It's a nicely made signal generator with incredible internal screening, a fanless design, which I always like. It's fitted with a rather excellent attenuator, too. Sadly this unit isn't fitted with the GPIB option, there's a blanked off rear socket and two empty card slots inside for this.
Chucking the signal into my 3585A reveals that the 2f harmonic is usually around 40dB down on the carrier and there's a persistent spur exactly -720kHz from the carrier at about -65dBc, not an exciting start, but when you look closer the peak is very narrow and the noise floor is very low. I think it must've been designed for low phase noise. Residual FM is ten times better than either of my Marconi 2019/2019As. I wasn't expecting that, because on the face of it, it's a fairly basic signal generator. All the numeric front panel programming happens using thumb-lever digital switches, they're like thumbwheels but you waggle a lever and the digit flings itself from 0-9 effortlessly. They're nice to use. You can also set a step size to increment or decrement the frequency, which is displayed on a green LED display.
Despite the sticker claiming that it's slightly dodgy everything worked properly except the internal reference oscillator's oven which was only drawing 9mA and not getting warm. It only warmed up using the heat from the nearby power supply and even then didn't get to the correct frequency.
The OCXO was a 10MHz Cathodeon part, it gets divided by 10 immediately and the external connections for reference in/out are all at 1MHz.
This was being supplied with +5V and +12V, with the +12V optionally present when the generator's front panel switch is set to "off".
As I had a stock of several Decca 10MHz OCXOs with a nearly identical footprint that I'd bought as part of a job lot of component cabinets at Milton Keynes a few years ago I thought I'd pop one of these in.
The Decca only needs +12V but has an additional pin that is used for oven current sensing, so you can warn the user that the oven isn't up to temperature with some sort of indicator light, klaxon or malodorous gas release mechanism. The Decca doesn't require +5V but has enough output to toggle the logic, the pin that had +5V on it is a +7.3V ovenised Zener stabilised output. For now I am not using this for anything and conveniently the current sense pin on the Decca is shrouded in ceramic, so I drilled the PCB to match this which had the effect of severing the +5V line at the same time.
The tuning voltage for the OCXO comes from the power supply board, possibly from a simple shunt Zener regulator. There's a multiturn adjustment potentiometer wired as a voltage divider. Drift is not amazingly great on that "stabilised" voltage and spends rather longer than I'd like wandering before settling down. I am considering moving the hot end of the tuning voltage pot to the +7.3V from the Decca. Having said that, it's not *bad* and half an hour after power-up there are 10 non-changing numbers on the frequency counter.
Here's a view of the rather tidy interior. The attenuator is canted to make it fit. There's a slab of aluminium between the PSU and the rest of it, this is also laminated with some of that mu-metal stuff, or similar. I'm surprised they didn't use a toroidal mains transformer. The PCBs have a solder resist and silkscreen on the component side (except the pictured oscillator, weirdly) but oddly they don't even have any of that on the solder side, which is a bit of a let down. Also, no gold, but everything worked bar the oscillator oven and that's not bad going for something made in 1981. It doesn't appear to have been worked on previously.
