Still issues with my HP 8903a, "mV" range change problem

[Agosto]

In my previous post ("HP 8903a, "mV" problem" dated May 5) I reported to have found a bad relay K1 on the A2 (Input) board. I replaced that relay (Magnecraft W107 DIP-1) with a Littlefuse HE721A0500. With the original relay, I was able to measure DC voltages down to 0.19V. Below that, I got erratic results.
With the new relay in, I can measure DC voltages down to 0.48V, then again below 0.189V down to 0.076V. Between 0.48V and 0.189V the left display turns blank and the right display shows "----". At the same time, I am hearing  rhythmic relays clicking, I believe from the A2 board. With the original relay this voltage range was fine but the range below that (0.189V) was bad. So it is reversed.I double checked that the new relay has the same functionality as the original.

Following table 3-8 (Page 3-35) of the service manual these voltage thresholds correspond to Special Function codes 1.15, 1.16 and 1.19.

Since then, I replaced relays K2 and K3 and confirmed that diodes CR16, 18 and 21 are good. Trigger voltages for K1 and K2 come from U15 Pins 2 and 12. Both outputs shift between H and L about once per second when voltages reach the "critical" ranges (0.189V - 0.48V) causing the rhythmic clicking, I believe.
U15 is a 7407. Is it possible that this IC is defect? I was not able to trace back the respective input of 7407 to the latch board (I believe this is where the signal is coming from).
Any suggestion where to look as a next step?

Many thanks

[Kim Christensen]

The 7407 is just x6 TTL buffers with open-collector outputs. Should be pretty easy to check that it's operating correctly.

The continuous relay clicking sounds like the 8903 is trying to autorange. But because it has a fault, it continues to search without success.

[Agosto]

Found the issue: The original relay has an internal connection between pins 2 and 13, the replacement has not. My oversight. Unit works fine now in Autorange mode across all voltage levels.
However, I believe I am still not done yet: I started to look at performance criteria and found that "Distortion" at 6V AC is around 0.47% and 0.53% at 0.6V. Spec is <0.01%. Measured with the built-in oscillator. If I use an external signal generator, I get 0.017% resp. 0.05%.
I guess I need to look at the oscillator next.