If a transistor doesn't sound like it should, the problem is probably operator error. Bad designer. *whack!*
Once a transistor of sufficient quality and suitability for the application is chosen, it'a all in the implementation. Any differences in "sound" between a C5200 and a MJL21194, for instance, are going to be very very very minor if you could even hear it at all.
Once a transistor of sufficient quality and suitability for the application is chosen, it'a all in the implementation. Any differences in "sound" between a C5200 and a MJL21194, for instance, are going to be very very very minor if you could even hear it at all.
Yeah it`s true but most of the h-power transistors are made just to put those db in and lower the quiality, that`s what happend practicly. For an example my marantz PM 440 sounds 10 times better that Technics Su-V90 D - and it`s not more than 60 w/per channel.
Also i think in our time we have too much copies ( originated from China or PRC ) and this is the main problem with this.
Louder mean worse in our case ?
Also i think in our time we have too much copies ( originated from China or PRC ) and this is the main problem with this.
Louder mean worse in our case ?
Different transistors sound very different in a certain design, much more than I would consider minor. You have to keep in mind that there are higher performence ones and run of the mill. One can change the sound of a amp by a change in transistors and even improve the messured performance by changing transistors. Changes in sound could offcourse also be made by changes in the design. If one looks at amps in the market youll see some that by design dont look like much but in fact sound very good as the design benefitted from high performance parts. A change of transitors usually also means slight changes will have to be made to a design, but its not always the case.
In my opinion japanese parts far outperform any european or american made parts especially when it comes to BJTs. For outputs look at tranies from Sanken or Toshiba and for small and medium signal Sanyo parts. Theres is also harder to get Hitachi/renesance or NEC parts avaialable that are extremely good. Sorry guys I feel parts from onsemi, motorola ect are run of the mill espesially for audio use. Onsemi make copies of japanese trannies to improve their range, look at all the variations that they make of a old toshiba part, the trusty and good 2sa1302 and complement. The performance gap is even bigger when it comes to medium and small signal devices.
In my opinion japanese parts far outperform any european or american made parts especially when it comes to BJTs. For outputs look at tranies from Sanken or Toshiba and for small and medium signal Sanyo parts. Theres is also harder to get Hitachi/renesance or NEC parts avaialable that are extremely good. Sorry guys I feel parts from onsemi, motorola ect are run of the mill espesially for audio use. Onsemi make copies of japanese trannies to improve their range, look at all the variations that they make of a old toshiba part, the trusty and good 2sa1302 and complement. The performance gap is even bigger when it comes to medium and small signal devices.
I agree categorically with Homemodder. It is obvious he has built and listened to many different designs and component mixes. Topology is the sexy issue, with contests drawn up daily in this forum to see who can produce the most incomprehensible schematic. But in truth, the standard Lin/Bailey/Self designs, with proper, fastidious parts selection and very careful selection of operating points will give absolutely outstanding results.
This is precisely what manufacturers do to produce a better amplifier. And the influence of current limiters to 'protect' the transistors has profound sonic impact, too. Better to select output devices with very high Vceo and superior SOARs, and then use simple means, such as fuses, to protect the speaker. It's not quite as effective as current limiting, but it sounds markedly better.
Furthermore, the issue of compensation (and phase lead) is another can of worms, seldom properly done. This is dependent upon layout, too, so PSpice is of limited use here. Time spent empirically figuring out the compensation is richly rewarded in good sonics.
Finally, too much speed/slew rate is bad, as we none of us have the ears of bats. LP filters should be placed on all amps beyond a certain speed to prevent intermodulation with local radio, TV and cellphone signals. This is very important, something I learned in practice and from PMA.
Japanese transistors are the best in my view too. And they can be had at low cost if you buy in SE Asia, particularly HK and Singapore.
I should add, whimsically, that I am not an engineer, and thus qualify as spectacularly incompetent. Pity really, my life has been wasted.....
Cheers,
Hugh
This is precisely what manufacturers do to produce a better amplifier. And the influence of current limiters to 'protect' the transistors has profound sonic impact, too. Better to select output devices with very high Vceo and superior SOARs, and then use simple means, such as fuses, to protect the speaker. It's not quite as effective as current limiting, but it sounds markedly better.
Furthermore, the issue of compensation (and phase lead) is another can of worms, seldom properly done. This is dependent upon layout, too, so PSpice is of limited use here. Time spent empirically figuring out the compensation is richly rewarded in good sonics.
Finally, too much speed/slew rate is bad, as we none of us have the ears of bats. LP filters should be placed on all amps beyond a certain speed to prevent intermodulation with local radio, TV and cellphone signals. This is very important, something I learned in practice and from PMA.
Japanese transistors are the best in my view too. And they can be had at low cost if you buy in SE Asia, particularly HK and Singapore.
I should add, whimsically, that I am not an engineer, and thus qualify as spectacularly incompetent. Pity really, my life has been wasted.....
Cheers,
Hugh
Janneman, sorry I dont truely understand your point.
You are an experienced designer and maybe do audio design for a living, dont you experiment with different trannies in a design to find the best possible sounding and mesureing part available??. Do you find that a bc, bd or 2n part sounds the same as a 2sa/2sc part or even measures the the same in a particular position of the design.
I find major differences. Building a simple diamond buffer with the same passive components as an example with bc or 2n is definetly not the same as building it with japanese trannies. Try it, not only do I find The japanese trannies version sounds better but it measures consideribly better too. Its not voodoo, looking at datasheets of the trannies tells most of the story, lower noise, better linearity, higher early voltage, lower capacitances etc.
I came from a country where it was eisier to get japanese components than eu/american parts and when I first came to this forum I saw some very good designs and my first thoughts were what it would sound like with decent japanese devices.
Imagine building a Porche by German design with african parts............ Do you think it would be the same as building it with german parts????
I find that most spectacular competent designers know this, and when I look at their designs I almost always find they using the best possible devices for a certain purpose and its always japanese in the case of BJTs, offcourse sometimes they have to make compromises because of price or availability. Ive given some tips to some here on component selection, mainly BJTs and all that have bothered to say thank you have written back to report excellent results, some are very competent designers from what i can tell from their designs and I suppose the problem was information and availabilty of these components. Some of these japanese devices I use are nearly 30 years old but still in production and vastly superior to anything western, and I still see bc547, bd140, 2n5401 being used, you gotta be kidding, well I suppose its a price or availabilty thing. I dont have enough experience with eu/american FETs so Ill wont speculate.
You are an experienced designer and maybe do audio design for a living, dont you experiment with different trannies in a design to find the best possible sounding and mesureing part available??. Do you find that a bc, bd or 2n part sounds the same as a 2sa/2sc part or even measures the the same in a particular position of the design.
I find major differences. Building a simple diamond buffer with the same passive components as an example with bc or 2n is definetly not the same as building it with japanese trannies. Try it, not only do I find The japanese trannies version sounds better but it measures consideribly better too. Its not voodoo, looking at datasheets of the trannies tells most of the story, lower noise, better linearity, higher early voltage, lower capacitances etc.
I came from a country where it was eisier to get japanese components than eu/american parts and when I first came to this forum I saw some very good designs and my first thoughts were what it would sound like with decent japanese devices.
Imagine building a Porche by German design with african parts............ Do you think it would be the same as building it with german parts????
I find that most spectacular competent designers know this, and when I look at their designs I almost always find they using the best possible devices for a certain purpose and its always japanese in the case of BJTs, offcourse sometimes they have to make compromises because of price or availability. Ive given some tips to some here on component selection, mainly BJTs and all that have bothered to say thank you have written back to report excellent results, some are very competent designers from what i can tell from their designs and I suppose the problem was information and availabilty of these components. Some of these japanese devices I use are nearly 30 years old but still in production and vastly superior to anything western, and I still see bc547, bd140, 2n5401 being used, you gotta be kidding, well I suppose its a price or availabilty thing. I dont have enough experience with eu/american FETs so Ill wont speculate.
Totaly agreed with you homemodder . Looking all the years at japanise amps and chineese - is like looking at well designed building , where you can do everything and it`s easy to be done - and - a village cotage with one door and no windows 
Most of the amps i have are at least 30 years old - and they sound like new ones ! even better
Most of the amps i have are at least 30 years old - and they sound like new ones ! even better
Hello homemodder
I have to agree with you, you can't build a Porche with african parts!
I recently had my ultimate experience in using japanese transistors
in a very well know design, the venerable JLH amplifier. I had built
my first one a few years ago and it always gave a lot of pleasure in
listening to music...
However, recently, I had to substitute the board for a more reliable
and smaller one so I decided to replace my BC's and BD's for a succession of japanese counterparts, including the outputs, Sanken
(4 in each channel).
I can't tell you the amazing difference I experienced, it sounds like a
different amplifier, the JLH had an intrinsic quality well know over
decades, but now it's a different amplifier. I used japanese transistors throughout, including 2SA970, 2SA1815 and finally the Sankens NPN
2SC3857.
Among different amplifiers I've built over the years this is the one
I consider true HiFi. Class A, of course!
Continue as suas considerações neste forum, são muito apreciadas...
I have to agree with you, you can't build a Porche with african parts!
I recently had my ultimate experience in using japanese transistors
in a very well know design, the venerable JLH amplifier. I had built
my first one a few years ago and it always gave a lot of pleasure in
listening to music...
However, recently, I had to substitute the board for a more reliable
and smaller one so I decided to replace my BC's and BD's for a succession of japanese counterparts, including the outputs, Sanken
(4 in each channel).
I can't tell you the amazing difference I experienced, it sounds like a
different amplifier, the JLH had an intrinsic quality well know over
decades, but now it's a different amplifier. I used japanese transistors throughout, including 2SA970, 2SA1815 and finally the Sankens NPN
2SC3857.
Among different amplifiers I've built over the years this is the one
I consider true HiFi. Class A, of course!
Continue as suas considerações neste forum, são muito apreciadas...
I agree categorically with Homemodder. It is obvious he has built and listened to many different designs and component mixes.
I’d say Jan and many others have also. Why insinuate otherwise?
Topology is the sexy issue, with contests drawn up daily in this forum to see who can produce the most incomprehensible schematic.
Some people are creative and have ingenuity.
But in truth, the standard Lin/Bailey/Self designs, with proper, fastidious parts selection and very careful selection of operating points will give absolutely outstanding results.
So? Outstanding results can be had from just about any topology given sufficient engineering competence and attention to detail.
This is precisely what manufacturers do to produce a better amplifier.
That’s pretty funny. The “high-end” market is in fact overwhelmingly dominated with marketing mumbo jumbo by just about every manufacturer (Krell, Halcro, Parasound, Dart-Zeel, etc, etc, etc) extolling the virtues of their “unique” proprietary or mysteriously tweaked topologies over all others.
Heck, that’s even what some around here do........
And the influence of current limiters to 'protect' the transistors has profound sonic impact, too. Better to select output devices with very high Vceo and superior SOARs, and then use simple means, such as fuses, to protect the speaker. It's not quite as effective as current limiting, but it sounds markedly better.
Wow, a quantum leap from component selection to current limiters.....
A current limiter only effects the sound when it operates of if it is incompetently implemented.
If the output stage of a HiFi amplifier is made powerful enough to drive a specified load without current limiting (which it should be)then there simply isn’t an issue WRT "sonics"
In such a case a conventional current limiter is perfectly acceptable and if superior protection is required the amplifiers protection circuitry can be designed to shut the amplifier down if excessive current is detected. This is far superior to just relying on fuses.
Furthermore, the issue of compensation (and phase lead) is another can of worms, seldom properly done. This is dependent upon layout, too, so PSpice is of limited use here. Time spent empirically figuring out the compensation is richly rewarded in good sonics.
An amplifers frequency compensation is absolutely critical to its correct function. Anyone who empirically “figures out” an amplifiers compensation is poking in the dark with little understanding of what they are doing and there is very little surprise that blind tweakers routinely report drastic influences on the sonic performance of their amplifiers.
Also, anyone serious about power amplifier building empirically optimises their amplifiers compensation in the prototype stage as there are many variables at play that are too difficult to predict. However, anyone with some degree of general analogue design competency and fluency in SPICE can tell you categorically that your assertion above is complete nonsense.
One of most versatile applications for SPICE in analogue design is the analysis of control loops and negative feedback systems. It can give a knowledgeable designer and excellent idea of just where to start before tweaking in the physical prototype stages for optimal results.
Finally, too much speed/slew rate is bad, as we none of us have the ears of bats. LP filters should be placed on all amps beyond a certain speed to prevent intermodulation with local radio, TV and cellphone signals. This is very important, something I learned in practice and from PMA.
You are confusing two completely separate issues and everyone bandwidth limits the input signal to their amplifier to some degree to suppress RFI. Why don’t you ask that John Curl fellow about “speed/slew rate” and his amplifiers.....
Japanese transistors are the best in my view too. And they can be had at low cost if you buy in SE Asia, particularly HK and Singapore.
When it comes to high power output devices the On-Semi RETs are amongst the best performing available (if not the best).
For one thing too many people here are bedazzled by bogus/misleading "60MHz" or similar fT marketing specs and ignore fT Vs Ic graphs..........
I’d say Jan and many others have also. Why insinuate otherwise?
Topology is the sexy issue, with contests drawn up daily in this forum to see who can produce the most incomprehensible schematic.
Some people are creative and have ingenuity.
But in truth, the standard Lin/Bailey/Self designs, with proper, fastidious parts selection and very careful selection of operating points will give absolutely outstanding results.
So? Outstanding results can be had from just about any topology given sufficient engineering competence and attention to detail.
This is precisely what manufacturers do to produce a better amplifier.
That’s pretty funny. The “high-end” market is in fact overwhelmingly dominated with marketing mumbo jumbo by just about every manufacturer (Krell, Halcro, Parasound, Dart-Zeel, etc, etc, etc) extolling the virtues of their “unique” proprietary or mysteriously tweaked topologies over all others.
Heck, that’s even what some around here do........
And the influence of current limiters to 'protect' the transistors has profound sonic impact, too. Better to select output devices with very high Vceo and superior SOARs, and then use simple means, such as fuses, to protect the speaker. It's not quite as effective as current limiting, but it sounds markedly better.
Wow, a quantum leap from component selection to current limiters.....
A current limiter only effects the sound when it operates of if it is incompetently implemented.
If the output stage of a HiFi amplifier is made powerful enough to drive a specified load without current limiting (which it should be)then there simply isn’t an issue WRT "sonics"
In such a case a conventional current limiter is perfectly acceptable and if superior protection is required the amplifiers protection circuitry can be designed to shut the amplifier down if excessive current is detected. This is far superior to just relying on fuses.
Furthermore, the issue of compensation (and phase lead) is another can of worms, seldom properly done. This is dependent upon layout, too, so PSpice is of limited use here. Time spent empirically figuring out the compensation is richly rewarded in good sonics.
An amplifers frequency compensation is absolutely critical to its correct function. Anyone who empirically “figures out” an amplifiers compensation is poking in the dark with little understanding of what they are doing and there is very little surprise that blind tweakers routinely report drastic influences on the sonic performance of their amplifiers.
Also, anyone serious about power amplifier building empirically optimises their amplifiers compensation in the prototype stage as there are many variables at play that are too difficult to predict. However, anyone with some degree of general analogue design competency and fluency in SPICE can tell you categorically that your assertion above is complete nonsense.
One of most versatile applications for SPICE in analogue design is the analysis of control loops and negative feedback systems. It can give a knowledgeable designer and excellent idea of just where to start before tweaking in the physical prototype stages for optimal results.
Finally, too much speed/slew rate is bad, as we none of us have the ears of bats. LP filters should be placed on all amps beyond a certain speed to prevent intermodulation with local radio, TV and cellphone signals. This is very important, something I learned in practice and from PMA.
You are confusing two completely separate issues and everyone bandwidth limits the input signal to their amplifier to some degree to suppress RFI. Why don’t you ask that John Curl fellow about “speed/slew rate” and his amplifiers.....
Japanese transistors are the best in my view too. And they can be had at low cost if you buy in SE Asia, particularly HK and Singapore.
When it comes to high power output devices the On-Semi RETs are amongst the best performing available (if not the best).
For one thing too many people here are bedazzled by bogus/misleading "60MHz" or similar fT marketing specs and ignore fT Vs Ic graphs..........
Whew! Glen got back to this before I did.
Let's not get too carried away here - we don't want you to end up in the 'bin again!
My personal experience with all this is that once sufficiently-rated and suitable components for the job are selected the things that influence the sound the most are all the things that are NOT typically put into a Spice simulation. (You could, but it would be a complete RF model including electromagnetic effects.)
Let's not get too carried away here - we don't want you to end up in the 'bin again!
My personal experience with all this is that once sufficiently-rated and suitable components for the job are selected the things that influence the sound the most are all the things that are NOT typically put into a Spice simulation. (You could, but it would be a complete RF model including electromagnetic effects.)
wg_ski is right offcourse their are a lot of considerations to be made in audio and a very important one to make is choice of components. Any design will benefit from better trannies. Look at Jmateus coment on a old but still good sounding soundin JLH design. But sufficiently-rated components dont give the best results subjectively or measured....... I still see bd140s used in vas stages, you gotta be kidding.
Roender part of the reason your amp sounds and measures so good, is the use of of about the best trannies on the market in the frontend of your design. I also see they are Japanese, I need say no more. Substitute the 2sa1360 with bd140s and fets to 2n types and come back and tell us the results, measured and sujectively. Your output trannies are basically a american copy of a old but still good Japanese part with a add on. Glen also uses these devices as well as farchild copies of some other excellent japanese devices KSA and KSC. BC550/BC560 are about the best non Japanese, not quite 2sa970 and I consider 2sa970 as a average small signal tranny, their are much better to be had from NEC and sanyo and Toshiba themselves. Some very important parameters are not mentioned in datasheets but have a very profound influence in Audio, Early voltage and turn/on/off speed, one will have to build some testing circuits to get values, but if you do youll find a even greater japanese advantage.
By the way your amp is an excellent design and I dont doubt for a second that your amp sounds as tremendous as you say it
does. The use of that cfp in the EF is very clever indeed, I first saw it in HK Av amp of about 7 years back and wondered about its performance soundwise. I am waiting for some FETS to further experiment that cfp but with bjt/fet combination. I have found that bjt/fet combination sound very good indeed.
Another example can be seen with fotios amp, he wanted to improve the rise time of his amp. A simple tip from jon curl on changing from 2n types for 2sa/2sc in the Ltp and improvement, he has some graphs on that thread to prove it.
Glen and Hugh Ill reseve the right to comment on some of your comments, not because I dislike or have anything against you or Hugh, but to avoid mudslinging. You and Hugh really dont get along do you? I have learned a great deal from both in these forum threads, its not pleasant seeing two competent designers going at each other.
Roender part of the reason your amp sounds and measures so good, is the use of of about the best trannies on the market in the frontend of your design. I also see they are Japanese, I need say no more. Substitute the 2sa1360 with bd140s and fets to 2n types and come back and tell us the results, measured and sujectively. Your output trannies are basically a american copy of a old but still good Japanese part with a add on. Glen also uses these devices as well as farchild copies of some other excellent japanese devices KSA and KSC. BC550/BC560 are about the best non Japanese, not quite 2sa970 and I consider 2sa970 as a average small signal tranny, their are much better to be had from NEC and sanyo and Toshiba themselves. Some very important parameters are not mentioned in datasheets but have a very profound influence in Audio, Early voltage and turn/on/off speed, one will have to build some testing circuits to get values, but if you do youll find a even greater japanese advantage.
By the way your amp is an excellent design and I dont doubt for a second that your amp sounds as tremendous as you say it
does. The use of that cfp in the EF is very clever indeed, I first saw it in HK Av amp of about 7 years back and wondered about its performance soundwise. I am waiting for some FETS to further experiment that cfp but with bjt/fet combination. I have found that bjt/fet combination sound very good indeed.
Another example can be seen with fotios amp, he wanted to improve the rise time of his amp. A simple tip from jon curl on changing from 2n types for 2sa/2sc in the Ltp and improvement, he has some graphs on that thread to prove it.
Glen and Hugh Ill reseve the right to comment on some of your comments, not because I dislike or have anything against you or Hugh, but to avoid mudslinging. You and Hugh really dont get along do you? I have learned a great deal from both in these forum threads, its not pleasant seeing two competent designers going at each other.
It's literally been years since I posted on this forum, and I didn't make many friends then. I don't expect I will now, but here goes...
Very few designers use 2 identical example boards to evolve a circuit, making changes and then doing an A/B comparison at every step. It isn't practical, especially if several listeners are to be used, as should ideally be the case. But going on the designer's auditory memory of a previous configuration is notoriously inaccurate. And then there is also the element of bias.
Many people who shell out large sums for 'better' componentry and then spend hours toiling to get everything built, want an improvement badly enough to convince themselves it is there. I can recall an example from my younger days when I sold hi-fi equipment part time.
One client had a very expensive system: Townsend Rock reference turntable, Spectral pre-amp, Krell KSA-150 amp, Acoustat speakers. We called on him to demo some new expensive interconnects, doing A/B comparisons all the way. At one point I mischievously hit the 'mono' switch on the Spectral. Nobody noticed. They carried on and on about the sound-stage and imaging and stereo separation... Needless to say I wasn't very popular when they found me out!
Besides the fact that I couldn't hear a difference between the cables, the 2 golden-ears couldn't tell the difference between mono and stereo!
So I would venture to say, that an assertion that transistor X sounds better than transistor Y is nonsense unless pedantically exhaustive, multi-listener A/B comparisons have been made, rigorously. And the same goes for cables, capacitors, resistors etc.
If no basic scientific comparisons have been made, such assertions are subjectively baseless. Simple.
Very few designers use 2 identical example boards to evolve a circuit, making changes and then doing an A/B comparison at every step. It isn't practical, especially if several listeners are to be used, as should ideally be the case. But going on the designer's auditory memory of a previous configuration is notoriously inaccurate. And then there is also the element of bias.
Many people who shell out large sums for 'better' componentry and then spend hours toiling to get everything built, want an improvement badly enough to convince themselves it is there. I can recall an example from my younger days when I sold hi-fi equipment part time.
One client had a very expensive system: Townsend Rock reference turntable, Spectral pre-amp, Krell KSA-150 amp, Acoustat speakers. We called on him to demo some new expensive interconnects, doing A/B comparisons all the way. At one point I mischievously hit the 'mono' switch on the Spectral. Nobody noticed. They carried on and on about the sound-stage and imaging and stereo separation... Needless to say I wasn't very popular when they found me out!
Besides the fact that I couldn't hear a difference between the cables, the 2 golden-ears couldn't tell the difference between mono and stereo!
So I would venture to say, that an assertion that transistor X sounds better than transistor Y is nonsense unless pedantically exhaustive, multi-listener A/B comparisons have been made, rigorously. And the same goes for cables, capacitors, resistors etc.
If no basic scientific comparisons have been made, such assertions are subjectively baseless. Simple.
homemodder said:
A transistor with a large nonlinear Cob that's on the order of the value of the miller cap ISN'T adequately rated for the job of VAS. If you need something with a 5pF Cob and a decade of Hfe linearity, you need something with a 5pF Cob and a decade of Hfe linearity. But any of them would do, and any differences would be very minor.
Glen,
You remain unable to put your points with any sense of grace or subtlety - I do not believe it is worth answering your points, except one.
You neither know, nor indeed are intererested, in any of my products, and your comments about ingenuity and creativity are merely childishly prejudicial.
I wonder if you treat everyone in your own family, who surely know less about amps than you, with the same contempt you treat those here who do not agree with you. I would hope not.....
Jan, with whom I get along fine, is well able to look after himself, and is probably exasperated you leapt to his 'defence'. It was a gentle chide, unlike your patronising, me-me diatribe, and I really feel you should think carefully about your troglodyte destruction of the tone of this forum.
You do not know everything, you still have much to learn, particularly about people, and your persona here is needlessly antagonistic.
Have a nice day.....
You remain unable to put your points with any sense of grace or subtlety - I do not believe it is worth answering your points, except one.
You neither know, nor indeed are intererested, in any of my products, and your comments about ingenuity and creativity are merely childishly prejudicial.
I wonder if you treat everyone in your own family, who surely know less about amps than you, with the same contempt you treat those here who do not agree with you. I would hope not.....
Jan, with whom I get along fine, is well able to look after himself, and is probably exasperated you leapt to his 'defence'. It was a gentle chide, unlike your patronising, me-me diatribe, and I really feel you should think carefully about your troglodyte destruction of the tone of this forum.
You do not know everything, you still have much to learn, particularly about people, and your persona here is needlessly antagonistic.
Have a nice day.....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.