Magnatone M30 output oscillation

[booj]

So far I've changed the outputs to 2N3055's (ST), Replaced the bias resistor which was originally set to about 130 ohms, to a combo of resistors set to about 320 ohms. I have replaced a 4.7ohm resistor that read 8 ohms and a leaky 470uf cap that was leaky. I also replaced the snubber cap which was a .01uf. (Maybe it should be a larger value?)
The schematic does not agree with the board in a number of places and is very hard to read. The gain and the center pots work and the output rail is centered and the gain is set for close to max.
Problem is, with a resistive load the power amp works fine. About 18 volts before clipping and no crossover distortion. With a speaker load, however, there is a really bad high frequency oscillation on the top of the waveform. The cooler I set the bias the worse it gets. the outputs aren't getting really very hot at this point, so I've got room for more adjustment, but there's no crossover distortion, so, well... I'm stuck.EsteyM30[1].pdf

[booj]

I tried clipping a .047 across the Zoebel cap and the oscillation went away. I was able to reduce the bias current without problems. Still fairly clueless about what's going on.

[booj]

Since the power amp seems to work now with the .01 II .047 cap, I hooked up the preamp panel and the vibrato/trem didn't work. Bulb was burned out (1869) the circuit uses two LDR's and one bulb in a metal enclosure. I tried replacing the bulb (10v @ .014A) with two blue ultrabright 3mm led's, one directly over each LDR, and a common 2.2k resistor. It works but has a bit of a hard edgy quality to it. I tried paralleling two red leds with the blue ones but only the red leds lit and it didn't sound as strong. Anybody got any experience with this circuit and using leds instead of a bulb?

[Enzo]

The thing is the bulb has some lag. If you turn the current on to a bulb, it takes a second for it to come up to full brightness. then if you switch off the current, the bulb takes a second to cool and darken. and if you pulse the thing, it cycles smoothly up and down in brightness. LEDs don;t work that way. APply current and they are instantly full on, then kill teh current and instant dark. That might be why your results are hard and edgy. Possible a cap parallel the LED might add some of that lag, don;t know.

LEDs have a natural junction voltage, and color makes a difference. The voltage across an LED has to go over that junction voltage threshold before the thing starts to conduct. And once it does start to conduct, the voltage across it tends to stay steady. If a red LED has a drop of 1.2v, and a blue one has a drop of something like 3v, and they are in parallel, then as the voltage across them rises, it gets to the 1.2v level and the red LED turns on. Increasing the voltage higher still doesn;t change the drop across the red LED, just the current through it. SO the voltage across the parallel blue one never gets up to that 3v it needs to fire. SO it stays dark. If you want both red and blue to glow, put them in series.


Bulbs produce more or less continuous light across the color range - white in other words. I don;t doubt the photocells react to fifferent wavelngths of light - colors - and since your LEDs operate at extremely precise frequencies, I am not surpised the red and blue LEDs might have different performance.

[Gtr_tech]

Cool white LEDs *should* have a closer light output to an incandescent. You'd need to play with an R/C circuit to mimic the performance....series resistance and a cap to gnd parallel with the LEDs. Or....replace the lamp

[booj]

I wound up using the two blue leds and a 2.2uf cap across the pair to slow things down a bit. Doh! I shoulda put the leds in series! duh. That way I could have had both the red and the blue going. I was so amazed that it worked first try that I left it in. I know Mouser has the bulbs.

[Gtr_tech]

Oops...that shoulda been "warm white" LEDs....not cool white.