Back in May, I dug up a Hallicrafters 8R40 from one of Seattle’s local electronics recyclers, 3R Technologies. It looked like it had been stored very well, had all the tubes, the dial strings intact, no rust to speak of and not even very much dust.

This one is a bit different from the others I’ve worked on, it’s a general coverage communications receiver as opposed to being a consumer/entertainment receiver. It’s built more sturdily, designed to be serviced more easily (down to the flip-up metal top for easy access to the tubes) and was designed with an eye for sensitivity and performance to receive ham radio signals. It receives the standard AM Broadcast Band + three shortwave bands covering most of the amateur frequencies from the 160M band (1.8MHz) down to slightly longer than 8M (44MHz). It uses 8 tubes, has a tuned RF amplifier stage and two IF stages (instead of one commonly found on consumer radios) and a device called a Beat Frequency Oscillator that means it can be used to receive Morse code, not just audio signals.

It’s really interesting to think about how capable of a device this is, given that its circuitry is of the same order of complexity of maybe a dozen transistors…about the same number a modern children’s noisemaker would have and nothing compared to the 731,000,000 transistors employed in a modern Intel Core i7 CPU.

An action shot of the underside, recapping in progress, starting from the right and working left. This radio was either built from bucket parts (unlikely, given it’s from a decent brand and wasn’t their cheapest model) or was repaired several times over its life, but always well. There were 6 different brands of paper capacitors: Aerovox, Micamold, Standard Condenser Co., Micamold, Cornell-Doublier, and El-Menco. I’ve replaced all paper capacitors with film capacitors with a uniform 630V rating ignoring the varying lower voltage ratings of the original capacitors. The small values were either replaced with 500V Mica or 1000V Disc capacitors of the appropriate capacitance.

Size comparison of old and new capacitors together.

All the old caps replaced, including the across-the-line cap (blue, top left) and the electrolytic filters (middle). I mounted a new terminal strip using a #6-32 screw through an existing hole in the chassis for two of the capacitors, and used an empty grounded lug from an existing terminal strip for the third. Some would remove the can capacitor, cut it open and re-stuff but I don’t find that to be a good use of my time.

A perspective shot showing the different components. It’s pretty busy down there!

All the replaced components. Four mica caps, one molded paper cap, 18 wax paper caps and two resistors.

So, flipping it back over for the first power-up:

First thing I noticed, the dial light is out. That’ll need to be replaced. The radio does crackle to life, though, and once I get the positions to life does weakly receive KIXI 880 in perfect alignment. But it has an overpowering volume-dependent hum and low sensitivity, missing many stations I know to exist there. We’re about six hours of hands-on labor to this point. Time for troubleshooting:

The hum is volume-dependent, which means it’s (almost certainly) not an issue with the power supply. Power supply wiring errors, like insufficient filtering due to a failed or missing filter cap, usually shows up as volume-independent hum. I’ll check the third filter capacitor, though, as it’s on a different ground lug. The radio has a Radio/Phono switch on the front which changes the audio source from the detector output (radio position) to the phono input on the back (phono position). When set to the phono position the hum disappears entirely. This indicates the problem lies in the RF section. I suspect the 6SG7 tuned RF amplifier has an internal heater-to-cathode short, which will put AC hum from the AC-powered tube heater into the DC cathode circuit. It could also be a short circuit to ground around one of the other RF tubes (converter, an IF amplifier or the detector) but I doubt this, a short around a functional tube would likely take it out entirely versus a leaking short internally.

Testing the tubes revealed they were all good. Several tested nearly new. All were identically branded Hallicrafters tubes to match the set which was even more interesting, I wouldn’t be surprised if they were the originals included when the set was sold. I relocated the ground from the above foreground electrolytic from the terminal strip to an unused solder lug but this didn’t correct the problem. Poking around, I discovered that the problem seemed to change with vibration and motion and the easiest place to see this effect was changing the position of the speaker. I initially suspected an intermittent in the output transformer (seen to the left of the black EPCOS capacitor):

but several tweaks around the transformer, including substituting another speaker, eliminated it from consideration. I desoldered the speaker leads and replaced with a bench speaker:

This didn’t fix the problem, so I reversed the modification and worked backwards from the output stage to discover that I’d missed a solder connection on the detector output. While there, I replaced two 200pF mica capacitors that were looking a little ragged for good measure.

With that repair made, I powered it up and it worked very nicely. AM comes on very well aligned with good tone and all the controls function. On the first shortwave band I picked up 5 stations but two of them are broadcast in Seattle. On the second, I picked up two copies of the BBC World Service and a news station in Spanish. On the highest shortwave band I picked up what I suspect are data signals and location beacons but no voice transmissions. These additional troubleshooting steps took about 2 hours, bringing the running total up to 8 hours of repairs.

I peaked up the IF transformers, but the RF alignment was “good enough” for me so I didn’t do a full RF alignment. This was built like a tank originally, and I didn’t need to replace any of the components in the RF sections, so it should be fine for quite a while. After reassembly, I put it up on my shelf:

Now I have a working general-coverage communication receiver. It’s right near my desk, so hopefully I’ll be able to spend some time cruising the dial and seeing what else is out there. Shortwave reception is heavily influenced by atmospheric conditions, it’s different every time!

Total project time for this one was about nine hours and consumed around $20 of on-hand consumable parts. For a parts list, see the service manual.