P3A-More upgrades

Member
Joined 2010
Paid Member
.......Could it be that shifting bias level up (relative to optimal) serves as some kind of subtle tone control, a kind of aural exciter?.......
I think you have it there. In fact, much tweaking we perform for subjective reasons is all about tinkering with the precise spread of harmonic distortion and response that is most pleasing to us, either for our personal music and tonal preferences or a certain demographic group like 25-55 year old males or more universal tastes, if we are manufacturers for wider markets.

These personal touches are often just interesting as a temporary sensation that becomes fatiguing over time but occasionally, someone designs an amplifier with a SQ that has more universal appeal and we get popular designs with a good reputation like this P3a. Most guys probably already know that great sounding amps are not about lowest THD but don't want to have pointless losing arguments about whether this is best conventional design or engineering practice.

It's not best practice but when was that a problem for DIY audiophiles? ;)
 
I need to and read my Self book again - thanks for giving me some new stuff to think about. I do find the CFP an interesting option - just don't see much talk about it these days. My perception has been that it doesn't sound as nice as the double EF but I've never built a CFP into the output myself, only as a driver stage or in the input stage. In the input stage it was a little harsh. In the driver stage it was fine but required compensation to stabilize. I know it's difficult to ***** a topology when the devil is in the details of the implementation but I do find it useful to hear about what other people think from their experience. Any more comments on the sound of CFP output ?
 
I need to and read my Self book again - thanks for giving me some new stuff to think about. I do find the CFP an interesting option - just don't see much talk about it these days. My perception has been that it doesn't sound as nice as the double EF but I've never built a CFP into the output myself, only as a driver stage or in the input stage. In the input stage it was a little harsh. In the driver stage it was fine but required compensation to stabilize. I know it's difficult to ***** a topology when the devil is in the details of the implementation but I do find it useful to hear about what other people think from their experience. Any more comments on the sound of CFP output ?

their good in the bass department....
taming them is another thing, unexplained heating of the output trannies despite correct voltages when checking....
that made me abandon the CFP and just use the darlington hence...:D
i might revisit it one day though....;)
 
Member
Joined 2010
Paid Member
:D
The data I referred to was in table 6.5 of Self's current (5th) ed. It is different in earlier eds like the 3rd which I seem to have lost in my jumble of a library but the basics are there. Randy obviously cribbed and acknowledged a lot of his material was inspired by Self. He at least made the point about EF2 sounding better at low levels, as you suggest, because the crossover region is huge, like 9V compared to 0.4V for CFP but all this depends on the bias level chosen and the specific compromise of crossover distortion spectrum V level.

We can see why designers like EF triples - think of a crossover region that means virtual class A operation at personal listening levels, without the label of a nasty current hog. High end, high bias AB designs just take the compromise a little further into the power range, based on EF designs.

For other CFP comments and different design, take a look at Calvin's review in LJM's L12 thread - quite surprising. :cool:
 
Last edited:
output transisitors for P3A

Hi sakis ,

I quote from the first post

VAS /DRIVERS
been trying many excluded from the process MJE series since sounded worst eventhough pretty rugged device, and realized that this process is seriously related to the choice of outputs ....For the record the best combination ever was the BD829-830 paired with 2SA1302-2SC3281...BD 139-140 with 2SA1943-2SC5200... obviously different combinations will produce different sonic signature.

Thanks for the posts regarding P3A. Your posts inspired me to build the amp and i bought the pcb from Rod.

Can you elaborate on the sound signature of the BD 139-140 with 2SA1943-2SC5200.combination ?
Is this a drop in replacement ?
I have TTA 1943 and TTC 5200 and are these interchangeable with 2SA 1943 and 2SC 5200?
thanks

kp93300
 
ΟΚ ..if i am allowed i will try to summarize this if possible ...

As said in many threads i am in the repair business and i repair amplifiers 400-600 per year ..of any kind ...as about listening and construction i only work with the P3A

Often what i do is give as a replacement amplifier to costumers one of the P3A i created for my shelf ..Don't fool your self this has not much to do with how service minded i am but mostly when i choose experienced listeners with sophisticated equipment to actually provide me information , test between A to B , and general feedback

now if i was able to post the comments of the costumers on the specific amplifier it will probably take a couple of days to write

Been working with almost any possible combination of supply , bias , and semis and posted my results ...for the record most people seem to like the 70 ma of bias been fooling them around with lower and higher versions but the favorite was 70ma .

For the record also the ppl that loved the P3A was previously owners of NVA QUAD AUDIOLAB GOLDMUND SANYO HAFLER from the big names and a huge list of other costumers owned consumer brands like technics pioneer Jvc AKAI and so on .

Also here don't fool your shelf it is not the P3A that is better than a JVC it is the amplifier construction that is 1000+ times better than a consumer amplifier so yes that test is not fair ....

That was my input
 
to answer the specific question the combo of matched BD829 830 and 1302-3281 original outs will sound perfect to my ears and i can describe it like a perfectly balanced 2Way speaker that produces everything in a perfect balance ...

the BD 139 140+1943 5200 will sound like a 3 way speaker and while both ends low and hi will sound proper it will be some dirt in very high pitch 6K and above and also sound like there is a small gap between 4-6K ...

kind regards
Sakis
 
.................Now I put the potentiometer after B1 (I believe CDP performance has small effect). The result is better than putting the potentiometer in front of B1. And the best result is for 20K potentiometer.
I wonder if this conclusion has been influenced by the filtering effect of the 5k output impedance of the 20k vol pot?
As the volume is turned up from zero, where the Zout is also zero, the Source Resistance (Rs) rises and thus the roll off of HF increases as volume increases. This has the effect of reducing the clipping of the louder HF signals riding on the LF signal waveform.

The follow up response from Sakis
well ... The B1 has less dynamics in this level than any preamplifier with a bit of gain Also to my ears the B1 tends to round a bit the ends of the graph
may well also be influenced by high Rs of an unbuffered volume potentiometer.
 
Last edited:
Hi Sakis
Many thanks for the clear reply.
If my music preference is jazz and vocal, what combination would you recommend?
I would be most grateful if you can post a picture of your P3A especially with regard to the layout. My electronic skill is basic and i can only copy from more experienced diyer .
I have high expectation in this amp after building the DOZ.
My speaker has a powered sub and so bass response is not so important .

thanks

kp93300
 
...I would be most grateful if you can post a picture of your P3A especially with regard to the layout. My electronic skill is basic and i can only copy from more experienced diyer...
You posted you have the boards from RodE, have you built those yet? You can get a good idea from building the amp according to Rod's directions, before you start making your own changes. It is much easier to have a circuit on a known working board to start, and to listen to, before making upgrades, and Sakis has left plenty of suggestions for component substitutions to try out.
 
You posted you have the boards from RodE, have you built those yet? You can get a good idea from building the amp according to Rod's directions, before you start making your own changes. It is much easier to have a circuit on a known working board to start, and to listen to, before making upgrades, and Sakis has left plenty of suggestions for component substitutions to try out.

HI
I am sorry for the misunderstanding.
I ask for a picture is to try to get an idea of component choices and power supply arrangement rather than to modify the circuit and make a pcb.
I am sure Rod board will sound fine and more than meet my needs..
On the other hand , I respect sarkis decision and intellectual property.
thanks

kp93300
 
HI
I am sorry for the misunderstanding.
I ask for a picture is to try to get an idea of component choices and power supply arrangement rather than to modify the circuit and make a pcb.
I am sure Rod board will sound fine and more than meet my needs..
On the other hand , I respect sarkis decision and intellectual property.
thanks

kp93300
I think Sakis has been very generous in posting his findings for everyone, especially since he is running a business, and he is really trying to strike a balance and respect Rod's intellectual property.

If you can be a little more specific about the power supply, I will post some pics for you in a few days (no time to disassemble mine right now). Pics of power supplies and wiring break no rules here, AFAIK... :rolleyes: It may take a week for me to get to it though.

In brief, this is what I did when I built mine:

I cut the single board in half, per one of the suggestions in Rod's directions. Then I added a decoupling cap to each board to balance the decoupling capacitance on the two rails (this is lost when the boards are cut).

The boards are mounted flat on a heatsink, with the power connectors toward the bottom of the board.

The transistor Sakis mentioned is bent/glued to one of the VAS power transistors (this is a pretty universal suggestion).

I started with ceramic Miller caps and later changed them to silver mica.

I did not add a common VAS transistor heatsink, they get only slightly warm at low power, which is how I am using the amp.

Bias is adjusted at the low-ish end of the range that has been discussed here (can't remember the exact value right now).

I am using BD139/140 and ON Semi MJL4281A/MJL4302A. (The latter is overkill for low power listening, but I had them already).

I tried 2x4700 uF per rail, 3x4700 uF per rail, 10mF+4700uF per rail, and CRC, CLCRC, and CRCRC filter. 3x4700 uF or 2x4700uF in CRC or 3x4700uF CRCRC configuration seems to work best for me. CLC seemed not to make much difference to sound. Heavy bass is not an issue for me, so your mileage may vary on that.