Gibson tremolo oscillator - where's the cathode resistor?

[Pierre]

image.jpeg

Hi all,
A few years back I fixed a Gibson ga-15rvt that someone had given to my son's guitar teacher. I remember the reverb and especially the tremolo sounding pretty good on this amp. I have been looking at the schematic and it appears that the cathode on the oscillator is connected to ground. Can someone explain how this works? Does this just shunt the signal at the phase inverter transformer to ground? Any explanations are appreciated.
Thanks,
Pierre

[martin manning]

The oscillator triode is grid-leak biased. Its grid leak resistor is made large enough so that a negative charge accumulates on the grid.

[Pierre]

Martin,
I'm confused as to how the cathode could be more positive than the grid? Is that unimportant in this application?

[martin manning]

The cathode is at zero volts (ground), but the grid becomes negative. See here under section 1-33: http://www.rfcafe.com/references/electr ... 1-1-40.htm

[jjman]

That one is placing a modulating signal on the grids of the output tubes via the PI transformer secondary. That pushes the bias up and down.

[Pierre]

The cathode is at zero volts (ground), but the grid becomes negative. See here under section 1-33: http://www.rfcafe.com/references/electr ... 1-1-40.htm

Martin,
Thanks for the link. That was a great explanation of how the tube can operate. I'm thinking of building a stand alone reverb and tremolo unit and using this oscillator. I'm going to breadboard it first.

[trobbins]

When setting that sort of circuit up, you need to check if the oscillator valve is acceptably biased. You normally can't do that by measuring the grid voltage, as the meter resistance will load the grid and change the bias voltage significantly. The plate voltage can be measured during idle, and the grid leak resistance modified to a value where the idle plate voltage is acceptable. However that also has issues, as the bias changes with signal. So the better method is to check the plate voltage swing (min and max values of the resulting signal waveform), and confirm they lie somewhat symmetrically between B+ for the stage, and the expected minimum voltage that the anode can swing to (estimated from datasheet curves and a loadline). A short-cut is to check the signal voltage waveform at the top of the intensity pot (as it has no DC voltage on it), and confirm it is not being clipped or grossly distorted across the frequency range.

One hassle is that you normally need a scope to check the signal waveform. You can use a soundcard based scope or spectrum analyser, but that would need something a bit special as the response needs to be down to circa a few Hz.

[sluckey]

A good meter such as a Fluke 87V can check the tremolo rms voltage and frequency down to a couple hertz. Scope is even better, but I've started annotating my schematics with the rms voltage at a particular frequency, usually 5HZ.

[trobbins]

Yes, we don't sort of worry about meter or measurement system frequency response for audio gear too much. Tremelo and vibrato raw frequencies are getting seriously low, and will catch some test gear out.

The only other application I recently had to go down to a few hz to measure was low frequency stability in hifi amp (Williamson), where some unwanted gain and phase shift can sometimes cause unexpected rumble issues. Pretty much gone are the days of record player rumble, but step transients or movie audio content for 5.1 subwoofers are more common now.

[Pierre]

I'm afraid that the most sophisticated piece of test equipment that I possess is an extech muiltimeter that I bought used on eBay. My build experience is limited to cutting and pasting parts of different circuits together and tweaking from there. I appreciate all of the knowledge and advice that I continue to receive from everyone here. I hadn't noticed a preamp stage without a cathode resistor before and wondered whether or not this oscillator relied on another part of the circuit to function. My thoughts are now leaning towards building the re-vibe instead of experimenting with designing another cut and paste build.

Pierre