Fluke 8060A with incorrect resistance reading

[nikifena]

Hi there. I'm having a problem with my fluke 8060A DMM.
The resistance readings are incorrect. Actually the reading value is higher. For example with 1.000k the DMM reads 1.01 and with 150k00 - about 250k. Tested with my fluke 187 these values are Ok. With a higher R value the readings becomes more incorrect.

I tested all DMM voltages. All of them are correct. R1, R2, RT1, Z1 has also correct values. Q3 and Q4 seems to be not leaky. In a Mohm range the meter shows OL and in conductance test the display shows 0.00
I checked Z5 resistor network and it seems to be  Ok.
Only in diode test with 1k resistance, the voltage drop across this resistor is 0.92-0.93V (it must be about 1V). I was replacing the existing CR1 with an external current source and with fixed 1mA current the value is accurate but the resistance measurement not.

All of the voltages from table 5.8 are correct too.
I'm using this user manual as a reference:
http://assets.fluke.com/manuals/8060a_3vimeng0200.pdf
My multimeter hasn't extra RMS PCB. The RMS chip is placed on mainboard.

So I'm asking you for some ideas how to fix this issue.

Thanks!

Niki

[retiredcaps]

It appears you have done a thorough job of reading through the troubleshooting section and checking everything so far. 

As you may or may not know, the 8060A seems prone to bad electrolytic capacitors.  The capacitors will generally leak from the bottom and it won't be obvious looking at them.  When they leak, they contaminate the pcb causing all sorts of funny readings.  I, modemhead and excavatoree have all experienced bad capacitors on the 8060A.

If you haven't done so, you can also give the pcb a complete careful IPA cleaning.  In the manual, it also says contamination could cause the problems you are seeing.

[retiredcaps]

I forgot to mention in my earlier reply that modemhead's 8060A repair is documented here

http://mrmodemhead.com/blog/fluke-8060a-repair/

From this blog "Resistance measurements are 10% or more off."  You can read the article for hints and suggestions on what to do.

[nikifena]

Thanks. I was reading the 8060A manual and the cropped part from you. everything seems OK.
 
I know  Mr. ModemHead's blog. It's pretty cool!
 
I made a visual inspection of all caps. They looks ok but maybe, I assume, some capacitors are bad.
There is no capacitor leakage.
I will desolder some caps to check their condition.

EDIT: I will check C13 and C14. Maybe some of them has an inaccurate value.

[Marvin]

The manuals (8060a___imeng0300.pdf and 8060a_3vimeng0200.pdf) that are available on Fluke site are from the time when the true RMS converter chip in the 8060A was changed from the Fluke engineered SC77174 to Analog Devices AD636 mounted on an extra board with some opamps. By a fluke I found this PDF from a French university site that has the original manual from 1982 with full schematics scanned @ http://www.ens-lyon.fr/DSM/AGREG-Physique/oral/Notices/N069-017.pdf

[nikifena]

^ Thanks guys!

Yes, I know Mr. Taylor. I was reading the entire thread before a couple of weeks. It's amazing to read an information about this pretty good DMM.

[nikifena]

Hi there

I change all electrolytic capacitors with a new. I tested the original ones and they were in a good condition and weren't  leaky. Anyway they were pretty old so with a brand new ones are better.
The voltages readings are OK with all test points and all reference.

What I found: When a resistor is connected (I'm having 150k reference resistor) the value starts decreasing slowly from about 220k to exactly 150k. The time for a correct reading value is approximately 5 minutes.

When I connect the resistor and measure the voltages at OREF+ and OREF-, the OREF+ value is stable, but the OREF- value changes due to measurement. Definitely  the problem comes from OREF-.
Q3 and Q4 are good. Any reference voltages are ok.

Any ideas how to fix this issue?

Thanks!

[retiredcaps]

I have a busy few weeks ahead with taxes due, migrating at least 2 computers from XP to lubuntu (or some other linux distro) + applications/data so my time and research/test capabilities are limited, but did you try removing, cleaning and reseating the main IC?  How about cleaning the 200k ohm switch?

Just thinking out loud, did you also try the 20M ohm range to see if it is accurate?  I usually just measure the mV  input impedance of another meter and usually get 10M.

PS. Modemhead is still on the road.

[nikifena]

The MAC chip and his socket looks OK. Yes, I cleaned the entire PCB. There were no any leakage from capacitors.
The problem occurs visually with high impedance resistor (in my case 150k as a reference)
With 1k5 reference the resistance looks good. I assume that there is the same problem but maybe the time for getting the proper measured value is smaller.

When 150k resistor is measured in lower -200k range the meter shows OL, and I change next range - megaohm, until the value drops under 200k. Than I can measure the falling value below 200k with 200k switch

[retiredcaps]

Now that I review this thread, I see that RJ1, RJ2, RJ3 and RJ4 were maybe not tested?

Since I posted almost 10 months ago, I have learned more about how the input protection works and when the MOVs or varistors are bad, I know it can affect readings.

So with your newly fixed Fluke 79, test and verify that RJ1, RJ2, RJ3 and RJ4 should all read open circuit.

[krivx]

I had to replace the charge pump IC, the socket for the MAC and the the caps in my 8060a after the originals leaked. I had to be really thorough in cleaning the board afterwards to get the thing back to reading correctly. Multiple washes with isopropanol and distilled water were not enough, eventually I had to clean the MAC itself in an alcohol bath to get things completely clean. No residual readings after that.