SPLIF Breakthru!
OK! I got it now. Using a SPLIF like topology, one output using conventional global feedback connects to voice coil #1. The second output stage, running open loop class A, and with a slightly greater voltage gain (maybe a Sziklai-CFP pair output to get slightly over unity gain), connects to voice coil #2. An adjustable gain pot. for the second output is adjusted until the output stage # 1 supplies a minimum of current. Output #2 is essentially driving the speaker and making up for the ohmic losses in the voice coil as well. Output #1 just provides any correction currents required to maintain its low distortion level.
Obviously, a somewhat half a--ed version of the above V-I composite/ superconducting speaker idea, but it should work. Maybe should put a little resistance in series with output #2 to increase its impedance so the two SPLIFed outputs don't fight each other. This looks a somewhat like the current dumping idea now. Just doing the output combination in the dual voice coil speaker instead.
Hmmm, I guess one could even dispense with the dual voice coil speaker and just connect the two outputs together, but one would lose the velocity servo effect ( and cancellation of voice coil ohmic effect) then as far as output #1 sees the speaker load. Just looks a lot like the current dumping design then.
One final thought: using Sziklai-CFP outputs on both output stages in the original SPLIF topology should help the two outputs match better since the local feedback improves linearity of each. Maybe would help to match up crossover effects better.
Don B.
OK! I got it now. Using a SPLIF like topology, one output using conventional global feedback connects to voice coil #1. The second output stage, running open loop class A, and with a slightly greater voltage gain (maybe a Sziklai-CFP pair output to get slightly over unity gain), connects to voice coil #2. An adjustable gain pot. for the second output is adjusted until the output stage # 1 supplies a minimum of current. Output #2 is essentially driving the speaker and making up for the ohmic losses in the voice coil as well. Output #1 just provides any correction currents required to maintain its low distortion level.
Obviously, a somewhat half a--ed version of the above V-I composite/ superconducting speaker idea, but it should work. Maybe should put a little resistance in series with output #2 to increase its impedance so the two SPLIFed outputs don't fight each other. This looks a somewhat like the current dumping idea now. Just doing the output combination in the dual voice coil speaker instead.
Hmmm, I guess one could even dispense with the dual voice coil speaker and just connect the two outputs together, but one would lose the velocity servo effect ( and cancellation of voice coil ohmic effect) then as far as output #1 sees the speaker load. Just looks a lot like the current dumping design then.
One final thought: using Sziklai-CFP outputs on both output stages in the original SPLIF topology should help the two outputs match better since the local feedback improves linearity of each. Maybe would help to match up crossover effects better.
Don B.
Re: Re: Re: Charles Altman's splif topology
How many mV should a very badly designed mike like a backwards operating speaker do you think will be dumped back by reflected sound?
It's allways a good idea to meat up sound reasoning with some numbers ....
Rodolfo
darkfenriz said:
How many mV should a very badly designed mike like a backwards operating speaker do you think will be dumped back by reflected sound?
It's allways a good idea to meat up sound reasoning with some numbers ....
Rodolfo
Re: Re: Charles Altman's splif topology
I think it is the cone excursion that would be proportional to output voltage, at least for an ideal speaker.
As for the cone velocity... would'nt this be an accelerated movement depending on moving mass and magnetic force, so... strongly determined by inertia ?
Would'nt that mean that cone acceleration is proportional to voltage rise, but largely determined by mass and magnet force, with practically the highest attainable velocity produced by a Lowther driver (light cone, strong magnet, rel. high excursion, and soft suspension) ?
Charles 🙂
hummhoom said:
I think it is the cone excursion that would be proportional to output voltage, at least for an ideal speaker.
As for the cone velocity... would'nt this be an accelerated movement depending on moving mass and magnetic force, so... strongly determined by inertia ?
Would'nt that mean that cone acceleration is proportional to voltage rise, but largely determined by mass and magnet force, with practically the highest attainable velocity produced by a Lowther driver (light cone, strong magnet, rel. high excursion, and soft suspension) ?
Charles 🙂
Back to Physics 101
The EMF voltage induced on the voice coil is caused by Volts=N*dPhi/dt where Phi is the magnetic flux, N the number of turns. Velocity sets the rate at which magnetic flux lines are crossed. So EMF voltage should be proportional to velocity, just like for a DC motor/generator analogy as Hummhoon suggested. For a given applied voltage, the voice coil motor will come up to speed until its back EMF is equal to the applied voltage. The force to accelerate up to this speed being caused by current flow * magnetic field strength, which ideally would drop to zero as the terminal velocity is reached. Additional complications such as coil Ohmic resistance and acoustic radiation power appear as a resistance across the terminals.The springy suspension system appears as an inductance. The cone and voice coil mass appear as a capacitance, drawing large currents to get them up to speed.
Don
The EMF voltage induced on the voice coil is caused by Volts=N*dPhi/dt where Phi is the magnetic flux, N the number of turns. Velocity sets the rate at which magnetic flux lines are crossed. So EMF voltage should be proportional to velocity, just like for a DC motor/generator analogy as Hummhoon suggested. For a given applied voltage, the voice coil motor will come up to speed until its back EMF is equal to the applied voltage. The force to accelerate up to this speed being caused by current flow * magnetic field strength, which ideally would drop to zero as the terminal velocity is reached. Additional complications such as coil Ohmic resistance and acoustic radiation power appear as a resistance across the terminals.The springy suspension system appears as an inductance. The cone and voice coil mass appear as a capacitance, drawing large currents to get them up to speed.
Don
need new smilie icon!
Suggestion to moderator/management.
We need a waving baton with shiny little stars coming off the end.
I need to use this icon more often! 😀 😀
Or, how about a smilie with a bishops hat too, waving the baton. Like the Pope giving his blessing!
Suggestion to moderator/management.
We need a waving baton with shiny little stars coming off the end.
I need to use this icon more often! 😀 😀
Or, how about a smilie with a bishops hat too, waving the baton. Like the Pope giving his blessing!
AKSA said:
Hugh,
Come on! What would building & listening accomplish? People on this forum built and listen to all kinds of amps, 300B SE, transformer coupled ss (Hi Susan), PP with tubes, susy etc. They all report that they sound great, good separation, excellent imaging, fantastic detail. I have seen the same report from somebody who built a splif. If somebody would built a splif without the dummy stage it would get similarly reports, you know how it is.
Now, thinking about the splif uncovered some interesting things. A very important contribution came from Steven (hi Steven), who noted that the current and voltage in the dummy stage are out of phase with those in the speaker load. Therefore, including the dummy stage in the global feedback loop does nothing for the xover distortion in the real stage. Now, I am still not sure why that dummy stage is there and why it should be the same as the real stage. Also, I dont know what the dummy load signifies. Charles has proved to be somewhat, shall we say, flexible, on this.
Anyway, it seems clear that the splif concept is not very valuable, although looking initially at the diagram it has a certain appeal to most people, including me. In this regard, you may wish to review my sig -😉.
So, by just thinking and talking to other people I learned a great deal about splif, what building and listening would not have given me at all.
Jan Didden
Re: Re: Re: Re: Charles Altman's splif topology
I think i don;t understand your question.
If you mean voltage caused by refllected acoustic wave, multiply it by open loop gain of the amp.
ingrast said:
I think i don;t understand your question.
If you mean voltage caused by refllected acoustic wave, multiply it by open loop gain of the amp.
janneman said:
Jan,
we are two here who have made notices about crossover distortion, I did it way over 100 posts ago when the SPLIF came up to discussion.
Post #350
BTW, Charles was fast to defend SPLIF by saying FET's having a "softer" crossover characteristics.
Et al,
and in general speaking, my point's with my posts in this thread so far has been more from a technical point of view and I reacted mostly because I felt some comments and analogies felt flawed, but I agree with Jan that there are so many people pleased with their amplifiers so sometimes arguing about the sound quality can be a difficult subject which has been proven earlier on this forum.
Cheers! 😉
Re: Re: Re: Re: Re: Charles Altman's splif topology
If I get it right, you imply the reflected sound waves impinging on the speaker make it to work as a microphone thus inducing an EMF which is seen in turn by the global feedback if present, adding to distortion.
My question is to what amount do you think / have measured this gets.
On the other hand if you imply this is multiplied by the OL gain well .... kindly I do not agree unless demonstrated by a formal analysis.
Rodolfo
darkfenriz said:
If I get it right, you imply the reflected sound waves impinging on the speaker make it to work as a microphone thus inducing an EMF which is seen in turn by the global feedback if present, adding to distortion.
My question is to what amount do you think / have measured this gets.
On the other hand if you imply this is multiplied by the OL gain well .... kindly I do not agree unless demonstrated by a formal analysis.
Rodolfo
Re: Re: Re: Charles Altman's splif topology
The cone velocity is proportional to voltage. The excursion is not proprtional to voltage. It's the velocity of the cone that creates the wave of pressurized and rarefied air in front of it. The higher the velocity, the higher the pressure of the air in front of the cone. So at the peak of the waveform, the cone will be at it's highest velocity, and just crossing zero displacement.
As for the Lowther drivers having the highest velocity, I do not know. The lightweight cone gives for good acceleration. While powerful magnets give you a high F/I (force/current) ratio, they also develop a lot more back EMF. In my opinion this is the best combination you could hope for in terms of sound quality.
Charles said:
The cone velocity is proportional to voltage. The excursion is not proprtional to voltage. It's the velocity of the cone that creates the wave of pressurized and rarefied air in front of it. The higher the velocity, the higher the pressure of the air in front of the cone. So at the peak of the waveform, the cone will be at it's highest velocity, and just crossing zero displacement.
As for the Lowther drivers having the highest velocity, I do not know. The lightweight cone gives for good acceleration. While powerful magnets give you a high F/I (force/current) ratio, they also develop a lot more back EMF. In my opinion this is the best combination you could hope for in terms of sound quality.
Jan,
Thank you for your post. You explain well where you are coming from but, as is my wont, I beg to differ.
We are operating in an area of patent subjectivity, where one man's meat is another man's poison. (or poisson, if you prefer...!) Naturally, of all the many types of amps around people will gravitate towards whatever floats their boat.
But Charles has presented here a new concept, and it deserves a thorough analysis - both objectively, in discussion and by measurement, and subjectively, in listening tests.
I have not yet built a SPLIF as he has patented it, but I've built amps with feedback coming off the VAS, the drivers, and the output - and mixtures of all three. I noted some extremely interesting results, which I struggle to explain objectively. I found that topologies with feedback taken from the drivers only gave improved imaging and a more delicate, twee and organic sound, akin to a tube amp. This comes back to a disconnection between the phase relationships at the load and those at the driver; obviously the output devices are buffering this anomaly, but many here insist the feedback should take full account of the phase relationships at the load. I would suggest that the differences I have heard, which were marked, would be even more exaggerated with the SPLIF, and I don't believe it is valid to insist that the feedback should reflect the phase relationships found at the load.
In closing, a scientific approach mandates that we observe a phenomenon and endeavour to explain it with known theory. It is difficult to do this, educated conjecture at best, if we fail to make the subjective observations first.
I don't claim spotless conception here, as I've not yet built the SPLIF. But I've built other topologies which are related, made subjective, useful observations, tried to delve into the reasons why, and extrapolated a little to confirm largely Charles' arguments. I believe vehement condemnation of the SPLIF and semantic argument about the type of feedback employed is not appropriate, and I can't see any point in further argument about semantics. It exists, it's like death and taxes, let's devote our energies to linking the subjectives with the objectives.
No offense intended, none taken, I love this forum........
Cheers,
Hugh
Thank you for your post. You explain well where you are coming from but, as is my wont, I beg to differ.
We are operating in an area of patent subjectivity, where one man's meat is another man's poison. (or poisson, if you prefer...!) Naturally, of all the many types of amps around people will gravitate towards whatever floats their boat.
But Charles has presented here a new concept, and it deserves a thorough analysis - both objectively, in discussion and by measurement, and subjectively, in listening tests.
I have not yet built a SPLIF as he has patented it, but I've built amps with feedback coming off the VAS, the drivers, and the output - and mixtures of all three. I noted some extremely interesting results, which I struggle to explain objectively. I found that topologies with feedback taken from the drivers only gave improved imaging and a more delicate, twee and organic sound, akin to a tube amp. This comes back to a disconnection between the phase relationships at the load and those at the driver; obviously the output devices are buffering this anomaly, but many here insist the feedback should take full account of the phase relationships at the load. I would suggest that the differences I have heard, which were marked, would be even more exaggerated with the SPLIF, and I don't believe it is valid to insist that the feedback should reflect the phase relationships found at the load.
In closing, a scientific approach mandates that we observe a phenomenon and endeavour to explain it with known theory. It is difficult to do this, educated conjecture at best, if we fail to make the subjective observations first.
I don't claim spotless conception here, as I've not yet built the SPLIF. But I've built other topologies which are related, made subjective, useful observations, tried to delve into the reasons why, and extrapolated a little to confirm largely Charles' arguments. I believe vehement condemnation of the SPLIF and semantic argument about the type of feedback employed is not appropriate, and I can't see any point in further argument about semantics. It exists, it's like death and taxes, let's devote our energies to linking the subjectives with the objectives.
No offense intended, none taken, I love this forum........
Cheers,
Hugh
Re: Back to Physics 101
V=N*dPhi/dt=N*d(B*S)/dt=N*B*section*dr/dt=N*B*section*V
right
I just want to say I am SORRY for makig the mess
Velocity is of course prop. to voltage
smoking-amp said:
V=N*dPhi/dt=N*d(B*S)/dt=N*B*section*dr/dt=N*B*section*V
right
I just want to say I am SORRY for makig the mess
Velocity is of course prop. to voltage
Ultima Thule said:
Sorry, I forgot, I just remembered Stevens post as it was recent. I have a pretty good memory but it is very short 😉
Jan Didden
AKSA said:
Hugh,
I think we understand _ and respect _ each others position, and I appreciate that. But looking at the quoted part of your post, I see some speculation and the danger of jumping to conclusions, words like "obviously" and "this comes back to" and "I would suggest". I want to avoid that, therefore my more theoretical approach. I am pretty sure that if you had found earlier that there is not really an engineering argument for splif, your listening experiences most probably would be leaning in a different direction.
But you are right, the semantic discussion was perhaps overdone a bit, although I STILL don't know what it is exactly what Charles claims. I searched for a patent in his name, couldn't find any. Did you find it?
Jan Didden
On patent issues
Okay, this one is on patent issues (my patents).
I recently stopped believing in patents because of the following reasons:
1) It can cost the inventor significant amounts of money to keep it up.
2) If the inventor is not extremely fortunate (in terms of a fortune to begin with), any larger company that just violates the inventor's patent can in most cases do so unhindered, as the inventor may not have the financial power to endure years of struggle in court... was there this Sony Walkman story ??
3) Great inventions are often quite stupid (simple) ideas. It actually sometimes is stupidity, ignorance and pure lack of knowledge that leads someone into unknown territory where he may find something significant. It is not an effort, it is rather the relaxation from the effort when the idea comes.
I have a rather limited knowledge about audio amplifiers (compared to some experts who know much more technical stuff about it) and the most interesting information (for me) I found on Nelson Pass's and Rod Elliot's websites.
But obviously it was the lack of knowledge plus lots of beer that led to my SPLIF amplifier invention. If I had had the knowledge, I would have stumbled over it, discarding the idea just before even contemplating it...
What I want to say is, that I would like to credit those who generously display their wisdom on the internet for others to learn and expand their consciousness and in this context I have stopped to support all of my own patents.
In other words: ITS FREE. EVERYBODY CAN BUILD IT, SELL IT... WHATEVER... and I will not sue your ***.
If money seeks you, it will find you anyway.
Charles 😉
ALTMANN MICRO MACHINES
http://www.jitter.de
http://www.amm.haan.de
http://www.altmann.haan.de
http://www.mother-of-tone.com
AKSA said:
Okay, this one is on patent issues (my patents).
I recently stopped believing in patents because of the following reasons:
1) It can cost the inventor significant amounts of money to keep it up.
2) If the inventor is not extremely fortunate (in terms of a fortune to begin with), any larger company that just violates the inventor's patent can in most cases do so unhindered, as the inventor may not have the financial power to endure years of struggle in court... was there this Sony Walkman story ??
3) Great inventions are often quite stupid (simple) ideas. It actually sometimes is stupidity, ignorance and pure lack of knowledge that leads someone into unknown territory where he may find something significant. It is not an effort, it is rather the relaxation from the effort when the idea comes.
I have a rather limited knowledge about audio amplifiers (compared to some experts who know much more technical stuff about it) and the most interesting information (for me) I found on Nelson Pass's and Rod Elliot's websites.
But obviously it was the lack of knowledge plus lots of beer that led to my SPLIF amplifier invention. If I had had the knowledge, I would have stumbled over it, discarding the idea just before even contemplating it...
What I want to say is, that I would like to credit those who generously display their wisdom on the internet for others to learn and expand their consciousness and in this context I have stopped to support all of my own patents.
In other words: ITS FREE. EVERYBODY CAN BUILD IT, SELL IT... WHATEVER... and I will not sue your ***.
If money seeks you, it will find you anyway.
Charles 😉
ALTMANN MICRO MACHINES
http://www.jitter.de
http://www.amm.haan.de
http://www.altmann.haan.de
http://www.mother-of-tone.com
Hi Charles,
Do not be discouraged because your idea falls apart on close inspection. Happens to everyone. If you don't try, you won't find. We all learn from the experience. (I know, some - OK, a few too many - of mine have flubbed too.) Many great ideas do start out from so called "dumb" ones, you just have to search out their limitations, sometimes the problems can then be fixed with another insight. This forum can often lead to that insight with a few other inputs. I don't worry about my ideas getting shot down anymore, I just come up with a better one. Eventually something successfully flies. And yes, patents mainly protect the Big guys. But if something flies well, the small guy might become the Big guy overnight.
On the original thread title: "anybody built a non fdbk ampl?"
I don't really want to read thru all 50 pages to see if someone already came up with this, but how about using current mirrors with gain to build a linear amplifier without feedback. This idea actually works quite well using vacuum tube current mirrors, but has not caught on yet in the incandescent component society.
*** Off Topic ****
Was Mikeks article on output stage safe area protection published in EW+WW? which issue?
Don B.
Do not be discouraged because your idea falls apart on close inspection. Happens to everyone. If you don't try, you won't find. We all learn from the experience. (I know, some - OK, a few too many - of mine have flubbed too.) Many great ideas do start out from so called "dumb" ones, you just have to search out their limitations, sometimes the problems can then be fixed with another insight. This forum can often lead to that insight with a few other inputs. I don't worry about my ideas getting shot down anymore, I just come up with a better one. Eventually something successfully flies. And yes, patents mainly protect the Big guys. But if something flies well, the small guy might become the Big guy overnight.
On the original thread title: "anybody built a non fdbk ampl?"
I don't really want to read thru all 50 pages to see if someone already came up with this, but how about using current mirrors with gain to build a linear amplifier without feedback. This idea actually works quite well using vacuum tube current mirrors, but has not caught on yet in the incandescent component society.
*** Off Topic ****
Was Mikeks article on output stage safe area protection published in EW+WW? which issue?
Don B.
smoking-amp said:
What !
I do not think Charles's idea has fallen apart in the slightest
My observation is that many people here that have NOT heard or tried this idea have critised or dismissed it while others that HAVE heard or tried this or similar ideas seem to like what they hear.
At the w/e I had the opportunity to listen to Susan Parker's Zero feedback amp as discussed in another thread and I have to say it was one of the best sounding systems I have heard anywhere. So for me the idea of keeping FB away from the amp output is very interesting.
So I would like to thank you Charles for this interesting and stimulating idea. I look forward to trying it sometime soon.
regards
mike
While I don't happen to think that Charles' design is something
I would pursue offhand, I believe his attitude and willingness to
experiment are healthy examples of DIY.
There are tons of ways to build NFB amps, and the content of
this thread has barely touched on the subject. 😎
I would pursue offhand, I believe his attitude and willingness to
experiment are healthy examples of DIY.
There are tons of ways to build NFB amps, and the content of
this thread has barely touched on the subject. 😎
- Status
- Not open for further replies.