I've been working on a spring reverb driver circuit, based mostly on figure 5 from this site: http://sound.whsites.net/articles/reverb.htm. While there's lots of information on the how of the circuit, there's not so much on the why - I'm not entirely sure what the op amp side of the circuit is doing - as far as I can see, C3 decouples, C6 bypasses high frequencies to the inverting input, and R5 and R7 set the level of bypass. R13 is in the feedback path, making this a current amplifier rather than a voltage amplifier. This arrangement should make it a modified non-inverting amplifier? Past there it's a fairly standard push-pull arrangement, with the reverb input coil connected across the 300R load.
I should note that I copied the circuit from the website onto a breadboard, and it worked fine (albeit noisily, because breadboard). I have a video and scope traces from the output stage, so I wasn't dreaming...
Here's the screenshots of my Kicad schematic - after the success of the breadboard circuit I felt comfortable ordering some PCBs from Seeed studio, which arrived a couple of weeks ago.
I built the op amp side of the circuit, omitting the push/pull stage, and hooked my 'scope up to pin 7. I applied a ~1V RMS 1KHz sine wave to the input, and tuned RV1 until I didn't see any clipping on the output. However, the output was saturated to -Vcc, with only the positive side of the sine wave appearing. Assuming I'd made a mistake with the circuit design I re-tested the original circuit from the website on a breadboard, with the same results. I don't recall this happening with my first breadboard circuit.
Can anyone spot where I'm going wrong, or are they able to explain the theory of operation for the circuit better than the website I copied it from, so that I can do some more in depth troubleshooting?