QS sent me a link to a Youtube video of a young lady who explained the use of a light bulb in a Wien Bridge Oscillator and who mentioned that HP long ago used a light bulb in its audio generator.
The HP 200CD audio oscillator was a part of the workbench of a guy I worked for. I could not figure out why they used a large dual ganged variable capacitor for the frequency dial. Sure, it didn’t have the wear problems that a potentiometer has, but HP could have used a custom made pot that had multiple wipers and whatever else it would take to make it hold up under heavy use. And I bet it would have still cost less than the variable cap. Nowadays, an air variable capacitor can cost fifty or more dollars U.S.
Later I bought a Heathkit audio generator that used a light bulb and used switches to control the frequency. This was all solid state (transistors) and was a nice piece of test gear. The HP 200CD used vacuum tubes, but even though it got quite warm, and was power hungry, it still was very stable.
I have built a few of the WBO circuits using transistors and a light bulb, They worked very well, a lot better than the PSOs (phase shift oscillators). The light bulb is low resistance, under 800 ohms (28 volts divided by 35 mA), so the output stage that also drives the light bulb has to be able to drive a few hundred ohms – the light, and any load on the output. One issue with the light bulb is that at the lower audio frequencies, 100 Hz and below, the light may have too fast a response and cause the oscillator to gallop or waver. The light’s response is getting close to the low frequency.