Bob, your questions are pointless. IF you want to continue to use coils, I sincerely hope that they are kept small in value like 0.5uH and a small value resistor is used in parallel, such as 1 ohm or less. You already do this, but you ignore the others who use several to many uH and 5 ohms or more. It is the higher values that concern me most.
Graham Maynard has given most probably the most insightful input here as to what the coil seems to do to the sound. However, he is ignored and harassed, just like I am, when I don't play along.
I learned about the audibility of coils when an insightful friend of mine (now deceased) made an A-B demonstration about 20 years ago. I was just as argumentative about it as many here are, but my late friend had the patience to show me anyway.
Years later, I found, or actually I was shown by the Taiwanese engineers that work with Parasound, that we did not need output coils in order to make a stable amplifier. At least, an amplifier stable with any value of capacitive load. This was not true for many power amps throughout history, and my first experience was with a Marantz 14, a solid state mono amp that I purchased for Ampex Research in 1969. It DID oscillate with a particular cap load, but the identical Marantz 14 that I personally owned DID NOT oscillate with a particular cap load, even with an output coil. Go figure!
Along the way I found that Julian Vereker (Naim) an old friend and now deceased, did not use an output coil, and Nelson Pass (an esteemed competitor) was talking about not using one, and Charles Hansen ( a good friend and esteemed competitor) was not happy about coils either, I felt myself in good company.
Other companies, using the basic input topology, that I developed in 1968 such as Bryston, still use output coils, but they appear to be smaller than what the nominal value on the published schematic. Go figure! And of course, Halcro, who appear to copy much of what Bob Cordell published more that 20 years ago, but with greater flourish, stayed with a 0.5uH coil, which is what Bob thinks is the be all and end all to the stable universe of amplifiers. Perhaps he is right! You folks will have to find out for yourselves.
Graham Maynard has given most probably the most insightful input here as to what the coil seems to do to the sound. However, he is ignored and harassed, just like I am, when I don't play along.
I learned about the audibility of coils when an insightful friend of mine (now deceased) made an A-B demonstration about 20 years ago. I was just as argumentative about it as many here are, but my late friend had the patience to show me anyway.
Years later, I found, or actually I was shown by the Taiwanese engineers that work with Parasound, that we did not need output coils in order to make a stable amplifier. At least, an amplifier stable with any value of capacitive load. This was not true for many power amps throughout history, and my first experience was with a Marantz 14, a solid state mono amp that I purchased for Ampex Research in 1969. It DID oscillate with a particular cap load, but the identical Marantz 14 that I personally owned DID NOT oscillate with a particular cap load, even with an output coil. Go figure!
Along the way I found that Julian Vereker (Naim) an old friend and now deceased, did not use an output coil, and Nelson Pass (an esteemed competitor) was talking about not using one, and Charles Hansen ( a good friend and esteemed competitor) was not happy about coils either, I felt myself in good company.
Other companies, using the basic input topology, that I developed in 1968 such as Bryston, still use output coils, but they appear to be smaller than what the nominal value on the published schematic. Go figure! And of course, Halcro, who appear to copy much of what Bob Cordell published more that 20 years ago, but with greater flourish, stayed with a 0.5uH coil, which is what Bob thinks is the be all and end all to the stable universe of amplifiers. Perhaps he is right! You folks will have to find out for yourselves.
john curl said:
Hi John,
Thanks for your respone. Now we have something we can discuss.
First, my questions were not pointless. Indeed, they were right to the point of whether coils cause audible degradation. Moreover, the answers to those questions would lend some insight into why they are audible, and whether the broad generalization you made is appropriate. I just don't take things at face value. I've repeated several times that I hold out the possibility that coils are audible, and would like to understand better the circumstances. If you answered that you were convinced that coils were audible in one test with an amplifier that used a 5 uH coil wound around an aluminum electrolytic, I would completely dismiss your generalization. I'm sure that's an extreme, but degree of credibility is what I'm getting at. So far, you've left me unconvinced.
In the audio business, old wives' tales get started and circulate and end up having a life of their own, completely undeserved. I'm not saying that is what has happened here, but I'll give it 50% odds that the generalization grew beyond reasonable breadth. Kind of like the bad rap that NFB got because some imbeciles used it wrongly. Nobody should use table saws because some idiot cut his finger off.
Because I hold out the possibility that coils may be audible, I do like to keep them small and very well-damped and carefully constructed. In some cases I have used an air core toroid. But I would never say that someone who chooses to incorporate a 2 uH-2 ohm combination in their amplifier is an inferior designer, as you have so often implied.
The later point that you made about the need for the coil is more interesting, even if not directly related to audibility of the coil. That is, whether it is needed and under what conditions it may be dispensed with. Maybe when one uses fast enough output transistors, and a large number of them in parallel, the need for the coil goes away, at least for a reasonable class of loads. If so, that is great, and it may be an area where we need to pay more attention, rather than just using a coil as a knee-jerk reaction.
Along those lines, maybe we could level the playing field by agreeing on a reasonable set of load tests that establishes a level of confidence in stability that we can all be comfortable with. This certainly might not be a 2 uF load. As a start, however, I would suggest that any amplifier should be stable with 1000 pf, or 0.01 uF or 0.1 uF right at its output terminals with or without an 8 ohm resistive load there as well. That is just six simple tests. Maybe the coil issue is mainly a matter of different degrees of conservatism by each of us.
What values of output capacitance have you checked the JC-1 with?
I'll try to go back and understand what Graham has been saying and comment on it. I sometimes have difficulty separating out the linear and nonlinear issues in some of those discussions. Or perhaps you could paraphrase for us what you think the main point there is.
Cheers,
Bob
PMA said:
mr.Curl (with all respect) as you see , "won`t?" give a STRAIGHT answer to an STRAIGHT question. As i`ve seen on many threads before, this is his politics: "I know how, but i won`t tell you". Just teasing and not giving answers is not very nice mr.Curl...
CHEERS FROM ALL DIYERS FOR HOLY NELSON PASS!
When the lumped indictance from the speaker cable and cross over can be reduced to close the 2-5uH of the old school output inductance designs, then I'll accept that there may be a case for removing the output coil.
The point of this whole thing is where do you stop? If a 10% change in the total output inductance (and I thinks its probably more like 1-2% at 2uH) has such an audible effect, then we need to go with bi/tri amping and integrate the amplifiers into the speaker boxes -ditch the x-overs etc
The point of this whole thing is where do you stop? If a 10% change in the total output inductance (and I thinks its probably more like 1-2% at 2uH) has such an audible effect, then we need to go with bi/tri amping and integrate the amplifiers into the speaker boxes -ditch the x-overs etc
Bonsai said:
When you have done all that is possible.
Then you lie awake at night, staring at the ceiling, trying to think of something new to try.
John,
As an individual you are more combative than I am at this stage of my life. I understand why you're fighting the good fight, but there comes a time when a rational man acknowledges that he cannot change minds that are already made up.
Their claims to the contrary, there is no argument you can present that will cause your critics to say, "Oh...I see...you're right."
They do not listen and since that is the impetus for these things, anything you say will be dismissed out of hand. They will keep asking for reasons after you have already given them the reason.
You don't have to do anything I say. We've never met and likely never will. I'm just one member among many on a website filled with enough people to fill a decent-sized city. That said, and with all due respect, I'd suggest saving your energy for fights that you can win and people who will listen...
In more than one sense of the word.
Grey
Bonsai said:
Thanks. Lissajous figures are OK, but I prefer to use the Variable Sweep control on my scope so that one cycle of the reference signal takes up the entire 10 horizontal divisions. I can then read the phase difference indexed in 7.2 degree increments (Or 3.6 degrees with 1/2 a cycle displayed).
Phase meters are easy to make. Here is one:
http://www.intersil.com/data/an/an9637.pdf
Working well out to 10MHz makes it a bit flash for audio, but by using the same toploogy, a lower performance clone could be made with common components and audio opamps.
Cheers,
Glen
G.Kleinschmidt said:
I'd love to discuss this topic, but I'm not going to... here (paraphrasing Bill Murray). You've stated your take on my comments, obviously there’s not even a slim possibility that there might be something behind them. From past experiences any discussion will go nowhere.
Besides I have zero credibility here, not an internet pro; to easily confundled and laid bare. Thankfully that can’t negate actual accomplishments.
For my presence here I’ve taken to adding comments when I feel a bit of food for thought might be appreciated by readers that might not necessarily buy into the fact that we now know everything there is to know about amplifier design. Kind of a 60's thing.
Simply put, the finer details of layout and grounding are the key to stability in ANY amplifier topology. This discussion keeps touching on the NEED for coils and complex compensation and trial and error to maintain amplifier stability. Look at your off-hand comment that the input filter is not functioning properly, why? Maybe the feedback isn't performing as per simulation, either... Whatever.
There is another dimension that isn’t clearly covered in any text I’ve been able to uncover, and I’m constantly looking. Therein lies my impetus to comment. We all bow to whatever God we chose.
Also I'll point out I did actually insert coils into my system and report back on what I heard. Bob needs to hear it from a higher authority, in DOE format. To bad John's not playing. I guarantee my system is badder than yours!(???)
Ya want to learn something, shut down the simulations and build something basic, without Super-amp garnishments, leave out all the compensation and try to make it stable without Band-Aids. You might learn a thing or two about what I’m talking about.
From the sidelines, Mike.
Mike, in thinking about this situation, it reminds of a difficult situation that I was in 35 years ago, almost to the day.
I was working as a sound consultant on a rock film called 'The Last Days of the Fillmore' for about 1 year and was into the final mix. We had all kinds of fun at Goldwin Studios taking the sound department apart and putting it back together. I drove everybody nuts as usual. ;-)
Finally, it was time to do the FINAL transfer at 20th Century Fox. They greeted me with about the same enthusiasm as they often do around here, but still I was there to make things work.
Well, I found that the transfer lost about 5 extra dB of S/N for no apparent reason. I finally traced it to 2 volume pots in series, and they were set wrong. I asked them to please turn 1 pot up, and the other pot down, and guess what? They wouldn't do it out of spite, and laughed in my face about it.
I realized then that you can't 'help' people if they do not wish it, and that often they would deliberately compromise a project, just to show their spite. I promised myself at the time, that just being 'right' was not enough, and that I should use more diplomacy, just to get people to do what is right and best for the project. I have temporarily forgotten this lesson here on the subject of coils, and I wish everyone the best, as far as they can go with their individual designs. Best to leave it at that.
I was working as a sound consultant on a rock film called 'The Last Days of the Fillmore' for about 1 year and was into the final mix. We had all kinds of fun at Goldwin Studios taking the sound department apart and putting it back together. I drove everybody nuts as usual. ;-)
Finally, it was time to do the FINAL transfer at 20th Century Fox. They greeted me with about the same enthusiasm as they often do around here, but still I was there to make things work.
Well, I found that the transfer lost about 5 extra dB of S/N for no apparent reason. I finally traced it to 2 volume pots in series, and they were set wrong. I asked them to please turn 1 pot up, and the other pot down, and guess what? They wouldn't do it out of spite, and laughed in my face about it.
I realized then that you can't 'help' people if they do not wish it, and that often they would deliberately compromise a project, just to show their spite. I promised myself at the time, that just being 'right' was not enough, and that I should use more diplomacy, just to get people to do what is right and best for the project. I have temporarily forgotten this lesson here on the subject of coils, and I wish everyone the best, as far as they can go with their individual designs. Best to leave it at that.
Well that certainly was an interesting response.MikeBettinger said:
"For my presence here I’ve taken to adding comments when I feel a bit of food for thought might be appreciated by readers that might not necessarily buy into the fact that we now know everything there is to know about amplifier design. Kind of a 60's thing."
Well I’ve never claimed that we know everything about amplifier design. I certainly don’t. I have no idea why you are taking issue with this in reply to me.
"Simply put, the finer details of layout and grounding are the key to stability in ANY amplifier topology."
It is not the ‘”key” – it is only one important part of it.
"This discussion keeps touching on the NEED for coils and complex compensation and trial and error to maintain amplifier stability."
So?
"Look at your off-hand comment that the input filter is not functioning properly, why? Maybe the feedback isn't performing as per simulation, either... Whatever."
I did not make an off-hand comment that the input filter is not working properly. So long as the input is properly terminated, the input filter works fine. This goes for virtually any audio amplifier. It applies to the audio stage of my TV, my 20 year old Rotell, my restored valve amplifiers, etc. I have simulated none of these, so I don’t know how your “performing as per simulation” comment applies.
"There is another dimension that isn’t clearly covered in any text I’ve been able to uncover, and I’m constantly looking. Therein lies my impetus to comment. We all bow to whatever God we chose."
Well I don’t know what point, exactly, you're trying to make here.
"Also I'll point out I did actually insert coils into my system and report back on what I heard. Bob needs to hear it from a higher authority, in DOE format. To bad John's not playing. I guarantee my system is badder than yours!(???)"
So you've got a gripe with Bob too......
"Ya want to learn something, shut down the simulations and build something basic, without Super-amp garnishments, leave out all the compensation and try to make it stable without Band-Aids. You might learn a thing or two about what I’m talking about. "
Now you are just being presumptuous. Except for the little I have revealed here, you have no idea of how many or what types of amplifiers I have built, or to what degree I have simulated them. I’ve simulated only a few, and I’ve built amps for various applications that are far less than “super”. Also, I'd really like for you to explain to everyone here how a decent solid-state amplifier can be made stable with all the compensation components omitted. That would be a real eye opener. Pray tell.
janneman said:
Hi Jan,
You've got me thinking there might be a simpler way of looking at this. The speaker response becomes dominated by cone mass (F=ma) above resonance and electrically looks like a large value capacitor. I don't recall the exact value but a figure in the region of about 10uF sounds right. There is still the speaker coil resistance and effective inductance between this capacitor and the output coil but maybe the interaction between the output coil and this capacitor is what is causing the problem.
Also, you mentioned the rise time of a 20kHz wave. The first quarter cycle of this wave will be 12.5uS duration so lets say 11us rise time. The rise time of a 10uF cap + 10uH L (by way of example) would be about 10us - depending on damping.
I'll post the TS speaker model when I get home.
Cheers,
Tony
tomtt said:
Hi Tom,
I have no idea what your referencing here, but as to the coil, a few things come to mind:
1 - Linear distortion. This could be caused by the rising impedance of an inductance in series with the load. The result would be frequency response deviations. As far as I remember the many posts, this would probably be at max a fractional dB at the frequency extreme;
2 - Non-linear distortion. Several causes could be identified: thermal effects, meaning the coil heats up and cools down with signal level causing harmonic distortions, or saturation effects meaning that the inductance varies with signal level. I haven't done the maths but my gut feeling says that with a hefty, air core coil, these effects would be swamped by similar effects in the speaker voice coils.
Jan Didden
GRollins said:
Grey,
I can assure you that many here, me including, share your frustration. I really AM ready to be convinced that I don't see it correctly. Many times at this forum I have told people, gee, you're right, I'm wrong. Unfortunately, here, no arguments are provided, unless you count "because I say so" as an argument. That's MY frustration.
Jan Didden
janneman said:
G'day Jan.
Speaking of frustration, I’ve just the whole content of a post ignored, and previous postings on amplifier stability, compensation and measurement blithely dismissed by someone who tells me that all we need to pay attention to is grounding and layout, thus allowing us to leave out any coil and “all the compensation.”
Oh, and if your amplifier acts funny with the input deliberately unterminated (signal ground floating too) and an abnormally high amount of RF injected, then your whole amplifiers grounding scheme is obviously incompetently implemented.
Serious question - Is it just me who is exceedingly inept, or do other designers / DIY’ers here have a great deal of difficulty making SS amplifiers with a decent amout of NFB stable without any form of frequency compensation and unconditionally immune to deliberately injected RF interference?
Cheers,
Glen
Hi Graham, are you still here ??
About the "first cylce distortion"simulation setup you use, is it this one ?
http://www.diyaudio.com/forums/attachment.php?s=&postid=1206545&stamp=1178903803&
Cheers
Hartono
About the "first cylce distortion"simulation setup you use, is it this one ?
http://www.diyaudio.com/forums/attachment.php?s=&postid=1206545&stamp=1178903803&
Cheers
Hartono
- Status
- Not open for further replies.