I think this idea was discussed a few issues ago in Stereophile, twice in the same issue! I think it was in some Conrad Johnson review, where CJ recommended this, but Art Dudley wrote something about it that seemed contrary to CJ's recommendation. Maybe I dreamt it all.

I've actually tried this with Leben CS600 push/pull and Harbeth Super HL5. Yes, the 4 ohm tap controls the bass better, but the sound kind of hardens and becomes artificially damped. Very effects-y. The 8 ohm tap doesn't control the HL5s woofers well when you're playing Stravinsky, but on the whole, the music sounds more natural. I've put it at 6 ohms and have psyched myself into thinking it's the best of both worlds.

Is there any potential down side from such a test? I mean can doing it damage the amp or speakers??

The effects on the performance, generally, are this: reduction in output stage distortion and reduction in the available output power. This may be good or bad listening wise but it is not likely to damage the amp or the speaker unless the reduction in power puts you into a constant clipping condition, which would be pretty audible. So don't be afraid to try it and see which you prefer.

Quote:
Feedback circuits are nothing more than automatic frequency attenuators. They cut back frequencies that have the highest voltage output. This in turn allows frequencies with a lower voltage output to surface, creating a flatter, wider frequency response. The more active a feedback circuit is, the more lethargic an amp will become. There is a time delay since this is a loop regulator circuit. Using the lower impedance tap on a output transformer can reduce the feedback circuit's effect on sound quality.

Performance of this experiment will vary according to how the amp is wired and how far apart your speakers' impedance peaks & dips are from one another.

What school taught you that about feedback? I want to know so I can keep those I know from wasting their money by attending.

Quote:
Feedback circuits are nothing more than automatic frequency attenuators.

No.

Just no.

Quote:
You just can't disagree and not provide a reason. How do we know your reason is valid?

Well, I've disagreed and provided only the reason that your statement is just wrong, so guess what, I can and I have.

As to "valid", well, sorry, you're making the extraordinary assertion, perhaps you should look into who has the burden of proof here.

I would suggest Linear Circuit Analysis
by Artice M. Davis

as a place to start. I have several good books, but they aren't in print any more and you can't have mine.

Quote:
Man, that was a cheap way out. You're no fun!

The book isn't cheap, but it is good. Go read it. Then you'll see what's wrong with your summary of feedback.

Quote:
Here is a link regarding feedback:

http://www.aikenamps.com/NegativeFeedback.htm

I stand by what I said. In fundamental terms, a feedback circuit is like a governor, a regulator. It works by monitoring voltages at the output stage (secondary of audio transformer). If the amp sees a higher impedance (which equates to a higher voltage) at a specific frequency, it will attenuate that voltage. With tube amps, this voltage is fed into a cathode circuit. As the cathode's voltage increases, the gain of that section decreases.

You just can't disagree and not provide a reason. How do we know your reason is valid?

I agree Niteshade. This is first year electronics/science. Using a lower tap or negative feedback will lower the output impedance of the amp and typically provide better signal voltage regulation vs varying load such as a typical speaker presents. Of course neither negative feedback nor switching taps are perfect.

Quote:
I have found that using the 4-ohm tap will help compensate for impedance dips. This only works if the speakers' nominal impedance is 8 ohms.

Again agreed. Using a 4 ohm tap instead of 8 ohm tap with an 8 ohm speaker provides a lower output impedance with better signal voltage regulation to a varying load, when a peak or dip occurs.

You have some nerve, Sammett... getting into a pissing match with someone whose chief vocation is audio processing! Christ almighty. you are like a dog that keeps running into parked cars...you keep getting knocked out, jaw broken, but always come back for more abuse.

you aren't choosing your battles very wisely, sammett.

Quote:
You have some nerve, Sammett... getting into a pissing match with someone whose chief vocation is audio processing! Christ almighty. you are like a dog that keeps running into parked cars...you keep getting knocked out, jaw broken, but always come back for more abuse.

?! Excuse me, what were you saying recently about others "turning a thread to shit"?! In a thread where, ironically, you were the first one to launch character attacks on other members? Once again, I see you're the first one to turn a thread into "shit", with yet another attack against Steve. That's an awful lot of hatred and anger for somebody who complains long and loud to others about their lack of "courtesy" toward fellow members. You never answered my question NC. Do you ever read yourself to see how foolish and hypocritical you sound to others? I'm guessing the answer is "no, I have neither the time nor the wherewithal to think about what I write". Because if you did actually think before firing off yet another angry missive of yours, you would realize the folllowing things:

a) Steve is responding to Blair Lamphear, of Niteshade Audio. They manufacture audio products, not "audio processing". So his chief vocation would be audio engineering. Which is more than anyone can say for you.

b) The topic Steve and Blair are discussing is feedback circuits. Not "processing". Last I heard, Steve was an audio engineer. Which is more than anyone can say for you. Steve produces amplifiers and is more than qualified to talk to Blair on the subject. But even if he wasn't, so what? When did that ever stop you from shooting your mouth off on subjects you were buck ignorant of? See what I'm getting at?

Quote:
you aren't choosing your battles very wisely, sammett.

You don't actually do anything very wisely, NC - so save your advice for those that need it. ie. Yourself. As everyone can see, Steve was having a friendly, on topic technical discussion with Blair on the very topic of his thread. So he wasn't "battling". You are. And it looks to me like you're the dog coming back for more abuse. I guess we'll have to see how many cars you need to be hit in the face by, before you stop starting fights, attacking people for no good reason, and trashing perfectly good threads like this.

Quote:

I agree Niteshade. This is first year electronics/science.

Interesting claim on your part. You're claiming time-varying behavior on the part of feedback?

Quote:

Using a lower tap or negative feedback will lower the output impedance of the amp and typically provide better signal voltage regulation vs varying load such as a typical speaker presents. Of course neither negative feedback nor switching taps are perfect.

Quite so.

But having nothing directly to do with feedback. Quite obviously a "4 ohm tap" will expect a different load impedence and have a lower output impedence than the expected load impedence(which does not have to be 4 ohms, but you build tube amps, you undoubtedly know that). This has to do strictly with the output transformer ratio, though, and is only secondarily related to feedback.

Quote:

Again agreed. Using a 4 ohm tap instead of 8 ohm tap with an 8 ohm speaker provides a lower output impedance with better signal voltage regulation to a varying load, when a peak or dip occurs.

True.

But this isn't the kind of description of feedback that was described as "Feedback circuits are nothing more than automatic frequency attenuators. They cut back frequencies that have the highest voltage output. " in any meaningful sense. This describes the results of some types of feedback, in some rather confused language, but does not describe the cause, rather than the effects.

This does not change the basic point of the thread, that a 4 ohm tap is extremely likely (I'm trying to imagine a setup that would not do this, and I think it would have to be a pathological design) to have lower output impedence as well as lower expected load (or power transfer) impedence.

I wonder if either of you has ever taken an amplifier with a 4, 8, 16 ohm winding, grounded the 4 ohm tap, and used the "0" and 16 ohm (those having the same winding ratio to the primary) taps as cathode taps.

This, if you have the gain structure right, turns the whole amplifier into a cathode follower, but of course with a very large gain from the plate circuit.

Works great with a pair of 8417's, or 6L6GC's, and ok with 6550's. Never tried it with 7027A's. Be fun to try it with 811A's but never had the chance. Be an interesting transformer design, that one.

Quote:
I'm glad we've started discussing feedback. I forgot that feedback was a somewhat touchy topic. My description was basic and does describe a gain decrease with an increase in voltage across a cathode created by a speaker having a higher impedance at specific frequencies. There are other things going on, so a blanket statement that feedback is nothing more than an automatic frequency equalizer was simplistic. I initially wasn't looking past feedback's effect on frequencies at high impedances. Use of the 4 ohm tap does have desirable effects depending on the speaker system. I have never tried grounding the 4 ohm tap and using "0" & "16" for the output.

The premise behind using the 4 ohm tap was this: Impedance dips cause attenuation. The 4 ohm tap should be close to most dips in a nominally 8 ohm speaker, alleviating the attenuation.

That is right Niteshade.

One of the effects of negative feedback is to lower the output impedance and provide better signal regulation, so once again you are correct Niteshade.

Me:

Quote:

I agree Niteshade. This is first year electronics/science.

J_J's response:

Quote:
Interesting claim on your part.

We are discussing Niteshades comments on flattening the response through better regulation, basic electronics 101. From page 1:

Niteshade:

Quote:
Feedback circuits are nothing more than automatic frequency attenuators.

J_J's response:

Quote:
No.

Just no.

Check this J_J comment out from page 1.

Quote:
As to "valid", well, sorry, you're making the extraordinary assertion, perhaps you should look into who has the burden of proof here.

Yes it was a generalized statement. But Niteshade's point is well taken.

So Niteshade has the burden of proof because he utilized mainstream Electronics 101. Yet you, who disagrees with mainstream Electronics do not have to demonstrate proof. Interesting logic to say the least.

And you claimed Niteshade has made an "extraordinary assertion". Again quite interesting logic since Niteshade was basically utilizing mainstream Electronics 101.

Me:

Quote:
Using a lower tap or negative feedback will lower the output impedance of the amp and typically provide better signal voltage regulation vs varying load such as a typical speaker presents. Of course neither negative feedback nor switching taps are perfect.

Quote:
Quite so.

Interesting since you disagreed with Niteshade on page 1, which was quoted above.

J_J

Quote:
But having nothing directly to do with feedback. Quite obviously a "4 ohm tap" will expect a different load impedence and have a lower output impedence than the expected load impedence(which does not have to be 4 ohms, but you build tube amps, you undoubtedly know that). This has to do strictly with the output transformer ratio, though, and is only secondarily related to feedback.

Using the 4 ohm tap vs 8 ohm tap with typical 8 ohm speakers, as well as using negative feedback, both reduce the output impedance (Z) of the component/amplifier and provide better signal regulation. This means the peaks and dips tend to fill in, just as Niteshade and mainstream Electronics 101 claims.

The rest is off topic. I would be careful of how you communicate with designers/manufacturers as this can adversely affect the public's opinion of that manufacturer.

Break it up, guys.

Quote:
The premise behind using the 4 ohm tap was this: Impedance dips cause attenuation. The 4 ohm tap should be close to most dips in a nominally 8 ohm speaker, alleviating the attenuation.

I'll have to pick a nit here, as well, the fact that the 4 ohm tap has smaller output impedence is why you'll see less attenuation, the "closeness" of the dip to 4 ohms isn't the point. What matters is the output impedence (as opposed to the preferred load impedence).

Typically the output impedence of an amplifier is much lower than the intended load impedence, although of course with tube amps this doesn't have to be true, and yes, I agree, the easiest way to do that is to use a decent feedback circuit, be it triodes with internal feedback (Yes, triodes have feedback, it's just built into the tube.) or whatever.

This is indeed a place where tubes are a heck of a lot more convenient to design with than transistors, too, in many ways.

Sammet, any comment on the particular circuit I mentioned above? I've found it quite interesting in its effects.

Quote:
I want to make sure you typed this right: Did you mean ground the 8 ohm tap and use "0" & "16" for the speaker? Grounding 4 would leave a disproportionate amount of windings on either side of it. Eight would be between 0 and 16.

Quote: "I wonder if either of you has ever taken an amplifier with a 4, 8, 16 ohm winding, grounded the 4 ohm tap, and used the "0" and 16 ohm (those having the same winding ratio to the primary) taps as cathode taps."

Hi Blair,

Grounding the 4 ohm tap leaves two feedback taps, one at 0 and the other at 16 ohm tap. This can be used to create positive and negative feedback loops to different cathodes to reduce the output impedance to very very low levels.

That is why he stated this.

Quote:
I wonder if either of you has ever taken an amplifier with a 4, 8, 16 ohm winding, grounded the 4 ohm tap, and used the "0" and 16 ohm (those having the same winding ratio to the primary) taps as cathode taps.

This, if you have the gain structure right, turns the whole amplifier into a cathode follower,...

I would not bother answering his question Blair as this is privileged information concerning my designs, and maybe yours. It could be another attempt to embarrass you while boasting himself, as he attempted to do on page one. Does he have something against you Blair?

Quote:
I want to make sure you typed this right: Did you mean ground the 8 ohm tap and use "0" & "16" for the speaker? Grounding 4 would leave a disproportionate amount of windings on either side of it. Eight would be between 0 and 16.

Um, the turns ratio is squared to give you the impedence so in fact I meant 4 and 16. I've made this mistake before, though, too.

Quote:
It could be another attempt to embarrass you while boasting himself, as he attempted to do on page one. Does he have something against you Blair?

so instead of answering JJ's harmless question, keeping the thread on-course, you insult and defame. Way to continue with this confrontational posting style, Sammett..

Quote:
I would not bother answering his question Blair as this is privileged information concerning my designs, and maybe yours.

I see that sasaudio is in fact unapproachable, unreasonable.

The approach is one I've used. I am asking if others have used it.

It does reduce both distortion and output impedence a lot, BUT it also makes overload performance more like "hard clipping". It's a case of "win some, lose some".

One reason I asked is that it helps eliminate transfomer distortions at low frequencies.

Would anyone care to answer? I don't want any kind of "privileged" information, please, only what you'd say to any other person in the world.

Quote:
so instead of answering JJ's harmless question, keeping the thread on-course, you insult and defame. Way to continue with this confrontational posting style, Sammett..

NC: Leave Sammet alone, he's unapproachable and unreasonable. Not only that, he would prefer to spread his unwarranted, defamatory hostility to others.

Mr. Nightshade guy: My only real gripe was with your serious oversimplification of feedback. Perhaps you forgot the n^2 impedence thing, we all do that once in a while (I dare say ).

If this is in fact a proprietary thing in your amp, no, don't answer it. I'm NOT trying to pry out proprietary information.

Mr. Mejas, I shall now regard Sammet as unapproachable and unreasonable. Hence, barring his continued, unprofessional and unethical defamation, I shall not bother to attempt to engage him in discussions.

Quote:
Like I said earlier, I haven't tried it, but may give it a shot.

Yes, all designs have pro's & con's. It's just a matter of finding which compromises you'd rather live with.

On that we can very much agree.

Quote:
so instead of answering JJ's harmless question, keeping the thread on-course, you insult and defame. Way to continue with this confrontational posting style, Sammett..

If you check page 1, the subject was not grounding the 4 ohm tap and using 0 and 16 ohm tap. Niteshades comment was quite different and to which I adhered to, until you and J_J diverted the subject.

Want to let it die now NC.

Quote:
NC: Leave Sammet alone, he's unapproachable and unreasonable. Not only that, he would prefer to spread his unwarranted, defamatory hostility to others.

Have you got some vendetta against individuals and manufacturers? This has happened before to others. First you give Niteshade a completely wrong answer, then after I respond and agree with Niteshade's assessment on page 2, your buddy NCdrawal attacks me personally.

Quote:
You have some nerve, Sammett... getting into a pissing match with someone whose chief vocation is audio processing! Christ almighty. you are like a dog that keeps running into parked cars...you keep getting knocked out, jaw broken, but always come back for more abuse.

And you call me unreasonable and spreading hostility. You use some strange logic.

Niteshade posted and you posted:

Quote:
No.

Just no.

And this:

Quote:
you're making the extraordinary assertion, perhaps you should look into who has the burden of proof here.

You know how negative feedback (increases damping factor) and regulators work (Electronics 101) and yet you responded with exactly the opposite and wrong answer (see page 1). That is why I responded to Niteshade. On top of that you claim Niteshade made an "extraordinary assertion" and demand proof from him. Sorry but Niteshade has electronics 101, first year electronics, on his side.

After all of this you then attempt to divert the subject by bringing up some obscure design to check if we know of such. What are we suppose to think but another attempt to defame Niteshade again, and me. Afterall, you just attempted it on page 1 and your buddy NCdrawal attempted to intimidate me on page 2.

Let's see it again.

Quote:
You have some nerve, Sammett... getting into a pissing match with someone whose chief vocation is audio processing! Christ almighty. you are like a dog that keeps running into parked cars...you keep getting knocked out, jaw broken, but always come back for more abuse.

So please forgive me for not trusting either you or NCdrawal's motives.

Cheers.

Umm, see post #2 following your OP. I've tried that for a while... it's definitely not a cure-all but I know what you're talking about.

I think all relatively low wattage tube amps should have 16 ohm taps; not having it would be too limiting. A lot of tube amp owners complain about not having enough speakers made for them, but there's a treasure trove of vintage full rangers or field coils out there for which you can utilize those 16 ohm taps. And many of them sound as good or better than a lot of what's floating out there...

Very good everyone for not posting the "source must match the load for max power transfer" argument.
If you really want to minimize load impeadance effects on frequency response just reduce source impedance to a few tenths of an ohm. And while you're at it remove any reactive components in the signal path (like transformers and capacitors).

Happy Thanksgiving as well to everyone.

Cheers

We need an arrogant prick forum. Maybe Stephen can create one just for you.