Keithley 179 multimeter zero problems

[The Soulman]

I've picked up a very used and sick keithley 179 multimeter for almost nothing,
It had different issues but re soldering the switches and al larger components, reseating the ic's and replacing al power supply related caps fixed all but one
it still reads 0002 with the inputs shorted on the 200mV dc range (therefore also on current range), on 2000mV dc range reads 0000.
Only real difference between the two ranges is the used internal reference (100mV vs 1V).

Question, is this a known "feature" of the 179 or something not so common and therefore probably fixable?

Will post more numbers and info soon.

[Jwalling]

That's a pretty old DMM. Although I have not worked on one that I can remember, I bet it has pots for calibration that may need adjustment.

Jay

[The Soulman]

That's a pretty old DMM. Although I have not worked on one that I can remember, I bet it has pots for calibration that may need adjustment.

Jay

Yes, pretty old, 1980 or so, and yes many pots for many things, even a thru the front panel accessible zero adjustment pot,
sadly only for the resistance range (?), at least it works fine on that range, not on dc mV.

Problem of not reading zero even remains when shorting C123 (it is on the icl7103 input)
so it must be something with icl7103 or the icl8052 or more likely somewhere in between.
Already played with r144 and r145 they pull the ground of the icl7103 slightly up or down but no luck..
Also replacing the integrator and reference cap with new wima mkp10 ones did not help.

Here are links to the schematics and the datasheet from the chipset it uses:

http://www.download-service-manuals.com/download.php?file=Keithley-5559.pdf&SID=2upvc9d64cvsvq4rka847tdb42

Wow, actually while writing I've found a app note for the chipset, going to read that!:

http://www.intersil.com/content/dam/Intersil/documents/an02/an028.pdf

The reason for repairing this old thing is because I can (it is getting personal..) and could actually (in a modified form) put it to good use.

[macboy]

I've picked up a very used and sick keithley 179 multimeter for almost nothing,
It had different issues but re soldering the switches and al larger components, reseating the ic's and replacing al power supply related caps fixed all but one
it still reads 0002 with the inputs shorted on the 200mV dc range (therefore also on current range), on 2000mV dc range reads 0000.
Only real difference between the two ranges is the used internal reference (100mV vs 1V).

Question, is this a known "feature" of the 179 or something not so common and therefore probably fixable?

Will post more numbers and info soon.

Unfortunately, your meter has no zero or Rel function. Have you adjusted R149 for zero with 200kOhm? Does yours have R112 installed? There may be interaction between R112 and R149, and at least R149 will affect the zero reading for voltage in addition to Ohms (I didn't find R112 on the schematic).

[The Soulman]

I've picked up a very used and sick keithley 179 multimeter for almost nothing,
It had different issues but re soldering the switches and al larger components, reseating the ic's and replacing al power supply related caps fixed all but one
it still reads 0002 with the inputs shorted on the 200mV dc range (therefore also on current range), on 2000mV dc range reads 0000.
Only real difference between the two ranges is the used internal reference (100mV vs 1V).

Question, is this a known "feature" of the 179 or something not so common and therefore probably fixable?

Will post more numbers and info soon.

Unfortunately, your meter has no zero or Rel function. Have you adjusted R149 for zero with 200kOhm? Does yours have R112 installed? There may be interaction between R112 and R149, and at least R149 will affect the zero reading for voltage in addition to Ohms (I didn't find R112 on the schematic).

Thank you for chiming in, r112 and r149 are used to compensate the dc offset of the "attenuator" op-amp, r149 as a fine adjustment "user accessible" through the front panel and r112 as a coarse adjustment on the inside (it is just right to r149 on the schematic).
They both work fine, but the "attenuator" is bypassed (at least it appears to be) in 200mV and 2V dc range mode so not helping there.
Also suspected the protection fets q106 and q107 of leaking but shorting c123 on the convertor input did not change anything compared to shorting just the inputs.
So the problem must be somewhere down stream..
Probably am going to replace r144 and r145 with something to compensate for the offset, but that's a bit of a bodge solution, rather track down
the real cause of the offset. 
Another increasingly more likely and less obvious culprit could be the + and - 15 volt regulators (7815/7915) that have drifted separate ways (over a volt out of symmetry) I ran out of fresh ones at the moment so may replace with 317/337 with resistors and trimmers but also ran out of small trimmers..
To be continued.

[The Soulman]

Replaced the 7915 with a fresh one, voltage changed from -15.5 volts to -15.05 volts a bit closer to spec..
The offset remains at .02mV.

quoted from the appnote:

a few modifications are
necessary to accommodate a 100mV reference
 (..........)
Third, a 300k potentiometer should replace
the 300k? fixed resistor in the comparator translation
network. With VIN = 0V, this pot should be adjusted until the
display reads equal intervals of positive and negative signs.

This resistor is fixed in the 179 and part of a resistor array (R135H) yes hurrah it thighs the zero point to the -15v rail    so might botch in a trimpot or use the original keithley solution pulling the analog ground pin of the icl7103
a bit above or below actual ground (towards one of the power rails) using r144,r145 and r146.
Either way long term 'zero' stability is dependent on the power rail stability, so might want to get that nice and stable first.

Next step is to remove the pushbutton range switches (really dickey) and the current shunts (drifted anyways..).
And substitute a nice rotary switch for input range switching (also loosing ohms and high voltage).
To leave a ac/dc voltmeter with ranges from 200mV to 200V and adding in a 24bit/96khz adc to
do some audio bandwidth fft stuff on a computer. 

Fun and learning new stuff all the way.