Fluke 185 repair attempt: "Uncal" startup message

[giosif]

Hello,

I've recently scored a Fluke 185 which displays "Uncal" every time I turn it on.
Testing accuracy, some measurements seem to be close to expected value, but not all.
I pinged Fluke support and they told me the message indicates "an internal memory error".
This seems to be confirmed by the rather time warped last calibration date of 77/9977 the meter is showing (MM/YYYY format).

Looking at the meter's board layout and using the nice details in the review post here
https://www.eevblog.com/forum/testgear/fluke-185-review/msg578824/#msg578824
and particularly this picture

I went ahead and desoldered the EEPROM IC (an ST M95160W).

Reading its contents with my trusty TL866 programmer several times, I got the same binary data on every occasion.
Assuming the calibration data, including the last cal date, are stored on the EEPROM IC, this means that either the actual contents of the IC have somehow been corrupted, or the contents are fine, but the main IC cannot read it from the EEPROM correctly (or at all).

So, my next step in the troubleshooting process was to re-solder the EEPROM IC back in its place and probe some of its pins during the startup of the meter:

• pin 1 - Chip select (active low)
• pin 2 - Serial data output (active high)
• pin 5 - Serial data input (active high)
• pin 6 - Serial clock (active high)

Please note the "active low" detail for pin 1 above.

If I now look at the single capture from my scope, I see the data in the screenshots attached ("one.png" is the overall capture, "two.png" is a zoom into one of the areas of interest, "three.png" is further zooming into an area of interest).
And, if my interpretation is correct, this looks like, at some point in the startup sequence, chip select is disabled (i.e. pin 1 is high) when data is being sent in and out of the EEPROM IC.
Is this normal behaviour?

Thanks!

[wictor]

Hi,
some Fluke DMM show uncal, when they have lost the calibration, because someone pressed the cal button or tried to do the calibration without finishing it or did not have proper tools.

Do you have access to proper calibration tools to try to calibrate it and see if it passes or to verify that eeprom has really died?

Wictor

[giosif]

Hi,
some Fluke DMM show uncal, when they have lost the calibration, because someone pressed the cal button or tried to do the calibration without finishing it or did not have proper tools.

Do you have access to proper calibration tools to try to calibrate it and see if it passes or to verify that eeprom has really died?

Wictor

This meter model does not have a cal button and calibration can only be done via the IR port.
So, I doubt someone managed to initiate, then screw up the calibration.
Also, as I said initially, the EEPROM IC reads fine; so, it's not dead, but either its contents have been corrupted (less likely) or the main IC can't read its contents properly (more likely, in my view).

[ErnestB]

Hi, I know this is an old post, but wanted to know if any progress has been made? I have the same type DMM with the same massage on the startup. One question, if the meter could not read the calibration values from the eeprom, then it would not work "almost fine", right? I suppose all values would be "all over the place"? I still have to do a check up of my meter to try to see on what functions and what ranges it is out of specs... I have done some DCV measurements comparing to a good HP3478A (1mV to 10V, 18 different values) and all values were within the specs. I still need to do the AC, but I do not have a very good reference source yet. My suspicion is that the AC part could be outside the specs, and that some part (like a capacitor) could have caused the problem... and that the DMM could not been calibrated at the certified Lab, and that therefore the Cal procedure could not have been finished. This meter was on a yearly calibration program, and came one time back with the massage that the DMM would need to be repaired....

[giosif]

Hi,

Yes, I put the EEPROM contents from another Fluke 183/185 (which was not complaining about "Uncal") into mine and the start-up message went away.
Please see here:
https://www.eevblog.com/forum/testgear/fluke-185-review/msg3672067/#msg3672067

Now, if you are certain your multimeter is fully functional, even if out of spec, then you can apply the same procedure.
But I suggest you confirm the correct functionality first.
And, if I recall correctly, using a different EEPROM content means the original calibration data is lost.

[ErnestB]

Hi, thanks for answering. I just did some quick tests on ACV. What I see is that the meter is showing less volts on the 5V range and more volts on the 50V range. Also I see a strong frequency dependency. While the good meter HP3478A is showing a "stable" 5.07 to 5.08 volts (range 200Hz to 20kHz sinus), the 185 is showing 5.2 down to 4.8V on the 50V range, and 4.8 to 4.4V on the 5V range, over the same freq. range. So for both ranges the value is going down strongly (same amount of 0.4V) with the freq. So a clear case of a low-pass filter 😁
De DCV is within the specs. So by loading an another cal values that would be lost of course. I will try first to find those coupling capacitors for the AC measurements. Any idea where they are to find on the PCB? Or any idea or thoughts on this specific error? Or schematics maybe?
Edit: I see that it also gives much more DCV reading from the Signal gen. on sinus then the HP does. 250mV instead of a few mV.

[ErnestB]

I suppose the series resistance of the AC coupling capacitor is higher then normal. That explains both different ratios for the 2 ranges 5 and 50V, and also the low-pass filtering, I think  I would need to localize that C.

[ErnestB]

Opened the DMM and I suppose the C72 would be the AC-coupling capacitor. When on DC I measure about 300 Ohm, and when on AC the resistance is high, what is all logical. When testing measured about 26n (in circuit) and it behaves like regular ceramic capacitor... Could de-solder it and try an another but... I am not convinced yet