12AU7 Hybrid power amp - Evil_Food

[joecool85]

**Note**

ORIGINALLY POSTED BY user Evil_Food

Something strange was happening with the forum software and this thread wasn't viewable so I had to re-make it.

______________________
Hi guys,

I was browsing Schematic Heaven and I found a Musicman PP power amp design that uses hybrid cascodes consisting of transistor on the bottom and a valve on the top.
I started analysing the design and decided to make my own low power version with a 12AU7 double triode.



The output transformer is a £12 3.5W Oxford Electrical Products Audio Transformer http://uk.rs-online.com/web/p/products/2106475/?searchTerm=3.5w+audio+transformer&relevancy-data=636F3D3226696E3D4931384E44656661756C74266C753D656E266D6D3D6D61746368616C6C7061727469616C26706D3D5E285C772B5B2D5C2E5C732F5D292B285C772B293F2426706F3D3926736E3D592673743D4B4559574F52445F4D554C54495F414C5048415F4E554D45524943267573743D332E357720617564696F207472616E73666F726D65722673633D592677633D4E4F4E4526

If you click on the "Range Overview" tab, there's a picture with the turns on each tap on each side of the transformer. The datasheet says 16k as maximum impedance, but if you do the calculations you can get other values.

What I think might be more interesting is the power supply. It uses a Cockcroft-Walton voltage multiplier to step up and rectify the voltage from 24VAC to ~280V DC. My main reason to do it with a multiplier was to save some money, as I already had a 2x12V power transformer. This way I can heat the tube with 12Vac and provide +/-12V DC.



What I liked in the voltage multiplier was one of its shortcomings - poor regulation. The power supply has high output impedance and when the amplifier enters class B operation, there's a voltage sag on the HT rail.

I haven't looked with an oscilloscope, but the DVM shows about 30V sag whit "harder playing".

I haven't got a preamp for it yet, but I quite like the sound when it's overdriven on its own.

I haven't got a camera or anything I can make a sound clip with, but next week I'll ask a friend (who's actually a musician) to bring along something to record what it sounds like.

If anyone is interested in the voltage multiplier power supply I can post the equations needed to design one.

[Pilot1]

very clever,
but why common grid output?
guitar needs 100Hz - 8kHz bandwidth
CG is very wideband...

[J M Fahey]

Hi Pilot 1.
Yes, you're right, Guitar does not *need*much above 6 or 8 KHz, which won´t be reproduced by the speaker anyway, but bandwidth does not hurt either.
Any popular Power Chip , from TDA2003 to TDA7294 easily meets 50KHz or more bandwidth, so no intrinsic problem with that.
The schematic posted above follows the very clever Music Man ideas (search http://www.music-man.com/techinfo/old-amps.html)
As an example:www.music-man.com/techinfo_old/old_amps/2165-rd_&_2100-rd.pdf
They could achieve incredible feats:
1) slamming brutal 700V on plates (650V on 6L6) without burning them; in fact they´re known for their reliability.
2) Pulling 65 REAL watts from a pair of tubes, up tp 130W from a quad.
3) no tube biasing or matching necessary, the SS side of the circuit takes care of that.
4) they could drive those tubes fully with a simple dual Op Amp and a couple mid power transistors
5) Loud, clean sound, way beyond what, say, a Twin could provide (the reference amp up to that point).
6) not surprisingly, it was a "real" Leo Fender product and concept ... although he couldn´t say so for legal reasons.

Leo, you are my man !! 

[Loudthud]

The bias seems to be set by Q3 and RV1. Probably not a big deal but that part of the circuit could be improved for better temperature tracking.

It is my understanding that Leo was not involved in the electrical design of these amps.

[J M Fahey]

It is my understanding that Leo was not involved in the electrical design of these amps.

Agree.
Don´t think even the earlier ones were 100% his work either, besides the first, simplest ones (which practically came straight out of datasheets and tube manuals anyway).
I see him more as a *concept* and *sound* genius than anything else, not forgetting his marketing and organizational genius too.
He was one of those rare, one of a kind guys, who can build a team of geniuses, each in his own area, and have them work together towards a common goal (his own).
He didn´t play guitar either; yet ..... ?

[Yazoo55]

I realise this is an old post - I saw this and wondered if there were any sound samples available now. I'd like to give it a try. Any additional information would be welcome. I would be especially interested in how to bias the 12AU7.

[J M Fahey]

The power amp is reasonable, but avoid the multiplier if you need any power, get HV from a proper HV winding.
Bias with a scope until you get rid of the crossover notch, we are talking very low current anyway, almost pure Class B

[blackcorvo]

The power amp is reasonable, but avoid the multiplier if you need any power, get HV from a proper HV winding.
Bias with a scope until you get rid of the crossover notch, we are talking very low current anyway, almost pure Class B

What about something like this smps with the MC34063 for the high voltage?

http://www.imajeenyus.com/electronics/20111010_40-400V_supply/index.shtml

And if I'm not wrong, you could use another chip to get -12v. I bet you could make it all run off a single 12v gel battery or a cheap laptop-type supply.

[EDIT]

I have run some simulations of a similar circuit on Proteus and I got to the following results, with the output adjusted for 280v with no load.

- 5K Ohms load: 280v, 56mA (15.68 Watts)


- 3K Ohms load: 247v, 82.5mA (20.37 Watts)


So, in theory, this could be a suitable power supply for small tube projects (probably up to 10w, depending on the tubes used).

[blackcorvo]

I have acquired the MC34063 and most of the components for the HV supply, but they did not have the MOSFET, fast diodes, or even the inductor on the local stores I went to. Even told me that high-current inductors were unavaiable for being "old stuff", go figure!
Gonna have to buy online.

Now, I'm very curious about the design process of such an output stage. Is there info anywhere about how to do it? I looked up "hybrid cascode push pull" but I didn't find anything close to this specific topology.
Also, seeing as each triode is biased separately, does it mean one could build a single-ended version of such design?

[Enzo]

Of course you can build a single ended.   It is just a triode stage, use two for push pull.

Look up either "grounded grid" or "common grid" amplifier, both mean the same.  The signal is fed into the cathode instead of grid.  Music Man amps are maybe the best known examples of this.  The various Peavey VTX series amps use it.  Classic VTX, Heritage VTX, etc.  And others.

[blackcorvo]

Of course you can build a single ended.   It is just a triode stage, use two for push pull.

Look up either "grounded grid" or "common grid" amplifier, both mean the same.  The signal is fed into the cathode instead of grid.  Music Man amps are maybe the best known examples of this.  The various Peavey VTX series amps use it.  Classic VTX, Heritage VTX, etc.  And others.

I couldn't find anything talking of this type of circuit using tetrodes nor pentodes, only triodes, and it didn't have any formulas, only brief comments on how the circuit works.
I was thinking about using a 12AQ5 SE with this topolog,y and the SMPS I posted here, so I could use a SS preamp with it.

[Enzo]

The principals applied to triodes still apply to pentodes.  Common grid is still common grid.


Maybe I am lucky, to me the cathode drive seems intuitive.  The tube works on the relationship between grid and cathode, not necessarily those elements and ground.  Hold cathode at ground and wiggle the grid, you get signal at the plate.  But hold the grid to ground and wiggle the cathode, the same thing happens.

In a common 6L6 push pull output, I might ground the cathodes and have -50v on the grid.  In a PV Heritage VTX, I might see +15 on the grid and +65 on the cathode.  In both those cases the grid is biased -50 with respect to cathode.  In practice, the amp examples I cited are running closer to class B than something like a fender twin Reverb would be.

[blackcorvo]

The principals applied to triodes still apply to pentodes.  Common grid is still common grid.


Maybe I am lucky, to me the cathode drive seems intuitive.  The tube works on the relationship between grid and cathode, not necessarily those elements and ground.  Hold cathode at ground and wiggle the grid, you get signal at the plate.  But hold the grid to ground and wiggle the cathode, the same thing happens.

In a common 6L6 push pull output, I might ground the cathodes and have -50v on the grid.  In a PV Heritage VTX, I might see +15 on the grid and +65 on the cathode.  In both those cases the grid is biased -50 with respect to cathode.  In practice, the amp examples I cited are running closer to class B than something like a fender twin Reverb would be.

I prefer colder-biased amps anyways. The best way I can describe it is that they're louder and clearer.

For now, I'll try to simulate a single-ended 6V6 @ 250v in Proteus with the cathode circuit from the amp in this topic, and just play around with values. It's what I usually do on my bench anyways, but until I have the SMPS working, Proteus will have to do.

My problem isn't really figuring out how common-grid works, but the math behind the values used for the circuit. I'm no engineer, I barely understand the math behind most simple amplifier stages (and don't ever ask me about filters, it's far too complex to me!), but I wanna start being more careful with my power amp designs.
Don't wanna make a tube-burner, after all.

[Enzo]

I am a master troubleshooter, not a theorist, so others will be more help there.  But if you look at some other schematics like various Music Man designs, or those Peavey Classic VTX and heritage VTX examples, you may gain some insights.  Look at what is different and what is the same. 

[J M Fahey]

Now, I'm very curious about the design process of such an output stage. Is there info anywhere about how to do it? I looked up "hybrid cascode push pull" but I didn't find anything close to this specific topology.

Compare it to playing and writing Music.

I doubt there's a book on "How to write Rolling Stones songs" (not holding my breath, given their popularity there *might* anyway) but there are tons of "general purpose" books on different aspects of Music, including chord combinations, rhythm analysis, probably explaining Rock/Folk/Country/whatever types of music, etc.

Here you will find general use Electronics book, plus some dealing specifically with Tube Guitar Amps, but I found the latter always missing something and in any case not going end to end from a clean sheet of paper to a glass and metal chassis on a bench, where you plug a guitar and a speaker and play.

So the proper method is to learn the basic Theory, experiment, see what Commercial makers do , how and why, design your own starting simple and advancing bit by bit.

Eventually your own designs will start to work better and better

Personally I can design a cathode driven power amp, with Tube/MosFet/Bipolar power devices, just with the proper datasheets, go figure.
But it took me many years.

Now, *explaining*  how to do that might take a couple posts, but *teaching*  how to , so you can do it fully on your own, end to end, starting with an idea drawn on a napkin and ending with a working amp, requires the full course, because that's what's needed.