Great thread starter Harry. I especially liked page 72 of the Linear app note.
None of the articles I read cover the basics, and audio hobbyists may not be well acquainted with the underlying physics. Everyone should read up on the subject of oscillating systems, either from the electrical angle of LC tank circuits, or from the mechanical approach of spring-mass oscillators. Of special importance are the concepts of underdamped, critically-damped, and underdamped oscillators. Critically damped systems settle to the desired level in the shortest possible time. Underdamped systems oscillate before reaching the desired level. This is seen as circuit noise, EMI, etc. Overdamped systems reach the desired level very slowly, and can be the cause of too high impedance in a circuit.
Without understanding this you cannot design a damper!
Cheers,
jwb
None of the articles I read cover the basics, and audio hobbyists may not be well acquainted with the underlying physics. Everyone should read up on the subject of oscillating systems, either from the electrical angle of LC tank circuits, or from the mechanical approach of spring-mass oscillators. Of special importance are the concepts of underdamped, critically-damped, and underdamped oscillators. Critically damped systems settle to the desired level in the shortest possible time. Underdamped systems oscillate before reaching the desired level. This is seen as circuit noise, EMI, etc. Overdamped systems reach the desired level very slowly, and can be the cause of too high impedance in a circuit.
Without understanding this you cannot design a damper!
Cheers,
jwb
It's easy to talk about specs, but it's hard to find a good sounding diode
No, it is not easy to talk about specs and interpret them. What constitutes a good sounding diode may be dependent on the circuit in which it is used. I started this whole dialog to address some common questions that seem to resurface from time to time. For example:
Are shottkys a better choice than fast recovery diodes?
How do snubber circuits work?
Will putting a snubber circuit on a conventional diode bridge work as well as a bridge made from fast recovery diodes?
Is putting a capacitor across each diode and/or the transformer windings a good idea?
What makes a good sounding fast recovery diode how do I pick one with having to listen to them all?
Are 30 Amp diodes better sounding than 8 Amp diodes?
What diodes sound best for digital supplies?
Will snubber circuits work with fast recovery diodes?
There are dozens of FRED type diodes, and listening to all of them is not practical or economical for most people. There seems to be a diode of the week syndrome that is very frustrating to many people. I believe that many of the fast recovery diodes chosen are not the best ones to use. There are many new types coming out with optimization of other parameters than the fastest reverse recovery times. The idea is to turn this into a little more engineering and a lot less mysticism. There are dozens of fast recovery diodes that might be worth listening to and hundreds of permutations of snubber circuits or parallel
capacitors that could be tried. Anyone have a few hundred hours to do this entirely by listening test alone. I have been messing with this diode stuff for about 10 years and still have a lot questions. I have also done a lot of work on EMI reduction in telecom and have some new perspective on diode noise and RFI suppression. I used the IXYS diodes, based on a little research, long before their word of mouth reputation spread. Those who wish to do all this based only on listening test else are free to ignore this thread. For those willing to study the literature, measure, model, AND LISTEN; I welcome and need your input. I would really read the references in this thread as you will have no idea what is being discussed without doing so.
A tutorial on resonant circuits and critical damping (those who call it dampening are all wet) can be found here:
http://www.ee.surrey.ac.uk/Teaching...circuit-theory/
The applet at: http://www.ee.surrey.ac.uk/Teaching...ampedcases.html
is very interesting to see the effect of damping factor on a resonant circuit.
No, it is not easy to talk about specs and interpret them. What constitutes a good sounding diode may be dependent on the circuit in which it is used. I started this whole dialog to address some common questions that seem to resurface from time to time. For example:
Are shottkys a better choice than fast recovery diodes?
How do snubber circuits work?
Will putting a snubber circuit on a conventional diode bridge work as well as a bridge made from fast recovery diodes?
Is putting a capacitor across each diode and/or the transformer windings a good idea?
What makes a good sounding fast recovery diode how do I pick one with having to listen to them all?
Are 30 Amp diodes better sounding than 8 Amp diodes?
What diodes sound best for digital supplies?
Will snubber circuits work with fast recovery diodes?
There are dozens of FRED type diodes, and listening to all of them is not practical or economical for most people. There seems to be a diode of the week syndrome that is very frustrating to many people. I believe that many of the fast recovery diodes chosen are not the best ones to use. There are many new types coming out with optimization of other parameters than the fastest reverse recovery times. The idea is to turn this into a little more engineering and a lot less mysticism. There are dozens of fast recovery diodes that might be worth listening to and hundreds of permutations of snubber circuits or parallel
capacitors that could be tried. Anyone have a few hundred hours to do this entirely by listening test alone. I have been messing with this diode stuff for about 10 years and still have a lot questions. I have also done a lot of work on EMI reduction in telecom and have some new perspective on diode noise and RFI suppression. I used the IXYS diodes, based on a little research, long before their word of mouth reputation spread. Those who wish to do all this based only on listening test else are free to ignore this thread. For those willing to study the literature, measure, model, AND LISTEN; I welcome and need your input. I would really read the references in this thread as you will have no idea what is being discussed without doing so.
A tutorial on resonant circuits and critical damping (those who call it dampening are all wet) can be found here:
http://www.ee.surrey.ac.uk/Teaching...circuit-theory/
The applet at: http://www.ee.surrey.ac.uk/Teaching...ampedcases.html
is very interesting to see the effect of damping factor on a resonant circuit.
Fred,
dunno what happened, but your last links seem to be wrong (at least for me). I think you meant this one for theory, and pointed here for the applet. Right ?
dunno what happened, but your last links seem to be wrong (at least for me). I think you meant this one for theory, and pointed here for the applet. Right ?
Gaincard Diodes
Hi Vladimir, You brought up a very interesting link for adepts to the gainclone showing the inwards of the Gaincard and clues to the construction of the Power Humpty.
The FE5D is a nick-name for the General Semiconductor BYV28-200 diode. Datasheet now at www.vishay.com (200V 3.5A )
http://www.vishay.com/search?query=byv28-200
vladimir said:
Hi Vladimir, You brought up a very interesting link for adepts to the gainclone showing the inwards of the Gaincard and clues to the construction of the Power Humpty.
The FE5D is a nick-name for the General Semiconductor BYV28-200 diode. Datasheet now at www.vishay.com (200V 3.5A )
http://www.vishay.com/search?query=byv28-200
applets and oranges
The correct URL for the damping applet is at:
http://www.ee.surrey.ac.uk/Teaching/Courses/circuit-theory/section7/dampedcases.html
Be sure to read the previous sections. It takes one from the definition of resistance, inductance, and capacitance to describing phasors, complex impedance, and finally to the RCL circuit and damping. It is a good intro or refresher to the concept of reactance and resonance.
Some people have remarked on the difficulty of measuring rf ringing in a power circuit and I will agree. The combination of scope and an RF detection and rectification circuit is probably the
best approach. An RF "sniffer" circuit can be constructed and a simple circuit is at:
http://sound.westhost.com/project74.htm
The correct URL for the damping applet is at:
http://www.ee.surrey.ac.uk/Teaching/Courses/circuit-theory/section7/dampedcases.html
Be sure to read the previous sections. It takes one from the definition of resistance, inductance, and capacitance to describing phasors, complex impedance, and finally to the RCL circuit and damping. It is a good intro or refresher to the concept of reactance and resonance.
Some people have remarked on the difficulty of measuring rf ringing in a power circuit and I will agree. The combination of scope and an RF detection and rectification circuit is probably the
best approach. An RF "sniffer" circuit can be constructed and a simple circuit is at:
http://sound.westhost.com/project74.htm
Re: The FE5D is a nick-name for the General Semiconductor BYV28-200 diode.
Yes you are absolutely right. I looked in the wrong column.
Datasheet of the FE5D is here:
http://www.vishay.com/search?query=fe5d
So it IS a 5A diode.
Hi fred,Fred Dieckmann said:
Yes you are absolutely right. I looked in the wrong column.
Datasheet of the FE5D is here:
http://www.vishay.com/search?query=fe5d
So it IS a 5A diode.
Diode dilema
I based my comparison on the data sheets from Vishay and will do some further reseach. I didn't mean to blow it out of proportion since they do appear to very similar even with the differences stated in the data sheet. I am also doing further investigation into snubber networks and their optimization with various diode types. Reading further articles on EMI generation from diodes reverse recovery effects has uncovered even more knowledge on this subject. I will try to conduct some Spice modeling and measurements to go along with the theory and diode data sheet numbers. The topic is really worth investigating since RFI in audio equipment can very much effect the sound. The noise from the rectifier bridge may be a major contributer to this RFI noise.
I have also invited alfsch to share his measurement results with us from the post in the "replacement diodes for 1N4007" thread:
http://www.diyaudio.com/forums/showthread.php?threadid=3955&highlight=BA159
I based my comparison on the data sheets from Vishay and will do some further reseach. I didn't mean to blow it out of proportion since they do appear to very similar even with the differences stated in the data sheet. I am also doing further investigation into snubber networks and their optimization with various diode types. Reading further articles on EMI generation from diodes reverse recovery effects has uncovered even more knowledge on this subject. I will try to conduct some Spice modeling and measurements to go along with the theory and diode data sheet numbers. The topic is really worth investigating since RFI in audio equipment can very much effect the sound. The noise from the rectifier bridge may be a major contributer to this RFI noise.
I have also invited alfsch to share his measurement results with us from the post in the "replacement diodes for 1N4007" thread:
http://www.diyaudio.com/forums/showthread.php?threadid=3955&highlight=BA159
Re: McCormack strategy...
I strongly suspect that D7 to D14 aren't "regular" 1N4007
They look like Telefunken (now Vishay) SF4007 "Super Fast Soft Recovery Rectifier" 1000V/1A according from the Sintered Glass Case SOD 57.
They are also sold by Partsconnexion at 50 cents each.
Thus Mc Cormak takes the best from ONE world....
Tube_Dude said:MackCormack strategy...for the two schools of thougt been happy.(fast versus regular diodes)...they use boths!!
D3,4,5,6---Fast
D7 ....to....D14 ---regular 1N4007...
So the best of both wolds!
We can called it...new hibrid rectification!!!
I strongly suspect that D7 to D14 aren't "regular" 1N4007
They look like Telefunken (now Vishay) SF4007 "Super Fast Soft Recovery Rectifier" 1000V/1A according from the Sintered Glass Case SOD 57.
They are also sold by Partsconnexion at 50 cents each.
Thus Mc Cormak takes the best from ONE world....
Peter Daniel said:It's easy to talk about specs, but it's hard to find a good sounding diode.
What I meant here is the specs not necesserily relate to the sound.
I was extremaly lucky that the first prototype of my gainclone was built with Motorola MUR860 diodes. I used them because they were on my bench and I didn't think twice about checking the specs. At the time it seemed to me that they must be better than regular bridges. The gainclone sounded extremaly well with them. So well that I put completing my Aleph projects aside and directed all my efforts into gaincloning.
Later I tried to improve on my original design. I built different supplies with seemingly better transformers and wires, but I could never achieve the same magic I got with my first amp. It just wasn't happening (because ea. time I was using different diodes). Eventually, after some time I decided to check what diodes I used in my first PS. They were Motorola parts and I couldn't get them anymore, so I ordred same diode but different vendor from Digi-Key (Fairchild). I'm sorry but it didn't deliver. It was the same type, MUR860, yet it sounded much harsher with some glare and I sent 100 of them back to the DK. In the meantime I checked at least 20 different brands and types. Some schottky, IRF, fast recovery, soft recovery, I checked most of the similar match to 860: 1520, 1560, 1060, 1045, 1620, 1020; none of them sounded as good as MUR860 with my gainclone. Just recently I found out that ON Semi is very close in sound to Motorola part, but again, Fairchild doesn't even come close.
So, if the same diode from two different vendors differs in sonics, what the specs are good for? Not that I'm undermining anybodys engineering authority, or being not scientific. All I'm saying is that listening with a given application is a final test for any part. And I agree that looking at specs will help in limiting the choices.
Damon Hill said:
That concerns the circuit they working in, not diodes alone. Then it leads to the question of the way of usage of high-speed diodes. Seems similar to the usage of high-speed opamps.
I learnt, or I believed that I learnt, Schottky diodes don’t have a recovery ringing. It reads so in the local University textbook. It reads the same here , and it also reads the same in the many places. Last night, in the spirit of this thread, I searched a bit through the documents on this topic. And I found this article , chapter Switching Transients , claiming, without the question or doubts about it, the Schottky diodes DO ringing!!! Well, literature can be quite confusing.
Pedja
Fred Dieckmann
Thanks again for an excellent thread. Peter's observations about the sound being the final arbitor are definitley on target but even he will admit that if he didn't get lucky the first time with the 860 part to use as a standard to judge the others, he very well may have settled for an inferior diode and stopped looking because it was the best he could find. After all, we couldn't expect him or any other person to go through the tedium of testing every diode or bridge on the market. I think this is a great way to get a feel of what diodes work in what scenarios!
Peter Daniel
May I ask what diode/bridge you use in your AlephX? The AlephX is an excellent amp to compare and contrast with the gainclones because they are so very different in their design philosophy. I have been playing around with the gainclones and find that as a general rule, more capacitance in the supplies effects the sound negatively but that some diodes work better with a larger capacitance than others. I hope this thread will begin to explain this curious phenomena.
Thanks again for an excellent thread. Peter's observations about the sound being the final arbitor are definitley on target but even he will admit that if he didn't get lucky the first time with the 860 part to use as a standard to judge the others, he very well may have settled for an inferior diode and stopped looking because it was the best he could find. After all, we couldn't expect him or any other person to go through the tedium of testing every diode or bridge on the market. I think this is a great way to get a feel of what diodes work in what scenarios!
Peter Daniel
May I ask what diode/bridge you use in your AlephX? The AlephX is an excellent amp to compare and contrast with the gainclones because they are so very different in their design philosophy. I have been playing around with the gainclones and find that as a general rule, more capacitance in the supplies effects the sound negatively but that some diodes work better with a larger capacitance than others. I hope this thread will begin to explain this curious phenomena.
The hard way
I believe that there are better diodes than the ones Peter has recommended. This belief is based on years of listening test and what parameters I believe may likely correlate to good sounding diodes. The use of snubbers may also totally change the optimum diode choice. The particular diode choice may also depend on the transforme and the decoupling caps used for the amp. Reading the links I have posted may shed some light on this. I do not mean to discourage those with deep pockets and lots of free time, from conducting listening test of diodes with a few dozen part numbers, some with multiple vendors. There seems to be indications that snubber circuits may work well with even fast recovery diodes. The permutations shouldn't require more that a few hundred hours of experiments and listening. I eagerly await the results while I try a more measurement and specifications oriented approach to reducing power supply noise. Yes, I plan on further listening test as well.
I believe that there are better diodes than the ones Peter has recommended. This belief is based on years of listening test and what parameters I believe may likely correlate to good sounding diodes. The use of snubbers may also totally change the optimum diode choice. The particular diode choice may also depend on the transforme and the decoupling caps used for the amp. Reading the links I have posted may shed some light on this. I do not mean to discourage those with deep pockets and lots of free time, from conducting listening test of diodes with a few dozen part numbers, some with multiple vendors. There seems to be indications that snubber circuits may work well with even fast recovery diodes. The permutations shouldn't require more that a few hundred hours of experiments and listening. I eagerly await the results while I try a more measurement and specifications oriented approach to reducing power supply noise. Yes, I plan on further listening test as well.
I also believe there are better diodes out there than the ones I mentioned. This one is only best sounding among the ones I tried. I'm open to any suggestions and always ready to try.
Accidentaly, that was first diode I used with the circuit and it proved to be quite OK. But one thing is certain now, when building an amp I will always listen to bridges and caps. I was never doing this before as I was assuming that the difference can't be that big.
As additional info, the system I'm using for testing isn't anything special: good Technics transport, Sonic Frontiers Trans Dac and my B&W 802 clones.
Accidentaly, that was first diode I used with the circuit and it proved to be quite OK. But one thing is certain now, when building an amp I will always listen to bridges and caps. I was never doing this before as I was assuming that the difference can't be that big.
As additional info, the system I'm using for testing isn't anything special: good Technics transport, Sonic Frontiers Trans Dac and my B&W 802 clones.
nania said:
As I remember the numbers correctly it was 30ETH06. I tried 15ETH06 with gainclone, but it was not as good as MUR860. I remember that I also liked MUR1520, but it was very soft sounding with lots of detail and was not that well tonally balanced as 860.
I bought some expensive 40A bridges from IXYS and I will try them in my next AlephX.
Re: The hard way
While it is true that the only thing I'm doing currently is Audio, it is also a cause that my pockets became empty long time ago. But at certain time in life, a person has to decide what's he's true calling is and follow the dream. It may end up both ways, but I would blame myself if I didn't try.
It also makes me wonder, that some people standarise on a certain part without even trying to listen, only because specs seem to be fine.
Fred Dieckmann said:
While it is true that the only thing I'm doing currently is Audio, it is also a cause that my pockets became empty long time ago. But at certain time in life, a person has to decide what's he's true calling is and follow the dream. It may end up both ways, but I would blame myself if I didn't try.
It also makes me wonder, that some people standarise on a certain part without even trying to listen, only because specs seem to be fine.
Peter Daniel
IXYS makes many diodes, can you tell me the part number? Have you tried the GA series from them? Tell me if you tried any others in the AlephX and how you thought they sounded. Also, I would like to know if you hear any difference between common cathode type and standard type diodes.
IXYS makes many diodes, can you tell me the part number? Have you tried the GA series from them? Tell me if you tried any others in the AlephX and how you thought they sounded. Also, I would like to know if you hear any difference between common cathode type and standard type diodes.
Fred Dieckmann
Since you brought up transformers in this thread, I have two questions about them. What is meant by a transformer with a "cap input"? Also, what do you believe is the best type of transformer for audio and please don't be curt, I prefer reading your concise explanations more than the verbose company promotional material of self interested corporations
Since you brought up transformers in this thread, I have two questions about them. What is meant by a transformer with a "cap input"? Also, what do you believe is the best type of transformer for audio and please don't be curt, I prefer reading your concise explanations more than the verbose company promotional material of self interested corporations
Nania -
A "cap input" is a power supply in which the rectifiers are followed, downstream as it were, by capacitors of sufficient size as to cause the voltage to rise to its max value, depending on which rectifier scheme is being employed.
This value will be greater than the "other" input supply, a "choke input" in which the rectifiers are followed by, get this, an inductor called a choke. For a nice little chart of rectifier schemes and input schemes, check the Hammond transformer site. - Pat
A "cap input" is a power supply in which the rectifiers are followed, downstream as it were, by capacitors of sufficient size as to cause the voltage to rise to its max value, depending on which rectifier scheme is being employed.
This value will be greater than the "other" input supply, a "choke input" in which the rectifiers are followed by, get this, an inductor called a choke. For a nice little chart of rectifier schemes and input schemes, check the Hammond transformer site. - Pat
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.