Repair/modify/upgrade on hot chassis GE solid state AM radio

[wa2ise]

Picked up at a garage sale a GE solid state clock radio, circa 1969 or so. All transistor, hot chassis, uses a high voltage transistor for the audio output stage. It worked, but weak stations couldn't be heard, moderate strength stations were distorted. Figured the detector diode, here the BE junction of a transistor, was too much.

I had some small shotkey rectifier diodes I salvaged from some small switching power supplies that had a very low diode drop, around a quarter volt. The transistor as diode measured about 0.55V. But I thought that these diodes tend to have excessive capacitance inside them, and make lousy detectors, but I decided to try one by touching the ends to the circuit board lands where the old detector was, once I removed it. It worked, and pretty well too. Better than the old detector. As it was a surface mount part I soldered it to the trace side of the circuit board.

[centralradio]

GE was famous for those line operated table radios and clock radios.Be careful around the HOT dropping power resistor plus the high voltages.

What was the big deal for them not putting a power transformer in the set.Copper was not that expensive then.

I'm glad you got it working better.

[wa2ise]

Originally Posted by centralradio

What was the big deal for them not putting a power transformer in the set.Copper was not that expensive then.

I'm glad you got it working better.

Thanks. I suppose the high voltage transistor was cheaper than the transformer. The clock is an analog mechanical Telechron movement, the same as used in their tube sets. And partly company tradition? When electronic digital clocks came out, those needed a power transformer, and then it was a short jump to install a radio circuit just like the kind used in battery operated portables.

Oh, on further review, I decided I should check the old carbon comp resistors to see if any drifted out of spec. Which may have caused the problem I "solved" with a better diode. I did find a few, like a 1.2K that went to 1.45K, a 18K that went to 22K, and a 1.5K that went to 2.2K. As these are around the detector circuit, I replaced them with new resistors that are the same as the drifted values, idea being that the new resistors won't continue to drift. Now that I modified the detector diode, new resistors of the old values may make it not work as well. If you know that you are doing this, it will be fine to do. Trouble comes in if you change something that accidentally compensates for other unfound defective parts.

Also replaced two small electrolytic caps in the front end with surface mount ceramics (lower ESR, less internal inductance, thus better RF bypassing)

[radiotvnut]

I just fixed a 4-transistor GE clock radio from around '68, one of the ones that uses a case that looks like the older tube sets. The oscillator was dead in mine and that was caused by a resistor in the converter that had way up in value and there were some bad electrolytic capacitors that were causing squealing. The radio now works fairly well, for a 4-transistor set.

GE liked using dropping resistors because I have a 1980 AM/FM with one; however, I have two early solid state ('64ish) GE's that have a power transformer.

[dieseljeep]

Originally Posted by centralradio

GE was famous for those line operated table radios and clock radios.Be careful around the HOT dropping power resistor plus the high voltages.

What was the big deal for them not putting a power transformer in the set.Copper was not that expensive then.

I'm glad you got it working better.

They all seemed to do that, including Zenith. The high voltage silicon audio output transistor had a little higher wattage and they were able to use an output transformer, that was used in a tube set.

[wa2ise]

Had to more work on this set, all related to the Telechron clock radio clock switch. The little plastic piece that forces the switch contacts open broke. The epoxy I tried to use on that plastic piece didn't work worth a ... Fabricated a new piece, and the mechanicals turn out to be more fussy than I expected. And that spring clip that holds the switch contacts to the clock decided to make a break for it, springing away. It was a real PITA to get to attach to the little holders anyway (how the hell did they do this at the factory without the boss screaming about productivity?) I ended up using a small tyrap, after a few iterations of the switch contacts never closing. This was a real PITA...