[SOLVED] Tektronix DMM916 voltage issue

[kerouanton]

My first post ever, here!

I have a Tektronix DMM916 which is excellent. However since a couple of months I noticed an issue with any voltage measurement. Other functions seems to behave normally.

The voltage displayed (either AC or DC) is way below the correct value.
I've done at least 5-6 tests either using AC or DC, and the results are always bad.

real => DMM916
5v => 2.28v
10v => 4.45v
12v => 5.45v
etc...

Is there anything to do, perhaps recalibrating the unit?
I found a service manual from Tek that states there are some ajustable trims on the PCB, but didn't try before asking here.
Thanks in advance
Bruno

[dave_k]

I'm going to ask the obvious question (and forgive me if it feels insulting), have you tried a fresh battery?
Several times I've questioned my own measurements, only to find the meter's battery was running low...

[kerouanton]

Oops, didn't check this, thanks for the tip.
I don't have a fresh battery with me, I'll buy one, test and post a feedback.
(My Beckmann multimeter has a "low battery" indicator when turned on, I wonder if the DMM916 has one too...)

[kerouanton]

Brand new Duracell battery, but still have the issue ;-(

Here is an example on 240V AC mains power:
- 232v on the Beckmann (good)
- 148v on the Tektronix (wrong)

At first I believed I made a RMS measurement thus the lower result, but it turns out the lower value is also display for anx currents, either AC or DC...

So there is probably a calibration issue ?

[kerouanton]

Has anyone an idea of the issue?

[edavid]

It's broken.  Good luck fixing it without schematics.

I would look for a partially shorted MOV or other protection component.

[kerouanton]

Thanks for the reply. Not easy to repair obviously. Sadly, because this is a nice meter.

[retiredcaps]

I would look for a partially shorted MOV or other protection component.

Yes, check the input protection first.  A MOV should be open circuit as measured by another working meter.

You have to take apart your meter and post clear focused pictures of your pcb.

PS. It is unlikely to be a calibration issue.  Don't touch the pots at all.

[retiredcaps]

Your 916 will be very close to the 912 as per

http://mrmodemhead.com/blog/tektronix-dmm912-teardown/

[LateLesley]

Since it seems to be affecting many ranges, it may be faulty protection circuitry on the input of the A/D converter chip. Many meters are a singe chip design. IT may be worth finding where the chip is, finding the model no off it, and getting a datasheet for it. Then you can find which pin is the input to the A/d converter in the chip, and then just trace back from there. Also they usually have a voltage reference on a pin, and it may be worth checking it is at the correct voltage.

Might be time to take the back off, and get some pics of the board up, if you are keen to fix it.

[retiredcaps]

In the 912, there are two spark gaps (labeled SP1 and SP2).  Both should measure open circuit (ohms).

There are two PTCs.  Both should measure between 500 to 1500 ohms.

There are also two power resistors in that area.  I'm not sure what it should measure, but take measurements and report them to us.

[The Soulman]

If ohms are 100% ok and all voltage and current readings are measured low by the same percentage, than to me it looks
like the internal reference voltage is to high, its probably set by a trim-pot, couldn't hurt to give that a wiggle.

[RandallMcRee]

If ohms are 100% ok and all voltage and current readings are measured low by the same percentage, than to me it looks
like the internal reference voltage is to high, its probably set by a trim-pot, couldn't hurt to give that a wiggle.

Good grief, don't follow this advice!
You are off by a factor of two, no reference is ever off by that much. Adjustment range for a vref is like +-1% not 100%!

Ignore this and look at the input protection, etc. All the meters i have that stopped functioning had this kind of problem. (Also have DMM916).

Currently, there is probably one problem and when you find it and fix it, all will be well. If you start twiddling with calibration then you have TWO problems.  After fixing the original problem your meter will be uncalibrated!

[The Soulman]

If ohms are 100% ok and all voltage and current readings are measured low by the same percentage, than to me it looks
like the internal reference voltage is to high, its probably set by a trim-pot, couldn't hurt to give that a wiggle.

Good grief, don't follow this advice!
You are off by a factor of two, no reference is ever off by that much. Adjustment range for a vref is like +-1% not 100%!

Ignore this and look at the input protection, etc. All the meters i have that stopped functioning had this kind of problem. (Also have DMM916).

Currently, there is probably one problem and when you find it and fix it, all will be well. If you start twiddling with calibration then you have TWO problems.  After fixing the original problem your meter will be uncalibrated!

By wiggle I mean slightly forwards slightly backwards returning to the original position, I'm more thinking of a mechanical issue with the trimmer, i.e. possibly open circuit.
Without a schematic it is impossible for me to tell what that the symptoms in this case are from such a failure mode, but its easy enough to give it a try.

My humble 2 cents.

[kerouanton]

Thank you for all your replies.
I will tear it down tomorrow to make pictures of the PCB and check the input protections.
This post may also help me, I guess:
https://www.eevblog.com/forum/repair/tektronix-dm254-(service)-manual/

[retiredcaps]

By wiggle I mean slightly forwards slightly backwards returning to the original position,

The problem is what is the original position.  If it is NOT precisely/exactly back to its original position, the meter could be a few counts off.

[kerouanton]

Teardown done. There are 2 boards, one for LCD and the other one for the rest.
Nothing seems damaged, as well as both fuses which I just re-tested.

There are 2 PTC near the V input (last picture). I guess I'll have to start there.

Some pictures:

[kerouanton]

In the 912, there are two spark gaps (labeled SP1 and SP2).  Both should measure open circuit (ohms).

There are two PTCs.  Both should measure between 500 to 1500 ohms.

There are also two power resistors in that area.  I'm not sure what it should measure, but take measurements and report them to us.

SPG1: Open
SPG2: Open
PTC1: 1.6K ohms
PTC2: 1.7K
R36 (upper power resistor): 67K
R37 (lower power resistor): 16Mohms (that's high, no?? but as I don't know the original value maybe not)
I didn't check R31, R33, R38 and R39 as they are connected to the A input, so probably not related to the issue.

[kerouanton]

The V input goes directly to R36 and R37 (and nothing else). They then go to the Spark gaps
So the high resistance of R37 does make sense in my understanding.

[LateLesley]

https://4donline.ihs.com/images/VipMasterIC/IC/MCHP/MCHPS00572/MCHPS00572-1.pdf?hkey=EF798316E3902B6ED9A73243A3159BB0

there's the datasheet for the A/D converter chip. There's a sample circuit on page 6, so it might shed some light on some of the stuff in the Voltage circuit.

[lowimpedance]

The V input goes directly to R36 and R37 (and nothing else). They then go to the Spark gaps
So the high resistance of R37 does make sense in my understanding.

These resistors are fusible types and the high value indicates they have done their intended job and now need replacing.
As you noted above with the link to the DM254 thread that meter had the same failure and the resistor value was meant to be 470 ohms..
So remove the 16 M ohm R37 and peel of the heatshrink and read the color code for the correct value (both R36 and 37 should be the same I assume).
I think some suitable replacements had been mentioned. You will need to get fusible resistors with a similar physical size as I guess detailed specs are not known, ie no service manual with parts lists etc.
 I'll bet that will fix it. But as noted in that 254 thread if you temporarily solder in the appropriate value (as determined by the color code you find), of a standard resistor just to test the meter functions correctly then that will give you confidence nothing else is busted and then order the correct resistor type for the final repair.

[kerouanton]

Thank you everybody, thanks very much for the tips.

It was indeed the 2 resistors that went bad. Replacing them solved the issue, my multimeter now works as brand new!

So, R36 and R37 are 470 ohms resistors, you were right.

I did a quick test by soldering common 1/4W 470ohms resistors I had, to check the values of the DV, AC and Ohmmeter, and all is now displayed correctly. (it was for a quick test, I won't use my multimeter like this!)

Now, I need to find those same resistor fuses, which I don't have as spare. Are those 2W or more (4.5mm in diameter, and 15mm long)? Any difference with common high-power 470ohms resistors?

Bruno

[kerouanton]

Got my answer for the fuse resistor type. RTFM, I know 
https://www.eevblog.com/forum/beginners/what-is-a-fuse-resistor/