So what's this varistor doing here?

[Leonc]

Being a Magnatone nut, I've mostly thought of varistors being used in Bonham's patented vibrato circuit. I've read that metal oxide varistors are used in some modern amps to provide sort of a "fail safe" for the fuse, from what I can gather.

I do also recall having seen one used in one or two of the highly regarded DeArmond/Martin amps but I don't know what they were doing there. Here's another example. The 12AX7 to the left is providing some kind of recovery function after reverb and tremolo. Between there and the phase inverter 12AX7, which is on the right, is one varistor, with no value or rating specified from what I can tell. Any idea what it's doing there?

 

[dbeeman]

try to ask over at amp garage?

 

[Leonc]

Haven't tried -- guess I should create an account over there...

 

[teemuk]

Seeing not only a small section of the schematic might help in answering more specifically but I suspect it maybe has something to do with duties of amplitude modulation, e.g. tremolo.

 

[diagrammatiks]

it's for the tremolo.

 

[guitarcapo]

Weird that it has a rating. How could it be a "270K varistor"? Isn't resistance supposed to vary with voltage?

I'm looking at the "A" feeding the plate and wondering if the varistor might be functioning to regulate the plate voltage on that tube to make it more steady.

 

[Leonc]

it's for the tremolo.

Ah yes, I can see this now;the lead coming into the varistor parallel to the .022 coupling cap is coming from the tremolo circuit.

Weird that it has a rating. How could it be a "270K varistor"? Isn't resistance supposed to vary with voltage?

I'm looking at the "A" feeding the plate and wondering if the varistor might be functioning to regulate the plate voltage on that tube to make it more steady.

That's the rating for the resistor going to ground; I believe the varistor is symbolized by like a resistor with the "squiggle" in it.

 

[Leonc]

The (S) is going to the Speed pot, the (4) is coming off the Intensity pot. The (A) and (C) are filter sections, as you'd probably guess.

Sorry for bits/pieces, but the schematic is huge.

 

[teemuk]

Any decent image editor software can reduce colours and resize images to make huge schematics not-so-huge.

Anyway, what you basically have there is a 47K + 270K resistive divider biasing one terminal of the varistor to a certain DC offset, a tremolo circuit output modulating this DC offset with a low frequency AC signal (thus varying resistance of the varistor), and an output from a gain stage passing through the varistor to another gain stage, which's input forms another resitive divider with the varistor, hence applying amplitude modulation.

 

[mark norwine]

I'm going to go way out on a limb here:

1.) That's not a "Varistor" (i.e. MOV)...the MOV as we know it didn't exist back when that amp was made.

2.) that's not the schematic symbol for a MOV

3.) I can't understand what purpose a MOV would serve in the circuit right there.

I believe....although I don't know the circuit in question....that what they're indicating is a "variable resistor".....i.e. a pot that isn't wired as a "potentiometer" [voltage divider] but is, instead, wired as a {2 lug} variable resistor.

Have you traced the amp? What does the actual device look like?

 

[Leonc]

Any decent image editor software can reduce colours and resize images to make huge schematics not-so-huge.

Anyway, what you basically have there is a 47K + 270K resistive divider biasing one terminal of the varistor to a certain DC offset, a tremolo circuit output modulating this DC offset with a low frequency AC signal (thus varying resistance of the varistor), and an output from a gain stage passing through the varistor to another gain stage, which's input forms another resitive divider with the varistor, hence applying amplitude modulation.

Thanks a bunch. That kind of makes sense to me (I'm quite the noivce, electronically). I wonder why they would decide to go this route rather than doing all the amplitude modulation with some kind of capacitor/resistor network, the way you see it done in so many other amps...? Was this a cheaper way to go?

I'm going to go way out on a limb here:

1.) That's not a "Varistor" (i.e. MOV)...the MOV as we know it didn't exist back when that amp was made.

2.) that's not the schematic symbol for a MOV

3.) I can't understand what purpose a MOV would serve in the circuit right there.

I believe....although I don't know the circuit in question....that what they're indicating is a "variable resistor".....i.e. a pot that isn't wired as a "potentiometer" [voltage divider] but is, instead, wired as a {2 lug} variable resistor.

Have you traced the amp? What does the actual device look like?

Hey Mark, no, I don't know what it looks like; don't have the amp (yet), but I'm guessing this varistor is similar to the Globar varistors used by Magnatone, Dearmond, et al back in the 50s and 60s.

 

[teemuk]

I wonder why they would decide to go this route rather than doing all the amplitude modulation with some kind of capacitor/resistor network, the way you see it done in so many other amps...?

You don't "see capacitor/resistor networks" doing it. FETs likely weren't invented yet, lamps LDRs, etc. may have been too expensive or rare. The way how the tube amps usually did this was modulating the grid bias voltage, which alters the stage gain. Unfortunately it's also possible to drive such scheme into annoying clipping and it can cause annoying wobble noise from modulation, especially if done in a single-ended stage.

So my hunch is...
- less distortion when AC signal from tremolo oscillator is hitting the peak of positive halfwave
- less annoying "wobbling" noise from amplitude modulation
- the designer felt most convenient doing it this way

 

[Leonc]

By "capacitor/resistor" network I was referring to something like what you see in the Gibson GA-40, where they appear to be modulating the signal to the grid.

Perhaps that's not the right terminology...but you can see what I meant.

Anyway, thanks very much for these insights. Interesting that this mfg (Lo Duca) may have actually had "good performance" in mind. I guess I've become a little jaded by having heard "that was the cheap way to do it" many times in relation to certain amp design/construction choices, LOL.

 

[teemuk]

You are misinterpreting what you are seeing. The RC network is a fifth order high-pass filter, incorporated to filter out the low frequency waveform introduced by the tremolo oscillator.

The preceding gain stage is fed with both (modertaly high frequency) input signal and low frequency output from tremolo oscillator. The magnitude of both waves will affect grid-cathode voltage bias and henceworth the stage's gain but in essence the circuit is configured so that tremolo signal has far greater magnitude in respect to plain input signal, thus tremolo oscillator's output signal will have defining effect on overall gain. The varying amplitude of thay signal will modulate stage gain and henceworth cause amplitude modulation.

If you watch this thing on oscilloscope the input signal can be seen "riding atop" a wave of much lower frequency and much greater magnitude. After amplitude modulation that low frequency content however must be quickly filtered out because it can (and will) easily overdrive following stages and its harmonics will cause nasty wobbling side effects in sound. This is where the resistor/capacitor thing steps in. It's a plain filter, a high order one to be ultimately effective.

 

[trobbins]

If you check Vibroworld, it indicates La Duco copied, or had some arrangement, with Magnatone. That tremolo circuit may have been a way to get close to same effect as Magnatone's vibrato/tremolo effect, without having to pay a licence fee for Maggies Bonham patent - or just to keep up with the other amp manufacturers tremolo offerings of the time.

As such, La Docu would have had stock of the varistors, and I reckon it would have been an 'easy' way to make a tremolo effect during that time. I don't know what the price of an LDR (eg. OPM61) at that time, but I'm betting they would have been more expensive than the varistor.

I'm also suspecting that the combination of LFO drive to the varistor and to the point A (HT to the 12AX7), would have somewhat nulled the LFO signal level, but retained the audio signal modulation via the varying varistor value. This would have minimised the typical issues with LFO signal going through the main amp and speaker.

Ciao, Tim

 

[Leonc]

You are misinterpreting what you are seeing. The RC network is a fifth order high-pass filter, incorporated to filter out the low frequency waveform introduced by the tremolo oscillator.

The preceding gain stage is fed with both (modertaly high frequency) input signal and low frequency output from tremolo oscillator. The magnitude of both waves will affect grid-cathode voltage bias and henceworth the stage's gain but in essence the circuit is configured so that tremolo signal has far greater magnitude in respect to plain input signal, thus tremolo oscillator's output signal will have defining effect on overall gain. The varying amplitude of thay signal will modulate stage gain and henceworth cause amplitude modulation.

If you watch this thing on oscilloscope the input signal can be seen "riding atop" a wave of much lower frequency and much greater magnitude. After amplitude modulation that low frequency content however must be quickly filtered out because it can (and will) easily overdrive following stages and its harmonics will cause nasty wobbling side effects in sound. This is where the resistor/capacitor thing steps in. It's a plain filter, a high order one to be ultimately effective.

Huh? Are you referring to the .005uF caps and the 1M resistors that follow the 2nd 5879? That has a filtering function, as I understand it, so I guess that may be what you're referring to. But I was referring to the "network" of .05uF caps with 47k, 240k and 510k resistors that connect to the plate of the 6SQ7...I assumed they were involved in producing oscillation involved in the AM.

 

[diagrammatiks]

the magnetone circuit is a true vibrato.

the filter circuit in the gibson is a tremolo.

 

[teemuk]

Huh?

Sorry, it seems I misunderstood what you referred to with the resistor/capacitor term. Yes, the circuit you are describing is indeed the phase shift network that triggers oscillation of the tremolo stage.
http://www.aikenamps.com/PhaseShiftOscillators.html

 

[Leonc]

Man, RA's site never ceases to amaze. Thanks for the link.

 

[trobbins]

the magnetone circuit is a true vibrato.

I'm pretty sure if you put a Maggie on the bench you will see some low level of tremolo along with the vibrato - it adds to the magic ;-)