Had an American Admiral hot chassis 120VAC AA4 radio, the kind that used an autodyne 12AU6 converter, no IF stage, and the usual 12AV6 50C5 and 35W4 circuits. It was insensitive, and it had a boring plastic cabinet anyway. So...

At a garage sale picked up a Sony AM-FM "transistor" radio that had a beat up cabinet. I figured I could use this radio's circuit board to replace the RF sections of the above AA4.

In step stages, I removed the FM specific parts from the SS board, making it AM only. Tested it to see if I broke something. Removed its tuning cap and measured it with a cap measurer. The antenna was 150μF max, and the osc 100μF max. The AA4's original tuning cap is antenna 400μF max and osc 100 μF max. This means I can use the SS board's osc coil, but tring to match the antenna tuning cap to the SS antenna needs work. Never did get satisfactory results, but I then decided I could make use of the AA4's loop antenna, and install a tap to feed the SS chip's AM antenna input. Looked at the data sheet of the chip, a CXA1019S and it had a spec for the AM antenna listing the number of turns for the LC circuit and turns for a secondary to feed the chip. Which gave me a turns ratio that I can then pick a point on the loop antenna to install a tap, here at 20%. And this allowed me to keep the original loop and tuning cap to preserve the original antenna LC circuit. The osc tuning cap was very close, and the SS osc coil is adjustable, so mating the original tuning cap to the SS osc coil was no problem. Tested it at this point, to see if any mistakes happened, and to tune things up.

With this success, now it's time to merge this above chip to the audio driver (changed it to a 5719 submini tube, you may need to lower its plate resistor, I used a surface mount resistor in parallel to get it to 270K), output and rectifier tube circuits. Need a 4V power supply for the chip, which I can get from the 50C5 cathode circuit. This cathode runs around 6V, so I used a resistor voltage divider to get the 4V. And this divider lets me use a cap to filter the 4V to remove cathode follower induced audio. But still letting me have some cathode follower feedback for the 50C5. The chip's speaker output now feeds the AA4's volume control. The "transistor" radio's volume control I set to give me a reasonable level of audio to feed the AA4's pot, selecting a undistorted audio level, aka "line level" that you'd get from the old tube AM detector circuit. Used fixed resistors to replace its old volume pot, once I determined the setting of it.

Used an AC voltage dropping cap to drop the missing 30 volts of now missing heater. This saves 4.5 watts of heat inside the radio cabinet. This means the radio must operate off AC powerlines, but have you ever encountered DC powerlines in the past 50 years?

Packaged it up, and externally the radio looks the same as before, but it really pulls the stations in. And we still have AA5 tube sound. Note that the tuning cap and the loop antenna is now a low impedance path to the powerline, through a 1.25V regulator circuit inside the chip. So be sure you can't touch it when the radio fully assembled, else you could take a shock and probably fry the chip.

As most Aussie radios use power transformers, this issue should not be present.

Well given the original lineup I can understand why you would do this!

Some years back I had a Japanese-made AA5 that looked like it had come from Singapore. The 230v to 115v part was original factory fit and consisted of an autotransformer, so still live chassis.

The 35W4 rectifer had shat itself and cooked the autotranny.

I found a 240v to 50v transformer (with some lower voltage taps) from Jaycar that was slightly larger than the original. It would fit, anyway.

The 50 volts powered the 50C5 heater, the other 3 tubes in series went to the 35 volt tap. The 50 volts also went to a full-wave voltage doubler with silicon power diodes to give me about 150v B+.

Worked a treat and now the chassis could be earthed! Much safer!

BTW, how do you post the pics in your postings? Brad's been a bit busy lately so I assume you did it.

Interesting exercise Wa2ise! Along time ago I scrounged a 5 * 9 pin valve set with two short wave bands and little sensitivity. I improved it by adding two transistors (BF115s?) and three 455kHz ceramic filters between the IF out and detector in. This livened it up and gave selectivity which on a middling to weaker signal could entirely tune out the carrier so the program sounded like SSB. The transistors were powered by a diode and filter cap off the 6.3V heater supply.

This the radio in all its glory in the kitchen atop the fridge.

I thought the heat from the tubes might cause the local osc LC circuit to drift (the can with the red slug is the L, not that far from the 35W4), but this radio seems quite stable.

To attach images use the img src html code for having images appear in web pages

"less than symbol" img src="path to the image" "greater than symbol"
(if I actually wrote "<" and the other one your browser would get confused)...

Thanks that works!

I had another Japanese AA5 that was sold to work on 240v - a tiny pink one.

The voltage drop to 110v was done with a series capacitor! What could possibly go wrong?

It was slow to warm up and, because the tube heaters have a positive temperature characteristic and were running from what was effectively a current source, they would glow more brightly the longer the radio was on. Hmmmmm.....

Hey, throw 'em together and get 'em out the door quickly!

An AA5 with a cap to make it run off 240V:
I did some playing around with that
(my web page describing how I did it).