With things like those it doesn't even matter what it is, what it is supposed to do and whether it still does it. Something made like that is just a pleasure.
Very true indeed.
Well I found the problem. As expected. Reverse power overload with attenuators switched in. Quick trick with a 50 ohm attenuator strip. If you switch out all the attenuators so it's 0dBm attenuation, then measure the resistance at the junction point at each node on the pi attenuator network it should be around 50 ohms. If it isn't, smoke has come out somewhere. Smoke most likely came from somewhere closest to the output.
![](https://i.imgur.com/QU2dJbg.jpg)
Soooo, this is a 96.3 ohm parallel pair and 71.2 ohm series. Time to go shopping for some suitable thin film resistors that have to be good up to 520MHz...
Quick shot of the whole assembly opened. Note the yellow things - they are transmission line stubs / gimmicks by the looks.
![](https://i.imgur.com/OQyY5Ip.jpg?1)
Edit: can I get a 96.3 ohm resistor? Can I hell!
![Sad :(](https://www.eevblog.com/forum/Smileys/default/sad.gif)
Edit 2: went through a massive box of resistors I have and found a used 97.6 ohm 1/4W MF one! Close enough. Have checked all the others and bunged that in and reassembled it and it works down to -60dBm. I can't measure any lower than that at the moment and I'm not sure if I care about miniscule powers like that for now. My power meter is reporting flat response to 405MHz as well which I was happy with. There's a 0.4dB error on the 10dB attenuator (4%) but I can live with that. And on that note, as it's 01:25 in the morning here, bed!
Edit 3: before I disappear, the enclosures for the attenuators are aluminium and one leg of the resistor is soldered to it. Rather than try and get a good joint on fresh alu which is a dick, I soldered it to the existing solder, adding some 62/36/2 silver bearing. This did the trick.