Sunday, December 05, 2021

10Ghz downconverter for 1.5Ghz spectrum Analyser

Had this build for some time, now it's time to show. 


 I was doing some experiments on the 10Ghz band and wanted a way of looking at the signals. Because the spectrum analyser I have only good to 1.5Ghz had to find a cheap way of doing it to get this:

Here looking at the third harmonic from an ADF4351 on 3.3Ghz after a pipe cap 10Ghz filter experiment.

The diagram explanation: a dbm mixer (Watkins-Johnson M80LCA) with a local oscillator based on a FVC99 10Ghz oscillator module (cheapest VCO I could find for 10Ghz). Some preamps on the input and output using 2Ghz preamp modules and replacing the MMIC amplifiers for the ones like Corvo NLB300 or ERA-1 that are good to 10Ghz.

The basic design:

 To this diagram I added a 6db directional coupler inline with the FVC99 VCO (used a Omni Spectra PN2023, good from 8 12.4Ghz) so I could measure the LO frequency and PLL it.

There is no stability control on the FVC99 oscillator, still working on a PLL system (maybe one of these days) but in my case I have two select positions, one: VCO is controlled by a single pot (like on the diagram) and the other position controlled by an EIP371 frequency counter (from the Lo Out via directional coupler) that makes the PLL loop. With EIP371 and since the output voltage of the loop is very small the control range seats near 9.5Ghz, there is an option of extending the range like on the EIP manual:

Or with a similar diagram, a multiply by 10 of the PLL voltage out of the EIP371, that would be enough to use the full range of the FVC99.

For now I use 9.5 Ghz if using the EIP371 for more stability and around 10Ghz set by the pot ("Flo" on the panel) if it's just a quick test.

Here the EIP working as external PLL controling the FVC99 so the LO gets more stable.

On the Rigol DSA815 spectrum analyser you can set the input offset to get the display right on the band of interest

Displaying here a 10Ghz signal using the 9.5Ghz Lo frequency

If you want to just check if the signal is around there, no need to use external PLL control to the FVC99, the "stability" with a simple potentiomenter  is enough.

Inside view:

The VCO adjust pot (top right in blue) is glued directly to the front panel

Some other images during prototype development:

Here one of the firsts tests, just an input amplifier, the mixer, the VCO and VCO amplifier and the mixer IF output directly to the spectrum analyser.

Testing during early days of the prototype with a 10Ghz homemade flange to SMA adapter and a pipe cap filter:



Anyhow, not a measuring device but it serves the purpose of checking if you have any signal around the 10Ghz band and for experiments, still very happy with the outcome and sensitivity.

Have a nice day!

No comments: