..back ontopic..
Inspected my "new" HP 4260A LCR meter.
First I discovered that the digital LCR-counter didn't go down to 0000 mechanically, lowest was ~0006.
So that was probably the reason why my small value caps (68nF/1nF) measured with a significant error.
To fix (aka calibrate) that the service manual said, I should loosen two screws. Incorrect. Version mismatch?
I loosened 17 screws in the end, took off all knobs and removed the front plate to be able to loose three more screws to get the counter gear separated from the potentiometer (R3) gear.
The manual said that at 0000-position the potentiometer R3 should have 5Ohms. In fact it had 22Ohms at 0000-position, but when counting up the LCR-counter it first went down over a minimum of 2.5Ohm to then go up to 5kOhm..
This potentiometer (5kOhm type) is so delicate you have to use an appropriate (=analog) meter otherwise even a 2 1/2 digit resolution of a DMM will drive you crazy.
To complete this calibration I had to work on another two grub screws in the hub of the gear-wheel sitting on the shaft of potentiometer R3.
Now at 0000-postion R3 has ~5Ohms on my analog meter.
After fixing that I poked around a bit, pulled all card boards out and put them back in...
Everything looked good -besides being relatively old.
Now, after putting the 4260A back together, the 1µF foil cap measures 0,97µF and the 68nF foil cap measures 68,5nF, both with D<<0,001.
Electrolytic caps with higher D-Factor of 0.2 - 0.4 all read to low, for example one of my new 820µF caps reads ~640µF with D ~0.35.
The 470µF electrolytic cap that measured ~70µF now measures ~386µF.
But not too bad until now.
Will inspect further.
Toddles off to bed..