Global Negative Feedback on AC30?

[Mr. Lime]

I‘ve been reading Aiken‘s explenation about the LTPI and was wonderig about the 3rd PI input he described:
https://www.aikenamps.com/index.php/the-long-tail-pair

Is there a chance to add negative PA feedback at the tail R of the LTPI? Similar like the injection at a cathode of a cathodyne driver stage.

It would be out of phase with one channel of the ac30 so it had to be disconnected when plugged into that channel.

[TUBEDUDE]

No, don't do it. A large part of the magic is no feedback from around the transformer.

[Mr. Lime]

It's more of a theoretical question.
I just want to know the options I would have with a LTPI using the two classical inputs for different channels..

Maybe a variable feedback R could act as a tone control when the feedback is canceling some frequencies of one channel?
Could be interesting in an ac30 with channel link to play both channels parallel.

[R.G.]

It will lower the overall gain ... that's what feedback does.

[roberto]

...and don't forget you have a high cut after the PI, and its effect will be dampened by the nfb.
I wouldn't do that. If you really want to apply any sort of nfb, around 20-25%v distributed load would be better.

[Mr. Lime]

It will lower the overall gain ... that's what feedback does.

Depends on the feedback circuit. If someone want to „boost“ the bass frequencies, a resonance circuit would only result in damping the highs.
Could be helpful for EL84s I guess.

...and don't forget you have a high cut after the PI, and its effect will be dampened by the nfb.
I wouldn't do that. If you really want to apply any sort of nfb, around 20-25%v distributed load would be better.

I see your point. The K60 or Concorde take use of both, hicut and presence. I have no AC30, I‘m just interested in options I have with LTPI.


I assume the tail R could be replaced by a pot as well for presence control? A coupling cap in the fbk loop should be added then..

[roberto]

Depends on the feedback circuit.

Talking about negative feedback, it always reduce gain.

If someone want to „boost“ the bass frequencies, a resonance circuit would only result in damping the highs.

No, it is reducing the feedback at low frequencies, exactly as presence does for mid-high frequencies.

Could be helpful for EL84s I guess.

IMO no, because EL84 doesn't like to be pushed on lows, but rather prefer to work flat and get some cut on lows earlier on the schematic.
At least this is how I prefer them.

[Roe]

NFB on an AC30 seems impossible since these amps feed both sides of the phase inverter. if you remove one side, then NFB will work. But it will change the tone and response significantly. Still, it might be a good idea to clear up the bass response somewhat, while removing NFB from the mids with a presence control

[Mr. Lime]

Please guys, don't hang on the AC30 circuit, it's just a good example where both classical inputs are used.
Imagine you want a clean amp and other big bottles are used instead of the EL84s..

NFB on an AC30 seems impossible since these amps feed both sides of the phase inverter. if you remove one side, then NFB will work. But it will change the tone and response significantly. Still, it might be a good idea to clear up the bass response somewhat, while removing NFB from the mids with a presence control

Yeah, that's my question due to Aiken's article, there's a third input at the tail R where feedback might be inserted.
You mean feedback at the tail R wouldn't work with two seperated channels at the other inputs?

Talking about negative feedback, it always reduce gain.

Basically yes, you won't add gain with it.
IMO the feedback in guitar amps is anyway very little and with a presence or resonance circuit maxed, my ear would say, only mids are damped.
I set the boost in brackets because it's auditive a boost in bass frequencies, while mids and highs are lowered in volume.
In some of my amps I had good results with fixed resonance circuits as I like to cut bass early in the preamp resulting in a tighter sound somehow.

[roberto]

NFB on an AC30 seems impossible since these amps feed both sides of the phase inverter. if you remove one side, then NFB will work. But it will change the tone and response significantly. Still, it might be a good idea to clear up the bass response somewhat, while removing NFB from the mids with a presence control

Feedback on PA can be applied in very different ways:
- OT to output tube screens
- OT to output tube cathodes
- OT to output tube grids
etc...

[roberto]

Please guys, don't hang on the AC30 circuit

I would say the title of the thread is then a little bit misleading, at least.

IMO the feedback in guitar amps is anyway very little

There are alot of different circuits, some with more, other with less, some with no gnfb.
As in Hi-Fi and every other amp, there are circuits with more, less, or no gnfb.

with a presence or resonance circuit maxed, my ear would say, only mids are damped.

You are reducing the feedback on lows and mid-highs, so the negative feedback is more at midband.

In some of my amps I had good results with fixed resonance circuits as I like to cut bass early in the preamp resulting in a tighter sound somehow.

Almost all high gain amps in the last 20 years do so with good results, but it is not easy to set all parameters for all speakers/cabinets.
I use eqs before the PI, and you can set them with this purpose in mind as well. It permits you to have more lows, without the looseness typical of a depth control on the PA.

[R.G.]

It will lower the overall gain ... that's what feedback does.

Depends on the feedback circuit. If someone want to „boost“ the bass frequencies, a resonance circuit would only result in damping the highs.

My bad. Yes, you can boost. But be very careful what you wish for. Simply slapping feedback on an amp designed without it carries a strong risk of making the thing oscillate or motorboat. Negative feedback is nearly always used to linearize things, and there are whole textbooks (I have a few of these) on this topic. Negative feedback can only reduce, by definition. Positive feedback can boost gain, but it's one of those "be careful what you wish for; you just might get it" things.

There are some nuances to that. One - a bass "boost" is nearly always done by arranging more feedback at high frequencies, lowering the gain there. Two - yes, feedback can be arranged to be positive feedback. This increases gain, but makes the whole thing much less controllable. Amplifiers with output transformers have issues built into them by the fact that there's a transformer there. Gain-phase oscillation happens (oversimplifying) when you have more than two time constants (alternatively, poles in the response) and an overall gain above one at the 180 degree phase point. Output transformers give you two poles all by themselves, so the third pole is certain to be there in the circuit. Perhaps more interestingly, OTs also give you two poles at the low end of the frequency response, so you can make it motorboat as well as squeal if you use too much feedback. That's why amplifiers with OTs can't use much feedback - they turn into oscillators.

So yes, you can "boost" with feedback by selective rolloff; you can even truly boost at some frequencies, but you're severely limited by the nature of a tube amp in how well and how much you can do this. The safer form of selective boosts is to use more gain and cut some of it back at certain frequencies. In general, guitar amps are limited to using negative feedback in limited amounts to linearize the power amp, and getting bugger signals out of the preamp to make up for the lower gain in the power amp. Positive feedback "resonance" flirts with instability and needs very close control.

The approach of amplifying everything and then cutting some of it back is in fact how the tone controls in most tube preamps work.

[Mr. Lime]

Just found this schematic of a 36W TMB.
The global feedback seems to be injected in the "3rd input" of the LTPI.
What if the other side of the PI got another channel instead of the 100n cap to ground?
I guess phase issues would occur?

[sluckey]

Just found this schematic of a 36W TMB.
The global feedback seems to be injected in the "3rd input" of the LTPI.
What if the other side of the PI got another channel instead of the 100n cap to ground?
I guess phase issues would occur?

Actually the NFB is applied to the grid (pin7) through the 100n cap as well as the tail of the PI. So there is no available input for another channel. You'll have to mix the channels together and apply to grid pin 2, something like Fender did.