Fixed strange fault in Philips / Fluke PM6306 with strange measurement results.

[Dekatron]

About a year ago I bought a Philips / Fluke PM6306 that I was told was working fully complete with almost all connectors, the only one missing is the LEMO - Banana plug wire, and with the PM9565 DC-option as well as the RS-232 interface.

It worked ok for a while but it was missing the PM9565 DC-option. I got it for a really good price and the seller promised to take it back if I couldn't get it to work and even offered a refund for the missing PM9565 DC-option.

What suddenly started to happen was that it started to deviate from the first measurements I had made of wire wound resistors and inductors, capacitors were working fine all of the time. Since I had written down the first measurements I had done of some high precision wire wound resistors and I have a calibrated HP34401a that I used I started to try to find the reason for this fault. I went through the complete process of selecting fixture set, performing contact check and performing the zero trim which all passed but the strange measurements were still there with varying measurement results.

I cleaned contacts and also removed some residue found on the circuit boards, probably from cleaning fluid during manufacture of the circuit boards, loosened and re-seated internal connectors but nothing helped. I then decided to check the battery as I had run out of ideas - lo and behold, the battery was at 0V! When I first opened the PM6306 I had checked the battery so I knew it was working fine at some 3V when I first received the PM6306, but after having been stored for some 5 years by the previous owner the battery must have given up as soon as I started using the PM6306. I decided to replace it with a socket and a standard lithium cell found in most PC computers as I had a few of those lying around, after having done that it worked just as it did from the beginning - so the solution to strange measurements was to replace the battery.

I guess the strange results have something to do with the results of the zero trim not being stored in the RAM-memory properly when the battery was low however I am not sure how a low battery could affect the instrument with power applied, it should only have an effect when powered off, but maybe there is some poor programming that means that uninitialized memory cells with random data affect the measurements.

Anyway my PM6306 works perfectly now that the battery has been replaced with a new one so I hope this might help others struggling with strange measurement results!

For my new PM9565 DC-option module that I made a new circuit board for you can check this thread: "https://www.eevblog.com/forum/testgear/making-a-pm9565-dc-option-module-for-my-philips-fluke-pm6306!/msg4043251/#msg4043251"

[HighVoltage]

Nice find.

Please show us some pictures of your battery modification.
 

[Dekatron]

Yes, a lot easier than having to do proper fault finding of the whole circuit!

Here are some photos of the battery with socket from a PC and the old battery on the side, the socket fit straight into the holes of the soldered battery so nothing has to be modified to get it to work.