Systron-Donner Spectrum Analyzer
Type 809-1 (10MHz - 12.4GHz), page 5

The photograph on the left shows the repaired 260MHz amplifier reinstalled. The long white SMB lead is the original and is presumably that length so as to allow the two halves of the RF-unit to be operated when ‘split’.

So I now had my Spectrum Analyzer back. But what about the faulty input attenuator? I was able to split it and verify that one of the elements is damaged. I had an idea that I might have a solution. I found a couple of rotary step attenuators salvaged from scrapped test equipment. One turned out to be 1dB steps, so that was a non-starter. The other was 60dB in 6 steps. It was a real 3D ‘jigsaw puzzle’ fitting it in. The only down-side of the Weinschel against the Widwest attenuator is that it is ‘backwards’ … i.e. 0dB is the final position as opposed to the first.

However, this solution proved to be problematic. Such was the tight squeeze that the rear connector on the attenuator actually pushed the adjacent PCB against the bulkhead. causing intermittent short-circuits. Also, I discovered that the 0dB position was intermittent. This might actually have been due to stress applied by the tight squeeze. Thus I have reverted to the original, but faulty attenuator.

I now plan to build the Weinschel attenuator into an external box that can be connected up to the front of the analyzer. There is no law that says that the attenuator must be internal.

Above - underside of IF-section. The faulty attenuator is the grey cube on the left.

Right - Weinschel attenuator ‘shoe-horned’ into place with bulkhead mounting N-SMA                        adaptor as input connector. This solution was later abandoned.