Audibility of output coils

Status
Not open for further replies.
Jump to Latest
#63
Workhorse said:

You all guys consider the output coil as a sonic degrader, well i agree with you all because the application environment of your amps is not in a Hell as pro-arena, rather than its the "Home Sweet Home" or a demonstration room or maybe your living room
Click to expand...



Well I certainly don’t. And my opinion stems from listening experience as well as theory.

The easily measured effect of an output inductor is a slight amplitude ripple in the frequency response when driving a real life, complex reactive load, as measured at the amplifiers speaker output terminal. The ripple typically increases with frequency, as the impedance of the inductor increases (but depends on the nature of the load).
However, the amplitude of the ripple (dB) is typically very small in a well designed amplifier with the output inductor appropiately proportioned and incorporated into a suitable output stabilisation network. There is very little evidence that such small amplitude variations are distractingly audible, least of all that if they were audible, that they’d actually detract from the reproduced sound quality. Amplitude variations in the frequency response caused by the acoustic behaviour of the loudspeakers and the listening environment are a great deal larger and more important.

To vaguely put the value of the output inductor into some kind of perspective here, consider what’s inside a typical speaker.
A standard 6db/octave low pass filter for an 8R woofer with a corner frequency of 400Hz will consist of a series inductor in the vicinity of 3.3mH. For a 12db/octave filter with the same corner frequency, it will be in the vicinity of 4.5mH – 2250 times greater than a 2uH output inductor.

At 1kHz, a 2uH inductor has an inductive reactance of 0.0125 ohms – a totally insignificant figure. Even at 20kHz, a 2uH inductor still only has an inductive reactance of 0.25 ohms.


Cheers,
Glen
 
#66
Historical artefact:
a couple of commercial power amps of the past had the option of installing a wire bridge on the output board to bypass the output coil. Doesn't GamuT still advocate this ?

Though subjective as usual, i've heard the effect of an output coil on a few commercial amplifiers connected to ESLs.
The user's option of keeping the output coil on-line or out seems practical to me, in contrast to the Orwellian attitude of Halcro.

I'm surprised Kanwar doesn't use any trick in the booklet, considering the electrical powerplant business he's in.
On the other hand, if Mr Curl can vouch for a safe operation of his amps without a coil on any loudspeaker, you'd have to respect that too.
Output coil and Boucherot/Zobel anyone ?

"Respectfull" is the true feature of mature intelligence and knowledge.
 
#67
G.Kleinschmidt said:
[snip]The easily measured effect of an output inductor is a slight amplitude ripple in the frequency response when driving a real life, complex reactive load, as measured at the amplifiers speaker output terminal. [snip]Cheers, Glen
Click to expand...


Glen,

I would submit that that freq response ripple is much higher when measured at the speaker terminals itself, after the speaker cables. In fact, even without the inductor, you still will have significant ripple at the speaker.

In other words, the presence or not of the coil will alter the ripple at the amp speaker terminals, but not at the speaker itself.

Therefore, I find it very hard to accept at face value any statements on audible effects of that coil.

Jan Didden
 
#68
janneman said:



Glen,

I would submit that that freq response ripple is much higher when measured at the speaker terminals itself, after the speaker cables. In fact, even without the inductor, you still will have significant ripple at the speaker.

In other words, the presence or not of the coil will alter the ripple at the amp speaker terminals, but not at the speaker itself.

Therefore, I find it very hard to accept at face value any statements on audible effects of that coil.

Jan Didden
Click to expand...


G’day Jan

I agree completely that the frequency ripple will be larger after the speaker cables, even without the inductor, but I wouldn’t go quite as far as saying that the output inductor only increases the ripple at the amplifier terminals. The ripple ultimately produced at the speaker terminals is the summation of the ripple induced across the speaker cable in series with the output inductor (assuming the amplifier has practically zero output impedance before the inductor).

On an interesting side note, there seems to me to be a few inductor-decrying designers out there who also like very low or zero global negative feedback (which is the significant design trait that actually allows their designs to operate stably without a load-isolating inductor).
Such amplifiers typically have an abysmally high output impedance, and frequency ripple into reactive loads like nuts – way worse than any decent feedback amp with a little inductance on the output :spin:

Cheers,
Glen
 
#69
It is possible to have low Zout, high bandwidth designs which use only interstage feedback, notably between output devices and the driver.

This is not strictly global nfb, nor local, some sort of hybrid, but it can operate extremely well. A CFP with bipolar driver and mosfet output can offer around 80 milliohms Zout across the audio range, but of course tends to be best in Class A. Such a design does not need an output inductor.

Hugh
 
#70
AKSA said:
It is possible to have low Zout, high bandwidth designs which use only interstage feedback, notably between output devices and the driver.

This is not strictly global nfb, nor local, some sort of hybrid, but it can operate extremely well. A CFP with bipolar driver and mosfet output can offer around 80 milliohms Zout across the audio range, but of course tends to be best in Class A. Such a design does not need an output inductor.

Hugh
Click to expand...

Hi Hugh,

I would say that such a design, effectively wrapping nfb around driver/output stage, classifies as a low feedback design from the viewpoint of the output stage, so to say, and therefore can get away without an inductor. But it also has a highish Zout and therefore much more ripple at the speaker than necessary.

Of course, the sound coloration generated by such freq response ripple may well be attractive to some listeners, which *might* erroneously lead to the conclusion that no inductor sounds better...

Jan Didden
 
#71
Even if Inductor reactance at 20khz is not audible, the effect is still measurable. For design that strive to bring the output impedance of the amp as close to 0 ohm at audio frequency the inductor reactance might be a set back.

consider active bi-amping situation for 4 ohm tweeter :

consider 4 ohm tweeter, while the impedance plot at 20 Khz might look large because of the inductance of the voice coil, the series resistance part is modulated by thermal heating, this with the series output resistance of the amp may led to less than desirable end result,so depending on the specific design goal it might be good if the actual coil reactance is not in series with the amplifier output.

Another issue is, in many case the transient capability of any active circuit element cannot be fully determined by looking at the frequency response alone, especially if the frequency response is very limited. (amp with 20Khz response will not be as good at handling fast rising 20Khz square wave compared to 2Mhz amp even with purely resistive load).
Inductance in series with amplifier output is definitely frequency limiting the voltage amplifier output.

It might be a good idea to start a new thread on how to remove output inductor while maintaining acceptable or even unconditional level of stability, rather than arguing about the inductor necessity. This way, time and brainpower is directed at removing the bulky component and hopefully shed new light and produce progress.
 
#72
It is clear, by all things people said, the need output coils

If a switched one was found will be even better, as there are others applications to professional units too.

The professional amplifier can be used in a home environment, small length of wires and not using passive crossovers.

No doubts coils are needed in the output....but not always.

Try an A to B comparison.... make the test using it and removing it...try this for a long time if not perceive immediatelly...differences exists.

A relay, of course, beeing near the circuit can switch and bypass the coil....not difficult and nor hard to do it by yourself, instead of discussing that.

Well...people has a lot of courage to discuss with John Curl...maybe people do not know who is John Curl....of course he is rigth to a professional application...to home use, some people that do not use coils are rigth too.

But he cannot make dangerous amplifiers, as he use to make for professional use.

regards,

Carlos
 
#75
Curl's stirring up the pot again!

Dunno about Curl's opinion, but he's LATE to my party!
I've been saying this for about 17 years now - John 😀

...Hey HEY, he he, get offa my cloud - Don't hang around cause two's a crowd, on my cloud... Micky Jag/Brian Jones

There's not any doubt in my mind that the output coil does effect the sound of an amp that is already pretty darn good to start with. That and the speakers that ur listening with, not to mention what is left of your so-called hearing this stuff has to work... problem is that a very high percentage of folks who are in the audio thing simply don't "get" many of the so-called "audibility" issues because they A) haven't heard it yet or B) can't (physically that is) hear it.

Ok, ya heard the message now - either ur deaf or ur gear sux... 😉

If you can not hear electrolytics, then stop right there, this discussion is meaningless for you.

Curl's probably deaf anyhow, don't trust him... :bigeyes: :hot: :clown: :whazzat: :Popworm: :whacko: :spin:


_-_-bear
 
#76
forr,

Yes I've read that article a few weeks ago , I already write something about it but I canceled to post, I don't want to ruin the theme of this new thread which is about output coil audibility.

I feel the article interpret the data in the wrong sense.
I have my own interpretation of the data , how about you ?

no I haven't measured voice coil resistance at 2 Mhz. Do you find anything unusual ? I only guess it might be quite low because of voice coil interwinding capacitance. I'm interested in your finding, or maybe u read it somewhere ? anything about audio is interesting,PM me
😀
 
#79
8 pages, almost 80 post, number of people actually doing A-B testing (preferably blind testing) to listen for sonic difference of coil = 0.

I use to think so many things could be ignored because it affects so little or can't have an effect. I don't think that way now. I think more is audioable or perceiveable than we are taught in school or Mr. Big PhD thinks. If you are up for it, try different coils in value, design, etc. and see what differences, if any, you may be able to hear.

-SL
 
#80
Hi John Curl,

Welcome to the very mixed world of diyAudio.


Hi Glen.

So you calculate 0.0125 ohms @ 1kHz for 2uH.
Let's take that to 10kHz where a supertweeter might crossover.
That becomes 0.125 ohms.
Take 1amp of current and you have 125mV. ?? % distortion ??

The point being that this error is generated by loudspeaker system current which has been reactively modified, and this voltage appears at the amplifier's output terminal in series with, yet is not directly related to, NFB node determined amplifier voltage output.

Now this voltage could be generated by mid-bass low pass crossover circuitry, and thus audibly distort the high-pass tweeter drive.

No choke - no error.

( John your hearing must be as bad as mine has been suggested to be. )

Bear - spot on !!!
( I go for not heard because of the equipment being used to listen for the problem.
.......... So what must the resulting audio sound like ??? )

Cheers ........ Graham I
 
Status
Not open for further replies.
Top Bottom