I have also reorganized my bench as I wrote some weeks ago (photos still to be made, not ready yet).
One positive effect is being able to pull out single instruments out the rack.
I used that for swapping the quartz in my HP 3456As that determine the integration time of the A/D conversion from the 60Hz-type to the 50Hz-type.
I ordered some custom made quartz built to HP specs some years ago and had them waiting for use since then -because swapping them in would have meant to tear down all the stack my HP 3456A were sitting in.
Since the change in noise level with the right quartz (50Hz type) was noticeable I decided to do some statistics using the built in the 3456A statistics functions.
(I also tested my shorting plugs with this.)
I checked my two working units #1 and #2, each with 50Hz-synchronous integration time and once with 60Hz-synchronous integration time.
Auto-zero was ON and filter was OFF.
Here's the raw data:
| #1 / 50Hz | #1 / 60Hz | #2 / 50Hz | #2 / 60Hz |
upper | -00.0000E⁻3V | +5.40000E-6V | +1100.000E-9V | +6.00000E-6V |
lower | -4.00000E-6V | -7.40000E-6V | -2.40000E-6V | -6.00000E-6V |
mean | -1586.778E-9V | -1634.096E-9V | -521.938E-9V | -602.368E-9V |
variance | 0.000049E-9V | 0.004099E-9V | 0.000036E-9V | 0.002787E-9V |
count | 141.8320E3 | 147.3180E3 | 124.5800E3 | 157.3460E3 |
It's clearly visible that the "right" quartz for the A/D integration time being synchronous to the mains power (50Hz in my place: Germany) has a significant effect; it's roughly a factor 100 looking at the variance.
I'm quite pleased about these results.
I started the same test with my 3457A and it seems to be worse by roughly a decade (~147E-9V);
It's significantly noisier -but that's no surprise when taking it's front end construction into account..
Conclusion: Do work with the HP 3456A only with the right integration time, use the right quartz for synchronizing the A/D-conversion integration time to the frequency of your mains power systems in your place. Otherwise you loose roughly two digits resolution to (unnecessary) noise.
Edit: Added conclusion