There is a strong case with simple amplifiers, for ignoring the ultimate in numbers and focusing on best subjective sound quality instead. When you are restricted to low parts count, simplest possible designs like P3, that has to be a no-brainer. According to the builders posting here who have simply copied the website P3a design and followed the photographic clues, most are still more than happy with the results regardless of the numbers. The OP is a very happy guy, believe me.
The problem with reducing existing distortion of any kind is that you are throwing out the baby with the bathwater, so to speak. Sometimes, in notoriously crap sounding products, that's a good idea. However, doing that to DIY designs like P3a, Aksa 55 or JLH 69 for example, is almost certainly a bad idea, whatever your views on distortion and whether it should be eliminated or not as a matter of principle.
The problem with reducing existing distortion of any kind is that you are throwing out the baby with the bathwater, so to speak. Sometimes, in notoriously crap sounding products, that's a good idea. However, doing that to DIY designs like P3a, Aksa 55 or JLH 69 for example, is almost certainly a bad idea, whatever your views on distortion and whether it should be eliminated or not as a matter of principle.
Hello , I did a test with theses resistors reduced from 220 to 100 ohms.. with bias at 100ma no much improvement on IMD (6.5khz - 7khz). next test I will try 10 ohms... however drivers will need a good heatsink.
Thanks
I don't think that going down to 10R will improve anything, including specs. Do not waste your energy trying to force this simple and elegant amp to become stellar performer specswise. The whole idea of P3A is to get reliable, good sounding amp from the simple and elegant circuit with minimum number of parts. If you are not satisfied with the result of your build, it's better to try another circuit.
I think that 100R will be just fine without risk for the drivers. Increasing output transistors bias and increasing gain of the amp will change the sound more.
I think that 100R will be just fine without risk for the drivers. Increasing output transistors bias and increasing gain of the amp will change the sound more.
Increasing the current through the drivers is one idea but still half the picture There a few things more to be done to achieve the target of lower IMD from this tweak only ..
Remember that transistors are like bells Melody will come if these ring altogether....
If this is done, another study is to be made to evaluate stability margins from scratch together with a few other tests like slew rate for example ....
Kindest regards
Sakis
Remember that transistors are like bells Melody will come if these ring altogether....
If this is done, another study is to be made to evaluate stability margins from scratch together with a few other tests like slew rate for example ....
Kindest regards
Sakis
As a small update (of the funny ones )
Client owns a LAB Goupen LP 2100 analog one the old models, well kept using it as a home amplifier with a tiny bit of bias increased .
Got it in for noisy ventilators replacement
While we searched for a ventilator that is quiet enough for the application ( home use ) so the vent need to be quiet on its own but also quiet in the specific ventilation channel we gave the client a P3A as a replacement amplifier.
The first night client realized that the P3A sounded 10 times faster than the Lab Groupen extremely wider ( probably in terms of bandwidth ) and out of the blue he realized that his speakers ( Dynacord VL262) Are no longer up to the task ....
Second night client decided to go and listen the CHario Cygnus speakers and at the shop they tested with advance acoustic Xi 160....
Third night client took the P3A with out my permission to the shop to test with the Chario ....
Beat the crap out of the advance acoustics from sound quality aspects though the Advance was a bit more powered while still the P3A with 60.000 ufd had no drive issues at all and left both the client and the shop owners .... breathless /......
Enjoy
Client owns a LAB Goupen LP 2100 analog one the old models, well kept using it as a home amplifier with a tiny bit of bias increased .
Got it in for noisy ventilators replacement
While we searched for a ventilator that is quiet enough for the application ( home use ) so the vent need to be quiet on its own but also quiet in the specific ventilation channel we gave the client a P3A as a replacement amplifier.
The first night client realized that the P3A sounded 10 times faster than the Lab Groupen extremely wider ( probably in terms of bandwidth ) and out of the blue he realized that his speakers ( Dynacord VL262) Are no longer up to the task ....
Second night client decided to go and listen the CHario Cygnus speakers and at the shop they tested with advance acoustic Xi 160....
Third night client took the P3A with out my permission to the shop to test with the Chario ....
Beat the crap out of the advance acoustics from sound quality aspects though the Advance was a bit more powered while still the P3A with 60.000 ufd had no drive issues at all and left both the client and the shop owners .... breathless /......
Enjoy
Member
Joined 2009
Paid Member
Actually, increased current through the drivers minimizes cross-conduction through the CFP output pairs. Not saying it's a problem with P3a as originally designed as it's passed the test of time but it's a general comment about this output topology for those interested in it.
Member
Joined 2009
Paid Member
Using this type of outputs its a total waist of parts .... Using lower rails than specified will have also some impact on the current flowing on the first stages
Rod may says that amplifier may start to operate from 12+12 volts but this doesn't mean that will also have a reasonable quality
And NO i have been listening to it and it plays fine is not a proper answer to the above comment .
Happy regards
Sakis
Rod may says that amplifier may start to operate from 12+12 volts but this doesn't mean that will also have a reasonable quality
And NO i have been listening to it and it plays fine is not a proper answer to the above comment .
Happy regards
Sakis
I disagree. The purpose of this build was to see if slower output transistors will work without oscillations with parts values from the original schematics. Based on my experiences, I would recommend to builders to use such transistors, especially if they don't have scope to verify the results. Fast Japanese transistors are not recommended if the builder is not experienced in modifying the schematics and if does not have scope. TIP35/36 (for 35V supplies) and MJL21193/4 (for 42V supplies) are perfect candidates.
The P3 began its existence with slow 2N3055 devices many years ago, so there is little doubt it can be made to work fine with other general purpose power transistors such as TIP35/6. The only question is, just how fine in audiophile terms, is the SQ going to be?
There is a lot of difference between the linearity of excellent modern devices like MJL21193/4 and older devices like the 2N3055 or TI's family of general purpose, TIPXX power devices dating from the 1970's. These have been improved slightly over the decades and you can certainly get good sound from these poor linearity devices but when they are upgraded to more linear types, I am usually a lot happier with the cleaner, sweeter sound. However, that does not imply that also being faster devices had much to do with it.
There are not many widely available, medium power transistors with good gain linearity but On-Semi's MJE1503X familiy of TO220 drivers will deliver up to 8A and are very good for small-medium power amps. These would be my preference for personal use, as the MJL series parts are a tad expensive for the extra and probably unnecessary power capability.
At an intermediate level of improvement, there is at least one manufacturer of good, inexpensive copies of old faithfuls like 2SD1047/B817. There are also several affordable and amazingly good LAPTs or even conventional types from Sanken's lineup that will really kick butt in the P3a too.
I suggest though, that there is no single standard or consensus of amplifier SQ that all audiophiles will agree to, globally. Like any personal interest, there are pockets of guru or expert consensus in various countries that will all tell a different story on their preferences, whatever the experience and influences. Japanese audiophiles, for example, generally have quite different preferences to mine and to other folk I compare notes with.
There is a lot of difference between the linearity of excellent modern devices like MJL21193/4 and older devices like the 2N3055 or TI's family of general purpose, TIPXX power devices dating from the 1970's. These have been improved slightly over the decades and you can certainly get good sound from these poor linearity devices but when they are upgraded to more linear types, I am usually a lot happier with the cleaner, sweeter sound. However, that does not imply that also being faster devices had much to do with it.
There are not many widely available, medium power transistors with good gain linearity but On-Semi's MJE1503X familiy of TO220 drivers will deliver up to 8A and are very good for small-medium power amps. These would be my preference for personal use, as the MJL series parts are a tad expensive for the extra and probably unnecessary power capability.
At an intermediate level of improvement, there is at least one manufacturer of good, inexpensive copies of old faithfuls like 2SD1047/B817. There are also several affordable and amazingly good LAPTs or even conventional types from Sanken's lineup that will really kick butt in the P3a too.
I suggest though, that there is no single standard or consensus of amplifier SQ that all audiophiles will agree to, globally. Like any personal interest, there are pockets of guru or expert consensus in various countries that will all tell a different story on their preferences, whatever the experience and influences. Japanese audiophiles, for example, generally have quite different preferences to mine and to other folk I compare notes with.
Some of the best sounding amps that I ever heard were not high power and used very slow conventional devices. First amp that comes to mind is NAD 3020, although it was long ago that I listened to it last time. I was never able to find direct relation of SQ with specs of parts used. Very often I find "inferior" parts better sounding. At the end, it all comes to how your amp matches with you speakers. I am using Tannoy professional studio monitor speakers with compression horn HF driver and it's absolutely amazing what can give the best subjective results with them. The other day I built miniature amp with TDA 2005 (bridge version) and small 30VA mains transformer. I was surprised how good it sounded with my speakers. Better than many high end amps.
The main reason that I used Toshiba 5200/1943 pair for my build using original ESP pcb was that they are high power. My main concern was reliability, not the SQ. But, it proved to be problematic concerning stability. They certainly sound excellent in P3A if you can verify stability and correct compensation if necessary.
The main reason that I used Toshiba 5200/1943 pair for my build using original ESP pcb was that they are high power. My main concern was reliability, not the SQ. But, it proved to be problematic concerning stability. They certainly sound excellent in P3A if you can verify stability and correct compensation if necessary.
You would be surprised to see where TIP35/36 are used. My friend and neighbor sold Lavardin IS Reference integrated amp couple of weeks ago. Look how much this amp costs in Australia! And yet it uses ST TIP 35/36 at the output. We are talking about very expensive audiophile product here.
Attachments
