I do agree that but each device has its own Ft and different FFTs like semisouth has a SIT transistor which was giving least distortion among any transistor available in the market so the above might have such characteristics..
drivers: scanspeak 8545k, tweeters ss aircirc 6600,
amp psu rcore, amp output stage simple pushpull
dac pcm1796
drivers: scanspeak 8545k, tweeters ss aircirc 6600,
amp psu rcore, amp output stage simple pushpull
dac pcm1796
here is some real input ....
P3a is running on my bench 5 years now and almost 12 hours a day
In these 5 years expect a zillion tests and combinations made to please the owner
Name it i ve done it
With in these tests is mix matching of all the parts included in the original schematic
So i have been spending lets say a period of one year keeping pretty much standard the rest circuit but playing around only with outputs .
Still i can give you my opinion but i have no means to explain why ....
2sa194-5200 sound muffled and rough even worst and in any other combination if used with 150xx drivers
fairchild 1943 5200 clones sound alike but still just a touch more rough ( tested with many drivers ) still other tests outside of your question proved that fairchild clones are more rugged than the toshiba
2sa1492 3856 sounded really nice far better than the 1943 c5200 and also be able to sound nice in a wide range of supply
1216-2922 combo sounded something in between the 1943-5200 and the 1492-3856 still an also for reasons i dont know the 1216-2922 combo started to sound better in higher voltage version 50+50 volts
MJW 1302-3281 was also very nice mine was made by Motorola (? ) and rumors say that these was re- mark of the original Toshiba 2SA1302 2SC 3281
from the above the best sound was the original Toshiba 1302-3281 by far
in the next section i ve done the same thing exactly with drivers and also mix matching drivers and outputs ( hell i ve been doing this for 5 years for all the parts of the amp ) in a 5 year run yes ( again ) the 1302-3281 was the best for reasons i cannot explain .
These are the tests i ve made and the results i have and as about SAP ..sorry no darlingtons for me ...and sorry also no mosfets in class AB ....
kind regards
Sakis
P3a is running on my bench 5 years now and almost 12 hours a day
In these 5 years expect a zillion tests and combinations made to please the owner
Name it i ve done it
With in these tests is mix matching of all the parts included in the original schematic
So i have been spending lets say a period of one year keeping pretty much standard the rest circuit but playing around only with outputs .
Still i can give you my opinion but i have no means to explain why ....
2sa194-5200 sound muffled and rough even worst and in any other combination if used with 150xx drivers
fairchild 1943 5200 clones sound alike but still just a touch more rough ( tested with many drivers ) still other tests outside of your question proved that fairchild clones are more rugged than the toshiba
2sa1492 3856 sounded really nice far better than the 1943 c5200 and also be able to sound nice in a wide range of supply
1216-2922 combo sounded something in between the 1943-5200 and the 1492-3856 still an also for reasons i dont know the 1216-2922 combo started to sound better in higher voltage version 50+50 volts
MJW 1302-3281 was also very nice mine was made by Motorola (? ) and rumors say that these was re- mark of the original Toshiba 2SA1302 2SC 3281
from the above the best sound was the original Toshiba 1302-3281 by far
in the next section i ve done the same thing exactly with drivers and also mix matching drivers and outputs ( hell i ve been doing this for 5 years for all the parts of the amp ) in a 5 year run yes ( again ) the 1302-3281 was the best for reasons i cannot explain .
These are the tests i ve made and the results i have and as about SAP ..sorry no darlingtons for me ...and sorry also no mosfets in class AB ....
kind regards
Sakis
thank you sakis...
I think andrew there are things which work in time domain and its all about aural evaluations... we have seen sonic variation with very slight changes in the component selections. Anyway sankens transistors seems pretty good overall I have even auditioned the MBL Amps they sounded nice checked and found they use the Sanken big transis..
I think andrew there are things which work in time domain and its all about aural evaluations... we have seen sonic variation with very slight changes in the component selections. Anyway sankens transistors seems pretty good overall I have even auditioned the MBL Amps they sounded nice checked and found they use the Sanken big transis..
I used many of the 2SC3281/SA1302 clones, but the best I found is the NJW0281/0302 pair. I think that the driver is almost as important as the output device. The NJWs sounds quited dead with the MJE150xx drivers, but the marriage with 2SC5171/SA1930 are very happy.
The Sankens are better little bit, but the price (from Farnell) is terrible high. My favorite is the 2SC3519/2SA1386, I feel that they are as good as the 2922/1216, with lower price, and if You don't want to build real high power amplifier, You can use them instead of MT200 devices.
Sajti
The Sankens are better little bit, but the price (from Farnell) is terrible high. My favorite is the 2SC3519/2SA1386, I feel that they are as good as the 2922/1216, with lower price, and if You don't want to build real high power amplifier, You can use them instead of MT200 devices.
Sajti
I don't think so, because I start a similar thread some times ago with the actually question:
Where are basically articles and papers concerning various internal technologies for the emitter structure ? - go to
http://www.diyaudio.com/forums/soli...istor-families-audio-power-output-stages.html
Last edited:
Here is a small real life example
Last night i had to repair a Rotel RB 830 typical amp with 100uf capacitor in the feedback chain ...there rod says : you may add a 100nf plastic capacitor if it pleases you but don't expect much to change ...
well Rotel in the specific amplifier uses a MKP capacitor bypassing a horrible quality 6.3V/100uf can anybody explain why ?
Then again similar practice goes to Luxman or Marantz equipment that selectively use type A capacitors in the signal path/compensation or type B in other applications .Obviously that is done to reduce the cost BUT also because it serves the sonic signature that the designer wanted to have .
So expect that they 've done the exact same choices with transistors to come up with the specific results or the opposite adjusted the circuit with proper compensation and miller and filtering to achieve the wanted sonic result ...
so transistors do sound different any way you slice it the point is how much of this is audible in an other wise same circuit of amplifier ....
Kind regards
sakis
Sorry. What I meant was this....You have, through years of endeavour, decided that A is better than B which is better than C etc etc. These tests would have been done in a particular sequence, perhaps by chance, as I presume you had no pre-conceived ideas as to which may be best. Eventually you came up with your preferences. Did you then compare your top choice with earlier rejected lower order ones to re-confirm your decision, preferably after a period of time has elapsed so as to get that chance again of 'first impressions'?
yes ... i actually had ...
since original 2SA1302 2SC 3281 are no longer really available though i have still about 100pcs closely matched pairs direct order from Toshiba dated 1996( which i am planing to use all for personal amusement )
and since i have produced a few P3A amplifiers which one was made for a friend some others for costumers and so on of course there i went with 1943-5200 in order to keep my fleet of originals i had the chance to verify my results once more
yet again i have produced other amplifiers that i have made similar tests with similar results but not extensive next to the P3A that i ma working with currently ...
For the record students in my back yard produced a P3A in a vero board for a school project and used carbon resistors ceramic capacitors 3055-2955 outputs at 24+24 rails Forget the power but it sounded and measure simply horrible...
kind regards
sakis
since original 2SA1302 2SC 3281 are no longer really available though i have still about 100pcs closely matched pairs direct order from Toshiba dated 1996( which i am planing to use all for personal amusement )
and since i have produced a few P3A amplifiers which one was made for a friend some others for costumers and so on of course there i went with 1943-5200 in order to keep my fleet of originals i had the chance to verify my results once more
yet again i have produced other amplifiers that i have made similar tests with similar results but not extensive next to the P3A that i ma working with currently ...
For the record students in my back yard produced a P3A in a vero board for a school project and used carbon resistors ceramic capacitors 3055-2955 outputs at 24+24 rails Forget the power but it sounded and measure simply horrible...
kind regards
sakis
how many treads are rhythmsandy going to start, that leves nowhere?
it is 87 threads no.
http://www.diyaudio.com/forums/search.php?searchid=5450674
it is 87 threads no.
http://www.diyaudio.com/forums/search.php?searchid=5450674
Well, that's interesting.
Would you be so kind to provide measurements of 2 P3, one using humble 2N3055/2955 vs. an exact same one using "audiophile" transistors?
Please, one variable at a time, let's not change drivers/capacitors/resistors/PCB/PSU/etc. so as not to pollute the comparison.
I accept your subjective tests ("horrible"), but not being in the same listening room as you (I'd love to visit Greece again
), I'd settle for the second best, a nice set of numbers.Thanks.
my version of P3A will make 40KHZ square wave on the scope flawless a variant with 2955-3055 will make no more than 16 before rounded up ( though this test was done 4 years ago and i have to make a lot of digging to find the measuring set of that long ago so i might be wrong )
For you pleasure though with in the next days i will upload a set with a couple of variants
All these have nothing to do with listening tests ,... that comes after measuring ...always
For you pleasure though with in the next days i will upload a set with a couple of variants
All these have nothing to do with listening tests ,... that comes after measuring ...always
yeah useless even the SemiSouth SIC transistors are also waste...
why was that Transistor sounded so good and why does Nelson has to collaborate with Semiconductor companies to make a new Transistor which no one ever find difference at all..
why does chord uses its custom built output devices...
why does all this happen when one is getting everything off the shelf.. when you say its waste thread it does really reflects how how much one is behind the material science and great insult to the leads in the audio industry...
why was that Transistor sounded so good and why does Nelson has to collaborate with Semiconductor companies to make a new Transistor which no one ever find difference at all..
why does chord uses its custom built output devices...
why does all this happen when one is getting everything off the shelf.. when you say its waste thread it does really reflects how how much one is behind the material science and great insult to the leads in the audio industry...
Ok, thanks.
If you find them, fine.
If it complicates your life .... don't worry.
At least, instead of a word (horrible) which by itself means nothing, now you qualify by saying "one passes flawless 40KHz squarewaves, the other starts rounding 16KHz ones"
Well, yes, that's a difference, no doubt, but it usually means the amp reaches at least flat 30 or 40 KHz (sinewave) and often beyond that.
Not bad at all.
Qualifies for "horrible"? Mmmmm, who knows?
What I'm sure is that it will take for a lot of *flawless* equipment to notice it by itself.
Good luck