I finished up repairs on the signal generator this weekend and did a brief performance check. I was unable to easily find an axial equivalent for the shorted filter cap so I came up with this arrangement for mounting a nice radial style with screw terminals. I also ordered the R712 diode pack in a TO-3 package, but until they arrive I’m using an NTE diode pack in a TO-220 package.
After confirming that everything was running nice and cool and at the correct voltages I let the generator burn in for a day before looking at the ovenized 10 MHz crystal time base.
After 24 hours I took my 10 MHz rubidium frequency source and compared then trimmed the 8656A oscillator to match. The rubidium source is mounted in an old hard drive enclosure. I added a meter and leds to indicate the status and when the rubidium cell is locked. To keep the operating conditions as true as possible I set the fan back in place after each adjustment. I waited an hour in between each adjustment.
Letting it sit for a few more hours I checked in with my HP 8566. This generator looks to be very accurate in both frequency and amplitude accuracy.
