Nico,
I like your regulator maybe replace R9 with a current diode for better rejection. simple and sweet not to mention excellent, You amp with the enhanced wilson mirrors looks impressive. Could you explain your current source in a little more detail.
This thread is getting good.
Jam
I like your regulator maybe replace R9 with a current diode for better rejection. simple and sweet not to mention excellent, You amp with the enhanced wilson mirrors looks impressive. Could you explain your current source in a little more detail.
This thread is getting good.
Jam
Last edited:
jam,
as mr spock would say,
"indeed"
as another source of ideas, i can suggest a look through bob cordell's amp design from quite a while back; there are some similarities in the basic topology:
http://www.cordellaudio.com/papers/MOSFET_Power_Amp.pdf
mlloyd1
as mr spock would say,
"indeed"
as another source of ideas, i can suggest a look through bob cordell's amp design from quite a while back; there are some similarities in the basic topology:
http://www.cordellaudio.com/papers/MOSFET_Power_Amp.pdf
mlloyd1
mlloyd1 , the "purists" might say we are getting off base too much if we venture into the realm of hawkford cascodes and minimizing the early effect (cordell amp). Still a good read , one of the first in my collection.
Back to the original ,simplistic , primitive Goldmund ..... I was checking the 3 mod's to it's input/VAS stage and have plotted them in 3 "modes" 1.standard (either KSA/C or MPSA) 2.darlington VAS and 3. Both current mirror AND darlington.(like nico ras's amp)
These are shown below. Phase margin and UG point remain the same but distortion decreases as you move to the most complicated arrangement. Slew also increased with the Nico arrangement. Would this make for better sound or just amplification with absolutely no coloration (near perfect "wire with gain") ?? Does increased loop gain make for better amplification (main question).
This brings up another question ... What do we want from an amp ??? to add "flavor" or just to amplify faithfully ? #1 would be adequate and you would hear the original SK134/J49's in all their glory (if the thing didn't oscillate and smoke out the room) and #3 is essentially a PPM amp (praise Nico) if the input/VAS is tested in isolation (no output stage).
I'm still debating the merit vs. simplicity aspect for each one ... MANY more simulations. BTW , all 3 would work flawlessly and make beautiful music without failure .. but 1 would sound just a little better.
OS
Back to the original ,simplistic , primitive Goldmund ..... I was checking the 3 mod's to it's input/VAS stage and have plotted them in 3 "modes" 1.standard (either KSA/C or MPSA) 2.darlington VAS and 3. Both current mirror AND darlington.(like nico ras's amp)
These are shown below. Phase margin and UG point remain the same but distortion decreases as you move to the most complicated arrangement. Slew also increased with the Nico arrangement. Would this make for better sound or just amplification with absolutely no coloration (near perfect "wire with gain") ?? Does increased loop gain make for better amplification (main question).
This brings up another question ... What do we want from an amp ??? to add "flavor" or just to amplify faithfully ? #1 would be adequate and you would hear the original SK134/J49's in all their glory (if the thing didn't oscillate and smoke out the room) and #3 is essentially a PPM amp (praise Nico) if the input/VAS is tested in isolation (no output stage).
I'm still debating the merit vs. simplicity aspect for each one ... MANY more simulations. BTW , all 3 would work flawlessly and make beautiful music without failure .. but 1 would sound just a little better.
OS
Attachments
Hi Guys,
There was no intension competing with anything or any other of the designs, use it or lump it. Just thought to throw in something that exists and maybe use pices of all the designs floating around to come up with a great amp. I do not want to hi-jack this thread at all. It is Kean's thread, he must decide.
I cannot take the credit for the CCS is not my idea but I hardly remember who gave it to me, I think it was from from Nikwal, I used it in some of the other amps and worked well.
As I have said, this amp is vanilla and even I had to get used to it at first, there is nothing impressive or musical about it, it is well, just an amp and not everyone's cup of tea.
Why did I regulate the whole amp and not just the front end - I did at first and then went one step further, it became more vanilla and in line with the design brief.
If I were to say that some of my amps run on 100kHz SMPS and they got better no-one would believe it either.
There was no intension competing with anything or any other of the designs, use it or lump it. Just thought to throw in something that exists and maybe use pices of all the designs floating around to come up with a great amp. I do not want to hi-jack this thread at all. It is Kean's thread, he must decide.
I cannot take the credit for the CCS is not my idea but I hardly remember who gave it to me, I think it was from from Nikwal, I used it in some of the other amps and worked well.
As I have said, this amp is vanilla and even I had to get used to it at first, there is nothing impressive or musical about it, it is well, just an amp and not everyone's cup of tea.
Why did I regulate the whole amp and not just the front end - I did at first and then went one step further, it became more vanilla and in line with the design brief.
If I were to say that some of my amps run on 100kHz SMPS and they got better no-one would believe it either.
Kean you are correct a ceramic resonator will look like a cap with the tester.
This is what I suspected as one contributor to the "vanilla" flavor, not just the VAS alone as ostripper stated.
Yes, I have never gotten used to this "straight wire with gain" sound. You said that even you had to get used to it at first. I'm not trying to criticize but you also have to remember that sometimes you don't design an amp simply for personal taste only.
Right. Here is my reasoning why I cannot accept the concept of straight wire with gain...
(1) There's no such thing as straight wire with gain. A wire doesn't have a gain, and an amplifier block is not a wire. If wire can contribute something to the sound, how about those parts in an amplifier? Change the power supply diode from ultrafast diode to another that I would call poor sounding one, what will you hear? Can you see it in LTSpice? May be. Here I expect we would be honest and agree that an amplifier will change/add something to the sound.
(2) What is a "vanilla" sound? Is it or is it not a FLAVOR? Output that is similar with input? How would you you define similarity here? Let me give an analogy...
A human can be used as an input to a magic box. The output can be a monkey with human's character and intelligence, or a human with monkey's character and intelligence. Is the new human better than the monkey because it is "more similar"?
I would like to have an output that is similar with the input, as long as it doesn't change or add something that I don't want.
AND there is another argumentation about the nature of this something subtle that I don't want in a sound reproduction of an amplifier and loudspeaker combo. But that is for another time...
Oh, BTW, this is on topic with the spirit of this thread. As I had a suspect that "improvement project" will most probably lead to something that doesn't have correlation with the original design used as a benchmark (here the Goldmund). Because people will try to implement what is in the book, and at the end get something that is "blameless". No old circuit, no simple circuit, no unstable circuit, no theoretical mistakes, nothing. Don't we have enough of such amplifiers?
Inside a practical built amplifier there is a rather hostile environment... trafos humming, capacitors charging....mechanical springs pressing relays.. ect ect.. and all this matters a lot....maybe even more so than if the distortion is so so low...
I know from my work in designing loudspeakers that coils are so so different.. and that the physical proprietaries of this, for a speaker vital component, affect the sound to an almost unbelievable degree...and so does little things such a lengths diameters and the isolation of cables. This project is very very nice...an can be a good foundation for an amazing amplifier...if only...!!
Jay sort of nails it to the point....no such thing as a neutral amplifier...Sounding good, is a human perception..and not a test occurrence.
Interesting here is how do we quantify the human goodness factor. how do we make sure that the equipment..sings the way we want it...??
For me...in designing loudspeakers energy storage and energy absorption as an absolute no no..so things like soft membranes and high moving mass..is a totally dead end.. same thing must apply to amplifiers..any trace of what just was must not be present...here thinking about memory distortion..(I think highly of the sound from Lavardin amplifiers). And in the same token..The shielding of high power devices from magnetic or paramagnetic materials..Bronze Copper or silver placed between the output devices and aluminium cooling fins.. works wonders...
This is not an attempt to derail this fine thread but merely to state that all we do matters and audio wonders are not only attached to supreme specs...
I know from my work in designing loudspeakers that coils are so so different.. and that the physical proprietaries of this, for a speaker vital component, affect the sound to an almost unbelievable degree...and so does little things such a lengths diameters and the isolation of cables. This project is very very nice...an can be a good foundation for an amazing amplifier...if only...!!
Jay sort of nails it to the point....no such thing as a neutral amplifier...Sounding good, is a human perception..and not a test occurrence.
Interesting here is how do we quantify the human goodness factor. how do we make sure that the equipment..sings the way we want it...??
For me...in designing loudspeakers energy storage and energy absorption as an absolute no no..so things like soft membranes and high moving mass..is a totally dead end.. same thing must apply to amplifiers..any trace of what just was must not be present...here thinking about memory distortion..(I think highly of the sound from Lavardin amplifiers). And in the same token..The shielding of high power devices from magnetic or paramagnetic materials..Bronze Copper or silver placed between the output devices and aluminium cooling fins.. works wonders...
This is not an attempt to derail this fine thread but merely to state that all we do matters and audio wonders are not only attached to supreme specs...
I think the problem is that the more audiophile we become the more we listen to the equipment instead of enjoying the music. As I said in an earlier post, it may be the flaws in the equipment that contributes or adds character to the reproduced signal and induces pleasure for the listener. Maybe we are removing the pleasure from the performance. I think we turning pleasure into a quantifyable science.
We listen to music, do we all listen to natural instuments only, I think not. Is there a lot of instruments that is electronically synthesized these days - absolutely. Are electric instruments modified by fuzz and distrotion boxes to create effects, is a Marshall or PeeVee amp the ultimate amplifier and produces absolutely the signal that the guitar string made - no. Then why are we bothered with amplifiers with zero distortions, infinite infinite this or that?
Only because we are obsessed with proving that we are ultimate in our assumtions of how something must sound.
We listen to music, do we all listen to natural instuments only, I think not. Is there a lot of instruments that is electronically synthesized these days - absolutely. Are electric instruments modified by fuzz and distrotion boxes to create effects, is a Marshall or PeeVee amp the ultimate amplifier and produces absolutely the signal that the guitar string made - no. Then why are we bothered with amplifiers with zero distortions, infinite infinite this or that?
Only because we are obsessed with proving that we are ultimate in our assumtions of how something must sound.
An experience I am sure other may share. At a site where a hundred or so African workers were digging trenches, they sang while they toiled.
These were not professional artists, they have no musical training nor practiced the song before, only simple laborers. The lyrics was maybe five or so words but they sang in absolute harmony and perfect rhythm while digging.
Hearing this mde the hair on my neck rise, gave me goosebumps and a slug in my throat. I have not in all my audio years, ever heard any recording chain, regardless of complexity, power, weight or cost to reproduce this sound or emotional experience - not even close.
But there is al lot of fun in trying
These were not professional artists, they have no musical training nor practiced the song before, only simple laborers. The lyrics was maybe five or so words but they sang in absolute harmony and perfect rhythm while digging.
Hearing this mde the hair on my neck rise, gave me goosebumps and a slug in my throat. I have not in all my audio years, ever heard any recording chain, regardless of complexity, power, weight or cost to reproduce this sound or emotional experience - not even close.
But there is al lot of fun in trying
How true, 20 years ago my DIY audio buddy stated he was building his last amp, after that he would listen to music only.
Took a very long time for me to fully comprehend his message, problem is that the amp thing keeps on itching.
Speaking of Lavardin, i did Gerard Perrot's fully regulated amp design (1990), MOSFETs for the regulator section and BJT's for the amp output.
50W continuous only, doing the same for the output of a +150W amp with a +100V/uS SR figure and a MHz bandwidth qualifies as rocket science and a money pit.
OS raises a good point.............are we making the amp too complicated?
Do we need a darlington VAS?
Is too much open loop gain bad?
Will trying to get the best distortion perfornance ultimately hurt the sound?
Einstein once said "Make everything as simple as possible, but not simpler."
The more junctions in the signal path the more problems we are going to face with stability. My thoughts are to choose the best devices possible, replace circuit elements such as current sources and mirrors and differentials with improved versions and fine tune the result .
The choice of output stages gives us some interesting choices
a) Keep the original
b) Replace it with a biplolar . ( As in OS's Mongrul-this intrigues me the most)
or
c) Mosfet drivers with bipolar outputs.
This is probably where we are going to have the biggest debate.
I think the basic topology is sound the key is refining the various circuit elements that make up the amplifier.
Jam
Do we need a darlington VAS?
Is too much open loop gain bad?
Will trying to get the best distortion perfornance ultimately hurt the sound?
Einstein once said "Make everything as simple as possible, but not simpler."
The more junctions in the signal path the more problems we are going to face with stability. My thoughts are to choose the best devices possible, replace circuit elements such as current sources and mirrors and differentials with improved versions and fine tune the result .
The choice of output stages gives us some interesting choices
a) Keep the original
b) Replace it with a biplolar . ( As in OS's Mongrul-this intrigues me the most)
or
c) Mosfet drivers with bipolar outputs.
This is probably where we are going to have the biggest debate.
I think the basic topology is sound the key is refining the various circuit elements that make up the amplifier.
Jam
Last edited:
I think design by committee is a bad thing, and this is certainly becoming that.
Everyone wants to chip in and improve the amp somewhere. Everyone has their favourite components or the "best" way of doing something.
Nagy's is right on one account, this amp will be a totally different beast if we continue like this.
I think we should focus on the flaws in the original design, and not try to "improve" the rest of the design. Just keep it simple.
PS: I use "we" but I haven't contributed anything to the design of course, just watching from the sideline.
Everyone wants to chip in and improve the amp somewhere. Everyone has their favourite components or the "best" way of doing something.
Nagy's is right on one account, this amp will be a totally different beast if we continue like this.
I think we should focus on the flaws in the original design, and not try to "improve" the rest of the design. Just keep it simple.
PS: I use "we" but I haven't contributed anything to the design of course, just watching from the sideline.
Nagy has his own Goldmund clone thread.
If you don't like improving the Goldmund and want to clone the original then this thread is not for you.
Now that is a good idea, configer what you think you need, upgrade as you feel fit or remove what you are unhappy with. An input with two the two differentials, and output/driver board one with MOSFETs and one with BJTs, a PSU with caps only, a regulator for overall regulation, a regulator for only front end, a speaker protection of sorts....
I would vote for that!
Mace,
I think you are quite wrong in this case. Keantoken is heading up the project and the rest of us are contributing. The main thing you fail to see is the exchange of ideas, so in the unlikely event that it does not get completed everyone is richer from the experience, this differenciates us from sheep, but if following the herd is your thing then I suggest the other thread.
So lets see what we agree upon
a) Keep basic topology intact
b) Nico's regualtor
c) Fet input
...........
The areas of debate are
a) Output stage
b) Darlington differential for second stage. (what if we raised the current in the first differential?)
c) ............
All the above is subject to change.
There is some merit to splitting the output stage from the front end but there is some elegance to building this on one board.
Jam
I think you are quite wrong in this case. Keantoken is heading up the project and the rest of us are contributing. The main thing you fail to see is the exchange of ideas, so in the unlikely event that it does not get completed everyone is richer from the experience, this differenciates us from sheep, but if following the herd is your thing then I suggest the other thread.
So lets see what we agree upon
a) Keep basic topology intact
b) Nico's regualtor
c) Fet input
...........
The areas of debate are
a) Output stage
b) Darlington differential for second stage. (what if we raised the current in the first differential?)
c) ............
All the above is subject to change.
There is some merit to splitting the output stage from the front end but there is some elegance to building this on one board.
Jam
Last edited: