Two unknown cap (so they seems to me) ----> GEM Schematic Added

[kagliostro]

Few time ago I got an old (~ 1970) italian guitar amp

as very often happens is impossible to find a schematic so I'm trying to obtain it via a reverse engineering of the circuit

there are two components that to me looks like disk capacitors

they have a color coded value on it (orange and gray)

Please can someone confirm if they are capacitor and say which is the value of it (I'm not able to find any color code table for caps that looks like those)

Many Thanks

Kagliostro

[sluckey]

I don't think those are caps. I'm thinking thermistors??? Can you draw out a partial circuit showing how they are connected?

[kagliostro]

Hi Sluckey

the circuit is very strange to me and if someone can explain which is its function I'll higly appreciate it

see V3 B - at the right end

Many thanks for any advice

Kagliostro

[sluckey]

Strange indeed! I don't think it's drawn correctly yet. Double check that please.

[birt]

this might help:
http://slfisica.host22.com/colores.html

i have a print of this above my workbench ;-)

[kagliostro]

That part of the circuit seems to be correct

there is a connection between V3 B (ECC82) and the vibrato circuit (the vibrato is implemented with a 6DR7 tube)

I'll go on with the reverse engineering and I'll post also the vibrato circuit when done

--

Thanks for the color correspondence table - anyway that is something to have

Kagliostro

[PRR]

> there is a connection between V3 B (ECC82) and the vibrato circuit

That is what I was wondering. The disks look like Varistors. The push-pull drive from V3b would vary them. Except, in your drawing, V3b isn't doing anything.

If it is connected to an oscillator, then it would make sense. The Varistor capacitance changes with voltage. If V3b is driven hard with a low frequency, the Varistors will var the gain and tone of the signal at V3a.

There was a USA amp using the same idea.

And same problem- you can not get new varistors made the OLD way.

[kagliostro]

Thanks PRR

Sluckey asked to draw a partial circuit so I posted the part of circuit that I've already done

---

OK if those are varistor must to find a new solution ..............

I'll go on with the drawing of the schematic

then I hope to solve the problem with the help of the forum

Kagliostro

p.s.: if someone knows more about the american amp that used the same method I'm interested about, especially to see the schematic - Thanks

[birt]

well they do look exactly like the varistors on that link i posted. value reads out as 338 i think. (orange orange grey)

[kagliostro]

For those who can be interested

Magnatone used Varistor in some of their amps

(thanks PRR without your input I'll not discover that)

Kagliostro

[kagliostro]

So this is the schematic I've draw till now (only PS and Anti Hum circuit are missing)

Please can you say if the draw looks correct ?

the circuit isn't simple and this is my first serious reverse engineering job

(the red cap to me aren't originals and were added after by someone)

Now what about those components ? are they 338v Varistors ?

I've read a bit about the problems about finding spare parts for Magnatone circuits that used Varistors

and would like to know if there is a way to have the same particular kind of vibrato that has now the (GEM) amp, using more usual circuit components

Many Thanks for any  

Kagliostro

EDIT: editable schematic added

[PRR]

See attached.

The 1Meg in the tremolo oscillator C-R-C-R-C-R network probably goes to ground.

The "220K" cap might be 22K? Note the sequence: 22K into ~~300K, 10K into 1Meg, 5K into 3Meg. The three sections should have similar R*C products. There is a slight improvement in performance if the resistors step-up in value; then the capacitors must step-down in value.

I separated the reverb driver cap ground from the tremolo Speed pot ground. They are probably going to the same place, but the drawing is easier to read if they do not "touch".

> a way to have the same particular kind of vibrato that has now the (GEM) amp, using more usual circuit components

No.

Not even using "recent" varistors: they have "improved" to have a sharp break-down, instead of the soft gradual break-down of the old varistors.

[kagliostro]

Thanks PRR

I corrected the error of the 1M resistor that isn't in parallel with the 10k capacitor (as you draw one leg is connected to ground, my mistake) and for more readibility I connected the 47k cap to one other ground instead with the same of the 1M pot

I verified the 220K cap connected to pin 6 of 6DR7, the value is right at 220k, not 22k (or better, in the circuit there is a 220k cap and it seems original)

The 4.7uf connected to pin 8 of 6DR7 tube looks really to be added after by someone, like all the red colored caps in the schematic

Note the sequence: 22K into ~~300K, 10K into 1Meg, 5K into 3Meg.

you must be patient with me, I'm not able to follow you, can you explain again in a different way ?

to which 22K into ~~300K, 10K into 1Meg, 5K into 3Meg do you refer ?

Not even using "recent" varistors: they have "improved" to have a sharp break-down, instead of the soft gradual break-down of the old varistors.

As you told also in a previously answer only old stile varistors can be used for that use

as can be seen in the attached image the response is very different

I was asking if there is a way to simulate an old varistor with new components

like:

Put more than one new varistor in series ?

Use zener diodes with an in series resistor (resistor with a paralleled capacitor) ?

Use LDR with some control circuit ?

Use Mosfet or Transistor with some control circuit ?

Use a bidirectional Transil diode ?

How to vary a resistance in accordance with voltage ?

---

does anyone knows if someone has replicated a Magnatone Vibrato using something else than varistor ?

Thanks

Kagliostro

[PRR]

> explain again in a different way?

The oscillator is the conventional Phase-Shift oscillator.

There is a tube and cathode/plate loads for gain, and three Capacitor-Resistor networks connecting tube output to tube input.

Re-draw it to make it clear: C-R-C-R-C-R network, amplifier, connection from Out back to In.

A C-R network alone reduces low-frequencies, starting from a frequency determined by the C and R values. It also introduces a phase-shift. One C-R is small shift, but three C-R will give 180 degrees of phase-shift at one specific frequency.

It is possible to change the resistance and get the same frequency if the capacitor is changed in the opposite way.

The three C-R networks should all be the same frequency, or fairly similar.

As a refinement, it works a little better if the later C-R networks use larger R (and smaller C) than the earlier networks. (Plate impedance is lower than grid impedance.)

1) 220KpFd + 460K is 4hz (180K to 1180K is 10.3 to 1.5hz)
2) 10KpFd + 1Meg is 15.9Hz
3) 4700pFd + 3.3Meg is 10.3Hz

C-R section 1 is tuned significantly lower than #2 and #3.

After thinking: I think it will work this way.

If you say it works with these values, then of course it does work.

[kagliostro]

Many Thanks PRR for the good explanation

and for redrawing the oscillator

that way it is much more readable

I'll redraw the schematic according to your redraw and I'll post it

Kagliostro

[kagliostro]

Here is the schematic redraw for more legibility

Kagliostro