Monster Mosquito 60, ca. 1976: Question about feedback (schematic)

[Elk_Instruments]

[Edit: fixed the wrong cap values on the schematic]

Hello all!

First post for me. I'm Jens von Weissenberg, luthier and tinkerer. I have been restoring/refurbishing/repairing this Monster Mosquito 60 hybrid head (Finland, mid-70's). The preamp + EQ are transistor based, but the PA section looks a lot like a 50W Marshall (see schematic). I have a couple of questions:

1. Has anyone seen a setup like this in some other amp: the G2 sources are additional taps on the primary side of the output transformer. The DC resistance between the center tap and the G2 taps is about 22 Ohms.

2. If I were to add a feedback loop here, what would your best suggestion for it be? Or would it just be a bad idea or unnecessary?

3. The amp innards have taken _a lot_ of heat, so the BIAS feed circuit is fried. What would, in your opinion, be the things I should consider in constructing a new BIAS feed? Voltage, current?

Thanks in advance!

Best regards,
-jens

[Elk_Instruments]

Update: I have been informed that the construction here would be referred to as an ultra-linear amp. Hence no feedback loop. I will keep hunting for more info, now that I know what to look for, but two of my questions still stand:

- has anyone seen this configuration in an existing amp (make/model etc?)
- any educated guesses on the bias for the EL34's in this configuration?

Thanks!
-jens

[sluckey]

The simple Marshall style bias circuit would probably work very well in that amp, but we would need to see a power supply schematic to be certain. Or, we may be able to tell you more if you post some high rez pics and/or a schematic of the power supply.

     http://sluckeyamps.com/misc/Marshall_Dual_50.pdf

But... It's very unusual to "fry" a bias circuit. Why not just repair your circuit? Probably should not need a schematic just to fix a bias supply.

[thetragichero]

holy crap does that really use 1uf caps to and from the phase inverter?!

[PRR]

Welcome.

You CAN run long-loop NFB with UltraLinear. While Williamson himself didn't, nearly all the "American Williamson" Hi-Fi amplifiers ran heavy NFB on UL to get show-off damping factor.

But this is a guitar amp? Mostly, heavy damping of a guitar stage amp makes it "lifeless". The classic Fenders ran about 6dB NFB. Using UL tends to give roughly 6dB NFB, boom, done. (Except you don't have a place for crazy NFB tricks like Fender's Presence knob.)

UL *tends* to survive stage abuse a bit better than Pentode (yet louder than triode). When the plate pulls down, G2 is pulled down too, which limits the peak current. IMHO the Fender "135W" on four bottles would have to be rated 100W if not for the UL connection.

Your schematic shows no "BIAS feed" and no photo of the inside of the oven, so I can't help.

[sluckey]

The Sunn Model T has a negative feedback loop that would easily fit into your schematic.

I recommend you fix what's broken before modding the amp.

[66Strat]

holy crap does that really use 1uf caps to and from the phase inverter?!

This does not appear to be the original schematic. Note the 7.12.21 revision date and the KiCad EDA 5.1.12 date. The 1 uF is likely a transposition error in converting nano-farads to microfarads. The actual value is most likely 0.1 uF.

[Elk_Instruments]

[EDIT: replaced the PSU drawing with the latest update]

Thanks for all the replies!

First off, the caps are 0.1 uF, that is, if I've read them right (see attached photo).

The Sunn Model T has a negative feedback loop that would easily fit into your schematic.

I recommend you fix what's broken before modding the amp.

Aye, that is what I'm going for as a first step 

I've attached the PSU, with some question marks. The old mains transformer is fried (and without any markings..), and doing my best guesswork, I've replaced it with a Hammond 290HX.
You can see that there has been no fuse(s) for the heater taps - I am going to add a fuse there.
As for the BIAS circuit, I've drawn what I could see from the old parts. I do not know what the trim pot has been - maybe 100k? Also, the diode there is unreadable, but I guess it should be a fairly fast one - any suggestions?

The ca 30VDC for the solid-state preamp and EQ has been made with two ZPY 30 V zeners, a bunch of resistors in front and two polcaps. I'll have to simulate and/or calculate the actual current need of the pre+EQ to see how much really is needed.

Although Monster Amps back in the 70's were sort of a factory, this thing sure looks, maybe not hand built but rather self built.

The choke has no markings on it, but I had access to a LCR tester and got roughly 13 mH (@50 Hz, 5V - I know, a far way from ca 500V).

This is so much fun - i might have to build my own amp next 

-jens

[Elk_Instruments]

I updated the schematic of the power amp with the correct cap values.
-jens

[sluckey]

As for the BIAS circuit, I've drawn what I could see from the old parts. I do not know what the trim pot has been - maybe 100k? Also, the diode there is unreadable, but I guess it should be a fairly fast one - any suggestions?

I would just use the component values in the bias circuit I posted in reply #2. You will probably need to lower the value of the range resistor since you will be using a 98v AC source from that 290HX. I would start with 100K and change value up/down as needed. Oh, and be sure to put a resistor in series with the bias pot. Your schematic is dangerous because it allows you to set the bias voltage to zero volts! 

[Elk_Instruments]

Thanks, Sluckey!

I’ll follow your lead and redraw the BIAS feed. I’ll come back when more questions arise.
This is a great forum and the amount of know-how and info here (and on the site) is quite something!

[PRR]

I updated the schematic of the power amp with the correct cap values.

The BIAS caps will be happier with positive grounded?

[Elk_Instruments]

I updated the schematic of the power amp with the correct cap values.

The BIAS caps will be happier with positive grounded?

Yes - and in real life they are. I will update my drawing. Thanks for pointing this out!

[Elk_Instruments]

Hello all,

one more BIAS-related question. Coming from a solid state world, I know I have a lot to learn about tubes. i have looked for and read several articles about power tube (EL34) biasing, and also found one very practical instruction on how to set the BIAS current.
however, in designing the BIAS circuit, it would really help if I knew what the voltage range should be. If I simulate this circuit (as suggested by Sluckey, see picture). I get a range of about -42V to -49V. The circuit feed is the 98 VAC tap from the mains trafo.

Various sources show the BIAS voltage, for a 50W Marshall-style 2xEL34 setup to be everything from -35V to -63V.

The tubes are JJ standard El34's.

What voltage range is needed, in your opinion?

[sluckey]

I would want an adjustment range of -25v to -50v. Then the EL34 correct operating bias voltage should fall somewhere near the middle of the pot range. I like to use a 50K pot with a 47K resistor to give a nice wide adjustment range. Then I only need to change the bias range resistor to make my bias circuit compatible with EL34s or 6L6s.

If you want to use that 25K pot in your circuit then change R3 to 27K. Otherwise, use a 50K pot and 47K resistor. This will give you a wider adjustment range. Then set the bias pot to mid range and experiment with the value of R1 to give -40V at pin 5 of the EL34s. You can safely do this with the EL34s removed from their sockets.

My bias circuit is attached...

[Elk_Instruments]

Thanks a bunch, Sluckey!

This was exactly the kind of advice I was looking for. I am used to working with opamps and transistors, but the sheer, um, dynamicity (if that's a word) of a tube circuit feels both exciting and overwhelming at times 

 

[PRR]

What current do you think is suitable?

G1 voltage needed will depend a LOT on G2 voltage.

There are two "extremes" (we may sometimes exceed these).

Cutoff: Vg1 ~~ Vg2/Mu(g2) will set plate current very very low. In any sane amplifier, "cut off". However over-volted guitar amps sometimes idle this cold.

Hot: Vg1 at half the cutoff value is very rich. In fact RCA's old litt suggests 0.6 as a happy class A trial point.

Mu(g2) is about 10 on 6L6 6V6 EL34 6550, but 18 on EL84.

[Elk_Instruments]

Thanks for your reply, PRR!

So, if I get this right: in this amp, being ultra-linear and Vg2 coming off the roughly 43% tap(s) of the OT, it (without measuring) is something like 490V. This would mean that cutoff in this case would be roughly 490V/10, ie 49V. And 0.6 of that would be roughly 30V.

Right! This actually lines up with what the old BIAS circuit was set to deliver. To summarize, I'll start cold (around 40V like Sluckey suggested) and then adjust as I measure and test.