Hi beppe61.
I've been playing around with this circuit topology for about 6 months now. I'm only using a 22V rail though. The linearity is good so i don't see the need for a higher rail, 2V RMS out will drive any power amp i have to full power and i can get 5V RMS with less than 0.1% DHT. At 22V it is easy to keep the power dissapation down.
Things I've found important.
1) to keep distortion low at high frequencies it is best to ommit C26
2) to reduce RFI suscepability use a RC filter before the pot (1k0 and 220pF seem to work well)
3) Most sonic problems come from interactions between the PSU and the input stage, to avoid this split R30 into 2 resistors and heavily decouple the centre of the two resistors. This can also be used to avoid turn on thump.
4)For super low output impedance buffer with a FET follower, but use a separatley regulated supply for the buffer.
5) Keep the collector current of TS7 low about 100uA, this allows higher values for R30 and R31 and raises input impedance.
The circuit deserves a good quality volume pot for best results, i'm currently using a 24W stepped attenuator.
Its a good fun little circuit which with care to detail can give great results. I've also used it as the gain stage for a 50W power amplifier
Chris
I've been playing around with this circuit topology for about 6 months now. I'm only using a 22V rail though. The linearity is good so i don't see the need for a higher rail, 2V RMS out will drive any power amp i have to full power and i can get 5V RMS with less than 0.1% DHT. At 22V it is easy to keep the power dissapation down.
Things I've found important.
1) to keep distortion low at high frequencies it is best to ommit C26
2) to reduce RFI suscepability use a RC filter before the pot (1k0 and 220pF seem to work well)
3) Most sonic problems come from interactions between the PSU and the input stage, to avoid this split R30 into 2 resistors and heavily decouple the centre of the two resistors. This can also be used to avoid turn on thump.
4)For super low output impedance buffer with a FET follower, but use a separatley regulated supply for the buffer.
5) Keep the collector current of TS7 low about 100uA, this allows higher values for R30 and R31 and raises input impedance.
The circuit deserves a good quality volume pot for best results, i'm currently using a 24W stepped attenuator.
Its a good fun little circuit which with care to detail can give great results. I've also used it as the gain stage for a 50W power amplifier
Chris
Dear Mr. Jennings,
Thank you very much indeed for your valuable instructions.
As I understand your are a great expert in the field, I would like to know your opinion on the subject of line preamps.
I have an old Nakamichi CA 5 - mkII that has about 30 active devices only in the line stage.
On the other hand there is the Nelson Pass Aleph P that has only one active component in the line stage.
And I understand that the Aleph P is much more appreciated within the audio people.
Why is some audio equipment so complex ?
Why don't audio manufactures prefer simpler circuits?
What is you opinion on the matter ?
Kind regards,
beppe61
Thank you very much indeed for your valuable instructions.
As I understand your are a great expert in the field, I would like to know your opinion on the subject of line preamps.
I have an old Nakamichi CA 5 - mkII that has about 30 active devices only in the line stage.
On the other hand there is the Nelson Pass Aleph P that has only one active component in the line stage.
And I understand that the Aleph P is much more appreciated within the audio people.
Why is some audio equipment so complex ?
Why don't audio manufactures prefer simpler circuits?
What is you opinion on the matter ?
Kind regards,
beppe61
Thank you for your kind words. I don't know about the 'great expert' I just know what i like 
Simpler circuits require attention to detail to get the best from them, so the inexperienced can reject them without spending the time required to optomize them.
Complex circuits often have more feed back, and this allows a low level of distortion to be reached with minimal effort. Unfortunately low THD is not the whole story. listen and tweak, then measure and tweak some more, Listen again and then go back to the start.
One reason people like complex circuits is that more satisfaction is gained when you finally get it to work, engineers tend to have too big egos and will therefore reject any design not worthy of their skills I'm too old for all that did a lot of my early work in milatary radio and satalite technollogy simplicity works for me.
Chris
Simpler circuits require attention to detail to get the best from them, so the inexperienced can reject them without spending the time required to optomize them.
Complex circuits often have more feed back, and this allows a low level of distortion to be reached with minimal effort. Unfortunately low THD is not the whole story. listen and tweak, then measure and tweak some more, Listen again and then go back to the start.
One reason people like complex circuits is that more satisfaction is gained when you finally get it to work, engineers tend to have too big egos and will therefore reject any design not worthy of their skills
I'm too old for all that did a lot of my early work in milatary radio and satalite technollogy simplicity works for me.Chris
Dear Mr. Jennings,
Thank you very much for your kind and valuable reply.
I am very interested in audio equipment but just because I love music and the well reproduced one.
I must confess you that I tried to learn some basic electronics but I failed unfortunately.
On the other hand it seems that engineers have failed in establishing a set of electric measurements to rank the sonic quality of different product.
For instance one of the biggest and neverending debate is about the use of negative feedback.
If you have to judge on the basis of measurements no question: negative feedback is a panacea.
But on the basis of some listening test it doesn't seem so.
What I can tell you is that I and a friend compared this 2 BJTs line preamp (very unelegantly built) and a commercial, complicated and much more expensive one and there was no contest.
The simpler and cheaper 2 BJTs won hands down: better image, voices reproduction, etc.
I strongly think that this is the way (ala ZEN style ).
Moreover if is time consuming to fine tune a simple device what kind of hell must be fine tuning a much more complex one.
I tell you this: I guess that some manufacturers have to fill their beautiful boxes with components to convince buyers to pay huge amount of money. It is all about marketing.
But marketing has nothing to do with soul.
Instead music concerns soul.
I also must confess you that I was expecting much more DIYers willing to build a prototype and report their outcomes.
May be their final aim is not music?
I have spent (sometime wasted) a good deal of money in this research just because I was not happy with the sound.
I tell you also that I am completing the regulated power supply of a Bride of Zen.
But I got already great results with the one pubblished on the Pass site.
I am waiting for someone with much more experience that builds one and tell me that is rubbish.
To end thank you very much for your kind and precious advices.
Your sincerely,
beppe61
Thank you very much for your kind and valuable reply.
I am very interested in audio equipment but just because I love music and the well reproduced one.
I must confess you that I tried to learn some basic electronics but I failed unfortunately.
On the other hand it seems that engineers have failed in establishing a set of electric measurements to rank the sonic quality of different product.
For instance one of the biggest and neverending debate is about the use of negative feedback.
If you have to judge on the basis of measurements no question: negative feedback is a panacea.
But on the basis of some listening test it doesn't seem so.
What I can tell you is that I and a friend compared this 2 BJTs line preamp (very unelegantly built) and a commercial, complicated and much more expensive one and there was no contest.
The simpler and cheaper 2 BJTs won hands down: better image, voices reproduction, etc.
I strongly think that this is the way (ala ZEN style ).
Moreover if is time consuming to fine tune a simple device what kind of hell must be fine tuning a much more complex one.
I tell you this: I guess that some manufacturers have to fill their beautiful boxes with components to convince buyers to pay huge amount of money. It is all about marketing.
But marketing has nothing to do with soul.
Instead music concerns soul.
I also must confess you that I was expecting much more DIYers willing to build a prototype and report their outcomes.
May be their final aim is not music?
I have spent (sometime wasted) a good deal of money in this research just because I was not happy with the sound.
I tell you also that I am completing the regulated power supply of a Bride of Zen.
But I got already great results with the one pubblished on the Pass site.
I am waiting for someone with much more experience that builds one and tell me that is rubbish.
To end thank you very much for your kind and precious advices.
Your sincerely,
beppe61
I'll just add that this topology was used in the line stage of the Dynaco FM5 tuner.
I also used the topology in a lab project for a 1 MHz amplifier and took extensive spectrum analyzer measurements to optimize the operating points. Small signal transistors should be used and you must watch out for voltage and power ratings. This design ran on 30V. We used 2N2222, 2N2219, and 2N2905s. Most of the larger power devices will have greater parasitic (non-linear) capacitances which will generally have higher distortion.
I prefer diff amp based designs for audio.
I also used the topology in a lab project for a 1 MHz amplifier and took extensive spectrum analyzer measurements to optimize the operating points. Small signal transistors should be used and you must watch out for voltage and power ratings. This design ran on 30V. We used 2N2222, 2N2219, and 2N2905s. Most of the larger power devices will have greater parasitic (non-linear) capacitances which will generally have higher distortion.
I prefer diff amp based designs for audio.
Hi again.
The Mr jennings is a little formal, Chris please
Listening tests are most important to me, if i want to build an amplifier for my measurement instruments to listen to then i may change my mind. I often find good technical reasons why a modification made through listening tests has made an improvement even though they have not show any difference in measurement.
I use amplifiers to enjoy music, this is there primary purpose, and how they sound to me is far more critical than how they measure.
I'm glad your recent build is giving you enjoyment, and not supprised at the good listening results you have been achieving. The benifit of DIY is that you don't have to follow what everyone else is doing.
Chris
The Mr jennings is a little formal, Chris please
Listening tests are most important to me, if i want to build an amplifier for my measurement instruments to listen to then i may change my mind. I often find good technical reasons why a modification made through listening tests has made an improvement even though they have not show any difference in measurement.
I use amplifiers to enjoy music, this is there primary purpose, and how they sound to me is far more critical than how they measure.
I'm glad your recent build is giving you enjoyment, and not supprised at the good listening results you have been achieving. The benifit of DIY is that you don't have to follow what everyone else is doing.
Chris
Hi beppe61,
Your position is not far from the view of rational value based engineering. A circuit should be only as complex as it needs to be to meet its requirements. A simpler circuit is more reliable (assuming no components are over stressed), lower cost, easier to manufacture and debug.
I adapted the 1 MHz design that I worked on more towards audio and simulated it. The performance is excellent, low distortion, fast slew rate, wide bandwidth. It's even fairly tolerant of transistor substitutions or parameter variations, but distortion is minimized with fast, small signal transistors. This version uses high bias in the output stage for speed and to keep it in Class A. There's really no reason not to build a well engineered version of that topology.
BTW, distortion is lower when the Miller cap is removed from the output device and a cap is placed across the feedback resistor instead.
Sorry, I never listened to the original version since it only had RF going through it and was mono.
Anyone know where to get those small extruded project boxes that were used in Audio Alchemy products?
Regards,
Pete B.
Your position is not far from the view of rational value based engineering. A circuit should be only as complex as it needs to be to meet its requirements. A simpler circuit is more reliable (assuming no components are over stressed), lower cost, easier to manufacture and debug.
I adapted the 1 MHz design that I worked on more towards audio and simulated it. The performance is excellent, low distortion, fast slew rate, wide bandwidth. It's even fairly tolerant of transistor substitutions or parameter variations, but distortion is minimized with fast, small signal transistors. This version uses high bias in the output stage for speed and to keep it in Class A. There's really no reason not to build a well engineered version of that topology.
BTW, distortion is lower when the Miller cap is removed from the output device and a cap is placed across the feedback resistor instead.
Sorry, I never listened to the original version since it only had RF going through it and was mono.
Anyone know where to get those small extruded project boxes that were used in Audio Alchemy products?
Regards,
Pete B.
hi,
same as my amp topology which I have been post here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=49927
same as my amp topology which I have been post here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=49927
Hi beppe61,
You have an excellent preamp in the Nakamichi. It is very quiet and sounds great. This is one of the preamps that deserved more attention. I used to do warranty on these and still have a high regard for them. BTW, every part in the Nak is there for a reason. It has local regulation and one transistor per regulator was used to keep ground currents to a minimum. Pretty smart concept actually.
PB2,
I wouldn't mind getting some Audio Alchemy boxes either, but they had a ventillation problem (another line I did warranty work for). You would be further ahead to make your own and redo the regulator heatsinks. Darn things were rivited on dry. Take them off, straighten them out and grease them. Use a screw to attach. A larger heatsink would not hurt some of these. Too bad A A didn't screw the boards down, they slid into a slot in the case. You could use right angle brackets with a notch in them to fasten the boards. You could then use a larger case to put multiple units in if you had them.
-Chris
You have an excellent preamp in the Nakamichi. It is very quiet and sounds great. This is one of the preamps that deserved more attention. I used to do warranty on these and still have a high regard for them. BTW, every part in the Nak is there for a reason. It has local regulation and one transistor per regulator was used to keep ground currents to a minimum. Pretty smart concept actually.
PB2,
I wouldn't mind getting some Audio Alchemy boxes either, but they had a ventillation problem (another line I did warranty work for). You would be further ahead to make your own and redo the regulator heatsinks. Darn things were rivited on dry. Take them off, straighten them out and grease them. Use a screw to attach. A larger heatsink would not hurt some of these. Too bad A A didn't screw the boards down, they slid into a slot in the case. You could use right angle brackets with a notch in them to fasten the boards. You could then use a larger case to put multiple units in if you had them.
-Chris
beppe61 said:
There are many factors, people don't all think alike or with pure logic. Some look at specs and reviews, others want to know how many transistors it has, and on, and on.
The engineers want to keep on designing. This design has been around for over 30 years, if management asks engineering for a new product line do you think they're going to say sorry this is good enough we don't need to do any better? I would not agree that one gain stage is all thats needed, this circuit has two and could use more current gain if it had to drive lower impedances, it does do the line stage job very well.
You might want to take a look at the Jensen 990 OP amp, some of it's complexity is there to make it general purpose but it is an excellent high gain design.
I added split supplies and a DC servo to my version so that I could eliminate the output capacitor. I like it.
What do you think of the wire with gain principle and amplifiers that provide a good null?
What if you had to choose between and amp that provided a deep null, and another with much worse specs but that sounded better, which would you pick?
Your in Torino__ITALY? - I like Italian food!
Regards,
Pete B.
Dear Pete,
Thank you very much for your kind and very valuable reply.
But I have to confess you that my level of understanding circuits is very very low.
I tried to learn some basics in simple circuit design but failed.
I have already read on the web about the Jensen 990 discrete op-amp and I am sure it is very good indeed.
Its success in the pro field speaks for itself.
At present I use a Bryston .4b line preamp that is similarly based upon a discrete op-amp schematic using a dual power supply +/- 24 V regulated by fixed regulators.
And I must tell you that sometimes I find it short on dynamics.
Clean but a little flat.
May be I am imaging things.
If I am not wrong you can see at the proposed schematic as a 2 BJT op-amp.
I am also very intrigued by the op-amps schematics proposed in the Nelson Pass article (3 active components only), but the schematics are not complete so I can't try for myself.
May be your are right when you say that two stage are needed.
But I have also experience with a prototype of the Nelson Pass design called Bride of ZEN. Just one mosfet, class A, no feedback, single ended.
If built with care (especially the power supply must be very good) the resulting sound is really amazing.
One problem however could be the slightly high output impedance (about 1 kohm).
I think that the best way to judge a sound reproduction system could be find some disks of nice music well recorded and then listen if the system can trasmit you some emotions.
Human voices are the best for this.
If it can, then "you have a damned good system" (Doug Sax in the Sheffield test CD).
Also the depth of image is very difficult to get.
And I find the Bryston a little short on the emotional side, especially with voices.
To answer to your question I think that the last test shoud be a listening test.
Kind regards,
Beppe
P.S. Yes, I live in Torino.
And I like Italian food too. Someone say too much.
Thank you very much for your kind and very valuable reply.
But I have to confess you that my level of understanding circuits is very very low.
I tried to learn some basics in simple circuit design but failed.
I have already read on the web about the Jensen 990 discrete op-amp and I am sure it is very good indeed.
Its success in the pro field speaks for itself.
At present I use a Bryston .4b line preamp that is similarly based upon a discrete op-amp schematic using a dual power supply +/- 24 V regulated by fixed regulators.
And I must tell you that sometimes I find it short on dynamics.
Clean but a little flat.
May be I am imaging things.
If I am not wrong you can see at the proposed schematic as a 2 BJT op-amp.
I am also very intrigued by the op-amps schematics proposed in the Nelson Pass article (3 active components only), but the schematics are not complete so I can't try for myself.
May be your are right when you say that two stage are needed.
But I have also experience with a prototype of the Nelson Pass design called Bride of ZEN. Just one mosfet, class A, no feedback, single ended.
If built with care (especially the power supply must be very good) the resulting sound is really amazing.
One problem however could be the slightly high output impedance (about 1 kohm).
I think that the best way to judge a sound reproduction system could be find some disks of nice music well recorded and then listen if the system can trasmit you some emotions.
Human voices are the best for this.
If it can, then "you have a damned good system" (Doug Sax in the Sheffield test CD).
Also the depth of image is very difficult to get.
And I find the Bryston a little short on the emotional side, especially with voices.
To answer to your question I think that the last test shoud be a listening test.
Kind regards,
Beppe
P.S. Yes, I live in Torino.
And I like Italian food too. Someone say too much.
Hi beppe61,
I have found most things with a high level of feedback very lifeless sounding. This is where most of the better designers are working. Some feedback is very benificial, but there is a point where the life gets sucked out of the music. The trick is to have good specs and low feedback. A balancing act.
-Chris
I have found most things with a high level of feedback very lifeless sounding. This is where most of the better designers are working. Some feedback is very benificial, but there is a point where the life gets sucked out of the music. The trick is to have good specs and low feedback. A balancing act.
-Chris
beppe61 said:
I tend to chose the amplifer that nulls, but this is not to say that your choice is wrong. I do this because I also enjoy speaker design and if I'm going to tailor the system to be more forgiving then I hope to do it in the speakers rather than the amplifier. My best system is not very forgiving by the way. My perspective is more from the designer's side rather than the listener's side.
As far as two stages goes I mean in most designs that uses feedback. And yes most simple designs do not have very good power supply rejection so it is more important to have a good supply. I only mention output impedance to have a design that is more general purpose, 1K out with a reasonable power amp input impedance is fine. Yes, I'm always looking for good recordings and I find that when the sound is good I tend to forget about the system and just enjoy the music.
As far as learning the electronics goes, I think hands on building is an excellent way, I like to mix study, building, and simulation. Did you build the Bride of Zen that you mentioned?
Best Regards,
Pete
Dear Pete,
" I tend to chose the amplifer that nulls"
Can you give me some more info on this "null-test" ?
I am very ignorant on the matter.
"When the sound is good I tend to forget about the system and just enjoy the music".
I noticed the same myself.
"Did you build the Bride of Zen that you mentioned?"
I have one built from a friend of mine.
It has the power supply shown in the Nelson Pass article.
Anyway I intend to build a new power supply based on the TL783 high voltage variable regulator from Texas which has very good specs.
Kind regards,
beppe61
" I tend to chose the amplifer that nulls"
Can you give me some more info on this "null-test" ?
I am very ignorant on the matter.
"When the sound is good I tend to forget about the system and just enjoy the music".
I noticed the same myself.
"Did you build the Bride of Zen that you mentioned?"
I have one built from a friend of mine.
It has the power supply shown in the Nelson Pass article.
Anyway I intend to build a new power supply based on the TL783 high voltage variable regulator from Texas which has very good specs.
Kind regards,
beppe61
anatech said:
Dear Chris,
Reading the very good reviews on preamp without feedback (Pass Aleph P, Conrad Johnson Art, etc.) I am inclined to think the same.
At least for preamps.
I have no information on power amps without feedback however.
I have one with low feedback that I like very much.
Thank you very much again.
Kind regards,
beppe61
analog_sa said:
Dear Mr. Analog,
You mean that the power supply circuit must be considered as and audio circuit ?
A preamp circuit in pure class A draws a low constant current from the power supply.
So the voltage regulator could work.
I was thinking to try because is very easy to implement a regulated power supply with a TL783, as you can see from its data sheet.
And I think that the proposed power supply does not reach the value of noise rejection mentioned in the article.
Also at Bryston they use the 317/117 couple in their top of the line preamp power supply.
I would like to know your opinion anyway.
Kind regards,
beppe61
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.