in from the right (close to where the one side of the fuse holder is) is the 200k resistor that measures 0.


Here is the bottom side of the volts input circuitry. You can see a few "domed over" solder joints. They're much more apparent in person. There is one opposite of the 47 ohm resistor I mentioned above.

[retiredcaps]

The battery life for the DM27XT is rated at 300 hours (9V) according to the manual.  Assuming a 9V battery is 400 mAH, that implies a 1.33 mA average.  So 80 ma is way too much current draw (implying 5 hour battery life).

To test all the input protection circuitry like the MOVs, spark gaps, PTC, and fusible resistor, I suggest testing them in circuit first.

A MOV and spark gap should read infinite resistance.  Most PTCs and fusible resistors measure around 1K ohms.  I circled all the components in the pictures.

Resistors usually fail open, not shorted.  If you are reading 0 ohms on the 200K ohm resistor, it may be in parallel with another component.  I suggest desoldering the 200K ohm resistor and measuring it out of circuit.

[corrado33]

Well, I think I've discovered the problem.

I replaced the 47 Ohm resistor and the 220k resistor (even though it wasn't bad, don't know why it wouldn't read last night. The cap is fine.

Replacing these made the capacitance and inductance ranges work!

But... it still draws 110 mA. So, I felt around to see what, if anything got hot. Unfortunately what got hot was the main processor. It got to 35C or so before I shut it off. I'm assuming that it's not supposed to get hot, and that something is shorted inside. Bummer. The funny thing is on the ohms range it'll zero out when you touch the probes together, but then rise when you take them apart, so it IS responding, just not in the way it's supposed to.

I think it's busted for sure 

[retiredcaps]

But... it still draws 110 mA.

Something is wrong if it draws 110mA.  If the 7108 is indeed fried, replacements cost around $20 USD which, in my books, makes it not economical.

I'll think about this over the next couple of days to see if we can somehow confirm that the 7108 is fried or some other component.

[corrado33]

Agreed, I don't consider it economical, especially since I was doing it for fun. Much better ways to spend $20. 

[amyk]

Isn't that just a standard '7106 clone?

[retiredcaps]

Isn't that just a standard '7106 clone?

I thought of that, but when I looked briefly last night, the datasheet for the 7106 and 7108 show a 44 pin and 48 pin package respectively.

[retiredcaps]

Just using some Ohms law quick calculations, assuming a 10V battery and 1mA draw, that implies a 10K resistance. Since you are getting 100mA draw from a 9V battery, that implies a 90 ohm resistance.  You might want to see if anything is around 90 ohms or less?

Since there doesn't seem to be a schematic, maybe look at the 7108 typical applications (page 6-46).

For troubleshooting shorts, you might get a few ideas from my fellow participants over at badcaps.

http://www.badcaps.net/forum/showthread.php?t=31683

PS. If you have no interest or time to repair this and your friend doesn't want the meter anymore, PM me.

[mirach]

hi, 
I have the same model !
I have a problem .. it is a burnt smd resistance and no longer reads the value, could you look inside the instrument, please? I don't find any service manual.. 

shuld be R18



thanks to those who can help me

[retiredcaps]

It looks like R18 is tied to pin 9 of the UM7108?  You can verify with a quick resistance/continuity check.

The datasheet for the UM7108 has an application circuit that shows a 24Kohm resistor.  I would try a 24Kohm resistor.

Pin 9 is V+.

[mirach]

It looks like R18 is tied to pin 9 of the UM7108?  You can verify with a quick resistance/continuity check.

The datasheet for the UM7108 has an application circuit that shows a 24Kohm resistor.  I would try a 24Kohm resistor.

Pin 9 is V+.

thanks for your response retiredcaps

Yes, it is connected to the V + pin 9.

I measured the resistance with an other Tester and R18 marks 25K ..

So.. what is the problem ?

The tester was powered with a tension of 17V.... ..... 
now turns on but there is always the symbol of low battery, and it doesn't  make measure

... do you think that I have to throw it in the trash?

Thanks a lot for you help and sorry for my english

[retiredcaps]

The tester was powered with a tension of 17V.... ..... 
now turns on but there is always the symbol of low battery, and it doesn't  make measure

Was that 170V and not 17V?

See if this thread provides any hints?

http://www.antiqueradios.com/forums/viewtopic.php?f=8&t=166996&start=30

If it always shows low battery, check current draw.

[mirach]

The tester was powered with a tension of 17V.... ..... 
now turns on but there is always the symbol of low battery, and it doesn't  make measure

Was that 170V and not 17V?

See if this thread provides any hints?

http://www.antiqueradios.com/forums/viewtopic.php?f=8&t=166996&start=30

If it always shows low battery, check current draw.

I used an external power supply to provide the 9VDC to the tester, because I had not  a battery and I needed a measure ... in a hurry.. but only after I realized that the tension was not 9VDC but 17Vdc

thanks for the link I read it immediately

[retiredcaps]

I used an external power supply to provide the 9VDC to the tester, because I had not  a battery and I needed a measure ... in a hurry.. but only after I realized that the tension was not 9VDC but 17Vdc

Okay, I see.  You mistakenly applied 17V DC into the 9V battery clips.

The datasheet for the UM7108 states 15V is the maximum.  You might have fried the IC?

What is connected to pin 40? (V-). The datasheet shows it connected to the BAT-, but on a 27XT, it might be connected to some other component first?

[mirach]

Okay, I see.  You mistakenly applied 17V DC into the 9V battery clips.

Yes, exactly. 

The datasheet for the UM7108 states 15V is the maximum.  You might have fried the IC?

Maybe the UM7108 is safe, I hope..

by connecting a battery, the display works but it doesn' t fit properly ..

What is connected to pin 40? (V-). The datasheet shows it connected to the BAT-, but on a 27XT, it might be connected to some other component first?

I followed all the tracks
In series from the battery clip to pin 40 (V) there is nothing.
In parallel there are:
- a diode D1that I checked and it measured (0.566) with diode test
- 22 uF x2 electrolytic capacitors that i have measured (18 uF)
- a track goes to the function selector..
- another to a PNP General Purpose Amplifier BSR16 and to a HEF4521BT


 

[retiredcaps]

Maybe the UM7108 is safe, I hope..

As I suggested, you should measure the current consumption of the 27XT.  You can put your other multimeter in series with a 9V battery.  Alternatively, you can hookup your external power supply and set it to 9V and see how much current the multimeter draws from it.

As I speculated earlier in the thread, most functions should take about 1.3mA.  If you see 80mA, then something is obviously wrong.

I don't know if 17V is enough to kill the IC or not.  If the 7108 is dead, getting a replacement is possible, but the cost of it is probably not going to be cheap.

- another to a PNP General Purpose Amplifier BSR16.

 May I have fired the BSR16?

You can easily test to see if it is shorted.  I usually just do a three way test between each pin to see if they are shorted.  That is, B to E, B to C and E to C.

[mirach]

I measured the current consumption of the 27XT


Selector                  Display                                    current consuption 27XT
Farad                          off                                                           5.78 mA
Ohm                            on                                                           3.64 mA
Henry                          off                                                            5.78 mA
Hz                                off                                                            5,78 mA
Volt                              on                                                            2.42 mA

Do you think that the UM7108 is broken?

[retiredcaps]

The current draw is slightly higher than what I expected, but it seems reasonable.

Since you are getting a display on some settings and not others, I suspect the 7108 is fried by the 17VDC.

Unless others have some suggestions, I am out of ideas.

[mirach]

The current draw is slightly higher than what I expected, but it seems reasonable.

Since you are getting a display on some settings and not others, I suspect the 7108 is fried by the 17VDC.

Unless others have some suggestions, I am out of ideas.

Also I think it is fried ..

What a pity .. it was a good friend since 2002

I will make a worthy funeral for him, may he rest in peace.

Thanks so much for your help and your patience

[retiredcaps]

The only longshot that might work is if you put a new 7106 44 pin IC somehow into the 7108 48 pin footprint. 

Something like

http://www.ebay.com/itm/251012139863

Obviously with 4 pins missing some functions won't work and the whole idea might be dumb to begin with?

Or you can spend more and get the right IC

http://www.ebay.com/itm/170956575753

The other obvious question is what else might be fried by the 17V?

[mirach]

My tester was especially useful for measuring the inductances. The other functions are not very important for me, because I have another economic tester.
The DM27XT MADE A fairly accurate measurement at 100 Hz for the inductors. It was useful to build the coils with magnetic core for the crossover of my speakers ..


The price of UM7108F is too expansive, the DM27XT may not work equally after his replacing!  On ebay for sale is a DM27XT for 50 dollars.

THE price of the ICL7106CM44 seems convenient to try a resurrection of the instrument but do you think that I can measure inductance if I replace the UM7108F with this chip?

[retiredcaps]

THE price of the ICL7106CM44 seems convenient to try a resurrection of the instrument but do you think that I can measure inductance if I replace the UM7108F with this chip?

I don't know.

[ModemHead]

The UM7108 and the ICL7106 are incompatible.  The 7108 uses a tri-plex LCD, the 7106 is direct-drive (unmultiplexed).