Single Ended Amps - Opinions

[roberto]

No, a zobel won't absolutely do the same. What a zobel does is flanne the impedance seen by the amp at high frequencies, when the speaker goues up and the zobel goes down. Feedback has a totally different effect.

PS
When you have read that PP cancels even harmonics, it doesn't cancel the even harmonics that comes from the preamp, but cancels even harmonics generated into the power amp (if perfectly balanced, and most instrument amps aren't). A SE amp can be 2nd dominant up to clipping, 3rd dominant up to clipping, or switch from 2nd to 3rd during the working range up to clipping. Other harmonics usually fall below those first two. A PP amp is 3rd dominant, then 5th, etc... but you can add even harmonics by unbalancing it.

[Tobyk]

Thanks for the explanation Roberto. May I ask what you prefer, SE or PP, and why?

[roberto]

I've used both and I still have both, with high and low gain sounds. Not a specific preference a priori.
PP are more common, but I've played with a 11W SE with the band, and indoor was louder than needed for pop-rock.

[bepone]

@tobyk
feedback in the meaning of reducing output impedance. why is this imortant? because single ended stage( and PP too) with "weak" output topologies with hi output impendance acting like current source. how is behaving current source and how voltage source with the speaker?

now here we are. current source is following speaker impedance curve. voltage on the speaker is i*Z. how is looking Z on average speaker? having raise on higher frequencies. how did you describe sound with poor SE stage? having too much presence?

if you measure with the oscilloscope on the amp output frequency curve while load is real speaker you will see immediatelly your problem.
lowering output impedance you are transferring from current to voltage source ( far from ideal ),
then your freq. curve is not following speaker impedance so strictly, bumps are only on speaker resonance (80-120Hz) rest is straight.

also NFB is reducing distortion (higher order 3H, 5H) but not 2H...so you can experiment and find what works best for you.. you can look also Shade NFB, or linearising pentode curves (not common in guitar world)

[Tobyk]

Thanks @bepone @roberto
When you say feedback around the output, where does one connect that? It’s normally connected to preamp cathode. Do you connect it to power tube cathode instead?
Would removing power tube bypass cap (creating local feedback) yield similar results?
Do you think zobel filter is still useful in any situations?

[bepone]

you can have simple NFB like in fender champ,
but also better versions like Shade NFB , Local NFB Anode to grid on output tube, and Local cathode feedback on output tube. I will not say which works best you need to try all of them. If you start now in a few months you will have all the answers!

[bepone]

zobel filter is suppressing speaker impedance raise on higher freq. providing more constant load to output stage. read it like reducing the highs, experiment and then see if ok. make it R+C, R should be normal impedance of output stage transformer (2.5k-5K depending of the tube)

[Tobyk]

I’ve never seen any schematics on how to do this kind of local powertube feedback. For anode-to-grid LFB (which I imagine is second best to shunt/Schade feedback), is it just a wire with a resistor of say 100K? between anode and grid (connected before grid resistor) on the output tube? And a mixing resistor of say 100K? before that?
Cathode local feedback (no bypass cap) I have tried, too little effect!

[R.G.]

Single ended
Pros: Preserves harmonics better than Push pull
Closer, more organic, direct feel. Often less bassy
Cons: Often harsher overdrive. More trebly.
More one-dimensional tone.
Agreed? But the cons can be remedied, I believe!

I don't agree with this as I read it.
Doesn't "preserves harmonics" mean something like "doesn't reduce the harmonics you already have"? Both SE and PP stages can have low distortion, preserving whatever harmonics exist in the signal coming in.
Can you expand on "closer, more organic feel"? What exactly does that mean? I have some guesses, but I don't think I understand what you really meant.
As to "Often less bassy", SE amps have a problem with bass response all built in, but it's an economic problem, not inherent in the setup. SE amps need much larger OTs than PP amps for the same bass response because they can only use 1/4 of the magnetic field swing in the core. Economics (... "MBA Disease") tells the makers to provide the least bass response they can get away with to save some money. To expand on that, hifi gear in the early 60s often had SE output stages, and did quite well with bass, as they needed to cover well under the typical guitar tuning's 82Hz lowest note. in my earlier post in this thread I mentioned my favorite amp for tone of all time. It used an SE output from a then-high-end home hifi. No lack of bass there. I would class the concept of SE= less bassy as the amp designer shooting themselves or their customers in the foot.
Same for "often harsher overdrive". There isn't any particular reason this has to be so. Overdrive can be controlled and set in the design from little to more to mind blowing. Just like plastic bobbins on OTs do not necessarily mean worse OTs, SE does not necessarily mean harsher overdrive. The trick is in being careful about which stages distort and how much.
"More trebly"? It's all in the frequency response. If the designer cheaped out on the OT, there won't be much bass and it will sound more trebly. Of course, a few caps in the preamp and the treble from the preamp can be cut down to match.
"More one-dimensional tone"? Can you describe that in more detail? I don't understand what you mean in a way I can relate to the actual amplifier circuit.

My net on this is that as the song said, you gotta know when to hold 'em, know when to fold 'em in terms of amp design. It's good that you're trying to not hold firmly to generalizations from few examples. There are many examples of SE amps like what I think you may mean; but I can peg a number of PP to those observations too.

[bepone]

I’ve never seen any schematics on how to do this kind of local powertube feedback.

you need to do some investigation, try to visit diyaudio.com

For anode-to-grid LFB (which I imagine is second best to shunt/Schade feedback)

shade would be from anode of output stage to anode of preceeding stage, you can start with 1M and then slowly to 470k

, is it just a wire with a resistor of say 100K? between anode and grid (connected before grid resistor) on the output tube? And a mixing resistor of say 100K? before that?

start with 1M, or 470k 2w s in series, connect anode to grid, and check bias if is changed,

Cathode local feedback (no bypass cap) I have tried, too little effect!

cathode feedback is localNFB around output tube and means isolated (special) secondary winding of output transformer connected in series with cathode!
i dont have time now to make new picture with the tubes for you, but you can see here some of my creation, transistor output single ended stage with source/cathode in tube world/ feedback.. Vfets, tubes, same principle of operation

vfet.jpg

[Tobyk]

I’ve never seen any schematics on how to do this kind of local powertube feedback.

you need to do some investigation, try to visit diyaudio.com

For anode-to-grid LFB (which I imagine is second best to shunt/Schade feedback)

shade would be from anode of output stage to anode of preceeding stage, you can start with 1M and then slowly to 470k

, is it just a wire with a resistor of say 100K? between anode and grid (connected before grid resistor) on the output tube? And a mixing resistor of say 100K? before that?

start with 1M, or 470k 2w s in series, connect anode to grid, and check bias if is changed,

Cathode local feedback (no bypass cap) I have tried, too little effect!

cathode feedback is localNFB around output tube and means isolated (special) secondary winding of output transformer connected in series with cathode!
i dont have time now to make new picture with the tubes for you, but you can see here some of my creation, transistor output single ended stage with source/cathode in tube world/ feedback.. Vfets, tubes, same principle of operation
vfet.jpg

Thanks bepone! Sluckey pointed me to Bassman AB165, which has anode to grid LNFB. It has only 220K resistor though.

[bepone]

you can do that ., but start with 470k and then reduce to 330k, 220k if necessary

[Tobyk]

Will try 470K, thanks. Strange that this late 60’s Bassman is the only Fender using this kind of feedback.

[roberto]

shade would be from anode of output stage to anode of preceeding stage, you can start with 1M and then slowly to 470k

anode of preceding stage is grid of the same stage, if you keep the RC out of the audio range. Being a guitar amp in SE, I suggest to split the 1 MOhm into two 470kOhm in series, 2W each (not because of dissipated power, but because of the voltage rating): a SE pushed close the the resonance frequency of the speaker can have quite a flat loadline and then go quite high in anode voltage swing. Better to be on the safe side.

[roberto]

Thanks bepone! Sluckey pointed me to Bassman AB165, which has anode to grid LNFB. It has only 220K resistor though.

It depends on the load of the previous stage, but I do not recommend to push this feedback too far as you will lower the Zout of the amp and flatten the response on lows and highs.