NYD one bottle, more saturation?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

aronaut

Well-known member
Joined
Apr 15, 2005
Messages
69
I'm going to experiment with my one bottle, trying to get more tube compression from it, maybe have it switchable.  am I correct in thinking if I increase the size of the cathode resistors I can under-bias the 12av7, reduce headroom and make it saturate earlier? 

thnx
A
 
:D
I had similar thoughts, installed a pot on cathode in my G9, practically no change in sound..
Now, overdriving is a different story, try slamming it with some strong signal...
 
You can bias colder or hotter, but bear in mind that cold biasing can sound a little more fuzzy/hard.
 
Yeah, people generally don't seem to get this.  You want that gtr amp/'50's R&R saturation thing you lose the NFB loops.  Most post-'55 designs seem to run 25+ dB of feedback, and it's clean until it hard clips kinda nasty.  Go 10 dB NFB or less.  Go none.  It's fine. 
 
NewYorkDave said:
Reduce or eliminate the negative feedback, you'll get all the saturation you want.

I remember asking about this a few years back, and you actually did a quick PDF, "no-feedback one bottle" or something. I did a quick proto of it and it worked well.

I guess a cool option would be simply to add a pot in the feedback loop, and you could dial in as much or as little feedback (linearity) as the task needs.
 
I searched for the 'no-feedback one bottle' pdf, but no luck: http://www.groupdiy.com/index.php?topic=16727&highlight=, what's up with the www.twin-x.com/groupdiy can never find what I'm looking for there.  anyone still have the pdf?  do you just remove R12, R11 and C5? 
Kingston, would you add the pot where R12 is?  I'm still learning how to reduce negative feedback, gonna get out my old navy electronics vol. 1....
onebottlepreampschem.gif
 
NYDave, I love my one bottles by the way, really really appreciate your contributions here... :)
 
The way you reduce the feedback is to increase the gain. (Well, actually it's the other way around: the way you increase the gain is to reduce the feedback. But the effect is the same.)

So turn the gain pot wide open. If you want even less feedback, make R12 100k instead of 10k (maybe switchable. Or put in a switch which disconnects the feedback loop (R12 plus the pot and coupling cap) altogether. Then you're running wide open, will have lots of gain and (almost) no feedback.

Peace,
Paul
 
aronaut said:
I searched for the 'no-feedback one bottle' pdf, but no luck: http://www.groupdiy.com/index.php?topic=16727&highlight=, what's up with the www.twin-x.com/groupdiy can never find what I'm looking for there.  anyone still have the pdf?  do you just remove R12, R11 and C5? 
Kingston, would you add the pot where R12 is?

Sorry I was thinking of mila-1, which has the gain pot between the first two stages.

But in one-bottle, a decent solution would be to leave the feedback as it is already, and just add another 100kOhm log pot between the two stages (like the first two stages in mila). If you set feedback pot to high resistance, you will have massive gain, which you can then control with the other pot between stages.

Another distortion option would be to starve the B+ line. give it only 120VDC and see if you like that.
 
ok, so I tried all the methods stated above (except for starving the B+), they all do the trick but the output is too hot and clips at the console before I can even hear the results, in fact It clips with the feedback pot all the way up and no modifications. 
I guess the only solution here is an output transformer or output transformer + attenuator
 
1. Make R7 a log pot, with the wiper to the second stages grid

2. Add a switch to remove the feedback loop

3. Add a switch to put a cap in parallel with r5.  Maybe start at 10 or 20uf.  Tweak to taste.

This will allow a few options:  Leave the 1M pot fully open and the feedback connected and the pre will behave as stock.  Crank the gain/feedback pot and back off the 1M pot and be able to vary to amount of distortion.  Disconnect the loop for more saturation, and put the cap in circuit for even more gain and saturation.
 
Thanks Baltimore, just tried it, almost works, but with R7 as a pot the 12av7 gets super microphonic at lower resistance levels, starts ringing... :-\
 
Kingston said:
Sorry I was thinking of mila-1, which has the gain pot between the first two stages.

But in one-bottle, a decent solution would be to leave the feedback as it is already, and just add another 100kOhm log pot between the two stages (like the first two stages in mila). If you set feedback pot to high resistance, you will have massive gain, which you can then control with the other pot between stages.

Interstage gain voltage divider pot in the middle of a NFB loop?  You realize they counteract one another?  This is why you don't see interstage gain in amps with overall NFB. 
 
> except for starving the B+

Why not?

Replace 10K 1W with 47K 2W. (47K 1W will work for months; 47K 1/2W will survive a few minutes before smoking, long enough to hear.)

> you don't see interstage gain in amps with overall NFB.

Actually, you see it far too often in Modern Design.

It's all about Features and Sound, not Smart.
 
emrr said:
Interstage gain voltage divider pot in the middle of a NFB loop?   You realize they counteract one another?  This is why you don't see interstage gain in amps with overall NFB. 

Yes of course they counteract. But this was band aid for more distortion. With a single pot switched to those two modes it would work better and then one pot would not make the other pointless. But as already noted above, the overall gain with no feedback is a LOT before it goes to any useful saturation, so output attenuator would be needed.
 
A stepped, ganged feedback/output attenuator control would allow you to vary the amount of feedback while keeping the output level relatively constant. It'd take a little fiddling to get it all to track correctly, but it can be done.
 
that sounds interesting Dave, but this is while using an output transformer right?  reading back through the threads I don't think it was possible to have an output attenuator without a transformer.  wouldn't it be more flexible and less complicated to have them on 2 separate attenuators?
 
Kingston said:
emrr said:
Interstage gain voltage divider pot in the middle of a NFB loop?   You realize they counteract one another?  This is why you don't see interstage gain in amps with overall NFB. 

Yes of course they counteract.

Just checking; I don't know who knows what, and most people don't get this, and could easily read this thread (up until that point) and never have a clue.    I can't tell you how many times I've seen someone hack an interstage pot into an RCA BA-1A or BA-11A and wonder why it was all f*(&ed up. 

Kingston said:
But as already noted above, the overall gain with no feedback is a LOT before it goes to any useful saturation, so output attenuator would be needed.


Actually, many of the classic 2 stage circuits lacking NFB loop, and with interstage gain (or possibility) have pretty similar headroom in each stage.  In other words, roughly the same gain would clip either the 1st or 2nd stages regardless of the others presence.  So, a real healthy input into a 1st stage will almost always overload the 2nd stage pretty quickly, and many times well before recorder headroom is reached.  All depends on the stoutness of that 2nd stage of course.    Plenty of vintage tube pre's can have a flatlined output several dB below +18 dbm, and an interstage pot in one lacking NFB can dial in whatever brown you want.   

Y'all be sure to do your own thang!
 
aronaut said:
...reading back through the threads I don't think it was possible to have an output attenuator without a transformer. 

I'm not sure but I think that a L-pad on the output would do the thing!? Maybe someone could suggest some resistor values?
 

Latest posts

Back
Top