Everything takes time, and this IF channel rebuild is no exception.
I thought building things in a separate chassis would make the job easier, but of course, it’s a tighter space to work in. And I spent, all in, about eight hours on this detour.
But victory, and a small defeat, are mine.
The old IF channel is pictured below.
Specifically, from the white ceramic tube socket and up. The two IF tuned circuits can be identified by the white (hot) and green (cold) wires coming out of the scrap IF transformer mounting cans. I didn’t think it was too askelter, although densely packed. That coax is from the new channel and wasn’t running through the middle of it.
But I have to admit, the new channel is better shielded and organized.
There is a brass shield separating the filter input and output sides, and another internal shield between the first and second stages.
And the outside looks pretty good.
The first IF stage was reproduced without difficulty. I wired the accessory socket on the chassis to provide B+, heater, ground and AGC voltages, and wired the sub-chassis with an eight pin plug to exploit those connections. This worked out well. When the sub-chassis is complete, I can rewire those connections inside the radio and free up the accessory socket for the next debacle. I also ditched the L matching network after further reading, and just fed the filter with a 360 pf capacitor directly, and directly out into the first IF stage, and this bought me some improved gain as well. The output seems a tad ripply, and I plan to add a 200 ohm terminating resistor to the output to see if that helps.
Stage two proved more difficult. I’d worked out the wiring I needed to add a second IF amp stage, which was basically just reproducing the first, but I used the second tuned circuit from the first stage as the tuned circuit for the second. Unfortunately, I got nadda out of the IF channel. Turns out in resoldering the tiny Miller 4407 coil, the winding opened up. This meant disassembling the entire second stage to try to repair the coil – successfully I’m happy to say, and when finished the stage did work.
I can still make the IF oscillate, but it’s much more stable than it was, especially when the bottom is on the sub-chassis. Voltage measurements suggest the improved standalone IF channel is capable of over 25 dB gain after the filter loss, so that’s about 20 dB better than it was, and it’s stable with a wee bit of detuning. This translates into being able to hear a -100 dBm signal injected into the IF channel, again a marked improvement.
On the downside, the extra tube may have pushed the filter choke beyond it’s limits. It is a 7 henry 45ish milliamp choke, and started dripping a bit of wax, getting hot, and I blew the half amp fuse after testing for about an hour. The choke will need to be upgraded soon. It looks like an 18 buck spend for a Hammond 10 henry 65 milliamp choke. I’m up to about 60 bucks at this point, with most everything coming from the junkbox.
Also on the downside, the mixture coil still does not peak, and I can barely hear a -100 dBm signal injected at 4 mHz. Remembering the loss from the IF trap, that’s still only a minimum discernible signal of about -105 dBm, which ain’t great, so the front end design will need more attention.
But first I’ll bolt the new IF channel in, and then maybe put in some plug-in bases to experiment with the front end coils. I’m thinking of going with HBR-style plug-in coils – this isn’t going to be a bandhopping receiver – and if I want to later, I can always work on converters or a band switching arrangement.
So progress, but it’s still a journey.