HP 5350B 10MHz output strange waveform shape

[matthuszagh]

I connected the 10MHz output of my HP 5350B microwave frequency counter to the input of my oscilloscope via a BNC cable and I was surprised by the waveform shape. I've attached an image showing this (the yellow trace is the HP 5350B 10MHz output and the green trace is a 10MHz standard from a GPSDO). Is this normal? I would have expected a sine wave. The 5350B is equipped with a high-stability oven timebase (option 010). I did some quick tests of the 5350 and it seems to work as well as I'd expect for a used Ebay item. I hooked it up to my HP 8341B (this is not calibrated, so the frequency is probably a bit off) and it read out close to the correct frequencies to the signals I gave it (it was always about 500Hz high, IIRC). The inside of the device appears to be in good shape. I also checked the 1MHz output which appears as a square wave with a fair amount of ringing on the edges. I can post an image of this if it's useful. Because I used a BNC cable directly I don't think this should be the result of the way I'm probing the signal (e.g., because of a long ground lead causing inductance).

[jfphp]

Did you put a 50 Ohms feedthrough before the scope input ?

[matthuszagh]

Did you put a 50 Ohms feedthrough before the scope input ?

Well, I feel stupid. Changing my scope input impedance to 50ohm yields a waveform that looks much better.

[jfphp]

What is the high stab reference ? HP 10811X ? Does your mesurement take place directly at the oscillateur output or is there a circuitry between. In the last case make direct connection to the oscillator output.
Apparently there is something wrong : shape and offset.

[matthuszagh]

What is the high stab reference ? HP 10811X ? Does your mesurement take place directly at the oscillateur output or is there a circuitry between. In the last case make direct connection to the oscillator output.
Apparently there is something wrong : shape and offset.

Yes, it's a 10811. I've connected the BNC cable assembly to the back panel 10MHz BNC connector. Is this not the correct way to do it?

[jfphp]

10811A floating through rubber O rings and connected by soldered wires and coax or plugged into a connector soldered on the mother board ?

[jfphp]

We are vasting our time : take the op and srvce manual (05350-90021) : you will see p 8.43 block diagram that the 10 Mhz out comes not direcly from the 10811A but from the time base buffer part of A1 throogh line drivers and compare your waveform(s) with those displayed on the various sheets . 10 Mhz comes from U9C pin 10 and cannot be a sinusoidal waveform !