NFB calculation question - 4-4-0 / October Studio

[tristanc]

Looking at the NFB / Presence section of the 4-4-0 / October Studio schematics - excerpts attached - I was interested in the frequency of the presence control, calculated using

f = 1 / (2PI C1 Rf//Rs).

By my calculation, using C1 = 0.68uF, Rf = 5.6k, Rs = 25k), the equation gives f = 52Hz. Seems rather low...

Also, the feedback fraction

B = Rk / (Rf + Rs)

comes out as B = 0.82. Which seems a bit high? Surely this makes the presence control more of a volume control?

I made a mistake on my recent build (also attached - along the lines of the October Studio) with the value of Rf, giving f = 136Hz and B = 0.35. Slightly closer to typical values of 350Hz and 0.1. I plan on changing my Rf from 33k to 5.6k but thought it would be worth checking - as the value of B would be 0.76. Again very high.

Is there some reason why such a value for B would be advantageous? Is is the choice of 12AY7 and 6SN7?

[tristanc]

And for interest, here's a Bode plot of a clean sine wave passing through my build with NFB off, NFB on with presence at 0, and NFB on with presence at 10.

So ~12dB NFB is being applied and the presence control acts down to ~150Hz. And the usual case of there being no impact until the pot is about 6 or 7 on the dial.

[Stevem]

Are you using the October output section also, if so you'll find that if your normal usage of the amp has your playing at 50% or more of the amps wattage then a presence / feedback circuit is pretty much a waste of time and parts.

Your selector switch is kinda one way around that , but your pickup outputs if using a few different guitars can throw a big Monkey wrench into the mix!
Just taking about my findings here are some 25 amp builds.

[tristanc]

Are you using the October output section also, if so you'll find that if your normal usage of the amp has your playing at 50% or more of the amps wattage then a presence / feedback circuit is pretty much a waste of time and parts.

I'm using the 6SN7GTB output as per the last schematic for the studio, but I've tweaked the 12AY7 LTPI to be more centrally biased. I've read through the discussion of its development and the goal was touch sensitivity and sounding big at low volume, I think. I can only assume using so much feedback sharpened the transition between clean and distorted or complemented the 6SN7 / smaller transformer to make it 'big'.

But the increase in headroom of the 12AY7 means it would be tough to get enough signal swing? By my reckoning 8Vpp open loop, to 32Vpp closed loop (4 times - 12dB feedback). Maybe it was to enable more useful range on the Master Volume knob? I did notice it was difficult to get power amp distortion when scoping, but I'll need to take another look. Not sure I get 32Vpp after the tone stack...

[tristanc]

I found the posts where Brian (GainMaster) talks about some of the values used in the 4-4-0.

https://ampgarage.com/forum/viewtopic.p ... 150#p64193

https://ampgarage.com/forum/viewtopic.p ... 150#p64214

And, yes, it seems the changes were to control the clean-to-mean element / blooming / feedback / instability. So more feedback and a presence control that operates over a wider frequency range.

At the higher gain settings you have to turn the presence pot down some to make it clean up well.

Maybe Brian will see this and comment.

[tristanc]

And for anyone in the future, here are a couple of posts from AX84 with a bit of detail on the values used. I'll quote the interesting bits here in case the links go dead. But yes, lots of NFB used to make the small output section sound big.

I've played with using a 12AX7 in my build and I actually prefer the sound at low volumes (using a scale control to get PI distortion). I'll try the 'correct' NFB values at some point and report back.

http://www.ax84.com/bbs/dm.php?thread=353719

(Iain) I've a feeling you could manage it with EL84s, or even something really tiny like a PP 6SN7, if you did it the same way Cliff has the new October Studio setup. Uses tons of NFB (well, maybe not tons, but a fair bit) and a voltage divider in each leg of the PI output to force the output stage to stay cleaner for longer.

May not quite get the 'smack in the face' character of a big PA (you need power to do that, generally), but it should be possible to get a similar tone and dynamic response from a small PA. How much tweaking you'd need to do to nail it exactly is another matter entirely ;-}

No, you need the negative feedback so the poweramp runs clean most of the time. What separates a big sounding amp from a small one is the punch the PA delivers. In a big amp you get that when you strike a note/chord with the amp turned up reasonably high, because the preamp signal is big enough to overcome the NFB and the transition from clean to PA crunch is very abrupt.

But as the note decays the NFB starts to come back into play and it tailors the way the note trails off, too. Both attack and decay are part of the 'high gain' dynamic.

With a small poweramp, you'd normally have a problem in that the signal from the preamp is so big that it doesn't have any trouble in driving the small output tubes into a squarewave before the attack even reaches close to it's maximum amplitude, so you don't get the 'kachunk' effect of the NFB suddenly being muscled out of the way at a point where the signal is large, loud, but until now much cleaner. And on the decay, you have to get down nearly to silent before the NFB kicks back in. Sounds OK, but doesn't have the right dynamics for you to really be happy with it.

With the October Studio setup, there's plenty of NFB to keep things clean until the signal is pretty damn big, and the voltage dividers help out somewhat, too, so you can get big amp response from a little one.

There's obviously more to it than that, because the low power has different psychoacoustic effects, can't drive the speakers as hard (which adds to the punch) and doesn't have the physical energy for the sound pressure to be something you can feel as a strong pulse in your chest. But it'll get you closer than anything else I've seen so far.

http://www.ax84.com/bbs/dm.php?thread=372478

Try using the feedback resistor value from the 4-4-0 or October Studio schematics (5K resistor instead of 100k)

The problem with low output power amps, is you have to feedback a higher portion of the output to get the same level of NFB.

The presence pot can only provide as much 'boost' as the NFB is suppressing, so unless you get the feedback into the 3-6dB range, it really won't do anything. IIRC from working on my October Studio, the 22k/5k pot was only providing something like 1dB feedback, which is hardly noticeable.

Actually, you might want to rip the whole power amp from the 4-4-0/Studio, rather than use the 2W PP. It's been pretty thoroughly debugged now, and sounds great.

The original design of the 4-4-0 studio used a 12AX7 in the PI with a 10K tail resistor. With that much gain in the PI, voltage dividers were used to control the signal level on the grids of the 6SN7. Through experimentation, Brian and Cliff discovered that a 12AV7 or a 12AY7 in the PI had lowered the signal enough to dispense with the dividers. There was a net increase in the useful tone range and character of the PA. In the end, the 12AY7 was selected because it is still in production, though somewhat expensive. The feedback resistor is 5K6 and the presence pot was increased from 5K to 25K, increasing the amount of NFB.

The 4-4-0 area contains a good bit of development history and experimental results for this amp. Cliff and Brian are the ultimate source of information about the 4-4-0.