To NFB or not to NFB ? - Geloso G1 1020 A conversion with Merlin Medium Gain pre

[kagliostro]

As someon knows having read the other thread about the search of a circuit with el34 and Concertina PI

a friend is converting an OLD Geloso G1 1020 A PA amp into guitar amp

I submitted him some preamp (many thanks to the friends that gived theyr contibute) the choice was the Merlin Megium Gain Preamp

that uses 3 triodes (1 and 1/2 12ax7 tubes)

Question is:

The original schematic has an NFB



but the use of a pre with that architecture as far as I know impede the use of the NFB because the TS is at the end of the chain, between the last triode of the pre and the Concertina PI



Is there a way to have the NFB avoiding to include on the pat the TS, if I'm correct a TS is not to be placed within an NFB circuit

Many Thanks

Franco

[sluckey]

The TS is not inside the NFB loop.

[kagliostro]

Ciao Steve, Grazie

On the original schematis I see the NFB connected to the  cathode of the first triode of V2, Pin #3 where is connected the 22K Cathode resistor

that triode is used in the Merlin's preamp to drive the TS so, the NFB if connected there will have the TS in the path and as far as I can know that is not a good thing

Franco

[pdf64]

Putting a tone stack and master volume inside a NFB loop is probably not consistent with what's normally considered good electronic design practice.
Certainly the effect of the NFB will affected by the settings of those controls.
But many guitar amps have presence, resonance and master volume controls inside the loop.

[kagliostro]

Grazie pdf64

Franco

[66Strat]

A possible option to global negative feedback could be a local NFB loop from plate to grid of the output tubes. Other modifications suggested are to isolate the master volume from the cathodyne, reduce HF roll-off, and prevent blocking distortion. See attached.

[kagliostro]

Many Thanks 66Strat

Very interesting

I had the schematic under eyes a lot of times, never noticed the difference between the resistors that feed G1 (1K vs 47K)

who knows why it was done like this

The mod is to add the plate to G1 resistor and cap for each tube, remove the components on the red boxes and add the components on the blue boxes

exuse my ignorance can you explain reduce HF roll-off please

I think that to try to control the level of local feedback via a pot isn't a good idea because of the High voltage on the pot (dual ganged one), correct ?

Thanks again

Franco

[pullshocks]

This may or may not be related to your question, but the response of that tone stack gets a little strange at the extremes. You can see it in the tone stack calculator so I don’t think it was my amp (SE 6V6, no NFB).
But overall I liked Merlin’s design a lot.

[kagliostro]

Many Thanks also to you pullshoks

We will mount the TS components on the pots so it will be easy to correct values if required

Thanks for the advice

---

To be honest, I would have preferred one of the pre that were suggested in the other thread, the one of the VOX NT15 because the Tone Stak is between the first and second triode, so I could have kept the negative feedback at the cathode of the third triode

But the amp isn't mine and the guy that has it want the Merlin preamp


Franco

[66Strat]

HF = High Frequency. The components in the blue rectangles act to shunt high frequencies to ground (about 21,000 hz.) These components were likely added by the circuit designer to stabilize the amp and prevent parasitic oscillations.

The 47K grid stoppers, in addition to preventing blocking distortion, work in conjunction with the tubes' miller capacitance to roll-off high frequencies in the 20k hz range. The components in the blue rectangles in conjunction with the 47k grid stoppers would lower the high frequency range to about 10,000 hz.

IMO, it is never a good idea to put high voltage on a potentiometer. One could swap the position of the series resistor and capacitor so that a potentiometer could be implemented. But, other than having an additional knob to twist, I can see no advantage.

[kagliostro]

Many Thanks

HF = High Frequency. The components in the blue rectangles act to shunt high frequencies to ground (about 21,000 hz.) These components were likely added by the circuit designer to stabilize the amp and prevent parasitic oscillations.

Do you mean in the RED rectangles ?

The 47K grid stoppers, in addition to preventing blocking distortion, work in conjunction with the tubes' miller capacitance to roll-off high frequencies in the 20k hz range. The components in the blue (or red ?) rectangles in conjunction with the 47k grid stoppers would lower the high frequency range to about 10,000 hz.

The spech sheet say a bandwidth from 30Hz to 15,000Hz this is in line with what you say

... other than having an additional knob to twist, I can see no advantage.

Well, the motivation is to have a control on the FeedBack to set the level you prefer, may be this operation can be done one time and the pot swapped for fixed resistors, after all we don't know the impact of this new (Local) FeedBack on the sound of the amp

One question is open ... why they put a 47K grid stopper in one tube and one 1K grid stopper on the other ???

Many Many Thanks Again

Franco

[66Strat]

Many Thanks

HF = High Frequency. The components in the blue rectangles act to shunt high frequencies to ground (about 21,000 hz.) These components were likely added by the circuit designer to stabilize the amp and prevent parasitic oscillations.

Do you mean in the RED rectangles ?

The 47K grid stoppers, in addition to preventing blocking distortion, work in conjunction with the tubes' miller capacitance to roll-off high frequencies in the 20k hz range. The components in the blue (or red ?) rectangles in conjunction with the 47k grid stoppers would lower the high frequency range to about 10,000 hz.

The spech sheet say a bandwidth from 30Hz to 15,000Hz this is in line with what you say

... other than having an additional knob to twist, I can see no advantage.

Well, the motivation is to have a control on the FeedBack to set the level you prefer, may be this operation can be done one time and the pot swapped for fixed resistors, after all we don't know the impact of this new (Local) FeedBack on the sound of the amp

One question is open ... why they put a 47K grid stopper in one tube and one 1K grid stopper on the other ???

Many Many Thanks Again

Franco

Yes, the components in the RED rectangles are there to squelch high frequencies.

Rather than using pots to find an appropriate resistor, one could clip resistors in parallel with alligator clips to find the desired level of negative feedback. A single 1 meg would be provide little feedback. Two 1 Meg resistors in parallel (500K) would have more negative feedback. One 1 Meg resistor in parallel with one 500K resistor would provide 333K resistance, still more negative feedback. This approach would be like going to the eye doctor and looking through the lenses that the optometrist would switch (better or worse?).

I don't why know the engineer(s) decided to use a 1K grid stopper on one tube and a 47K grid stopper on its compliment. The 47K grid stopper has about the same HF roll-off as the 50pf capacitor in parallel with the 150K grid leak resistor. However bigger grid stoppers have the advantage of better mitigation/prevention of blocking distortion.

[kagliostro]

Ciao 66Strat, Thanks Again

This approach would be like going to the eye doctor and looking through the lenses that the optometrist would switch (better or worse?).

 

I've been to eye exams many times, you couldn't make a better comparison as an example


 

A friend told me that also some Marshall has difference in grid stoppers, this is because the result will be that in the transient you have a momentary increase of the second harmonics, for this reason, in his opinion, they should be kept different


Thanks Again


Franco

[sluckey]

The TS is not inside the NFB loop.

My mistake. I only looked at the ORIGINAL schematic. 

[kagliostro]

Absolutely NO problem 


Franco