Edit time gone :
Why do that ...as about the ground issues ...learn to produce and install audio circuits properly and you will never see this problem in your life again ...Learn it the hard way which is only one and its called PA applications where every Tom Dick and Harry will mix match and connect anything with everything .
Sorry G can i be mean for a second ?
Ok thanks
This one is for Alex MM with no hard feelings but with a meaning that produces most of his pcb in the same model ...The topology and equal distribution symmetries as i like to call them can and will have an impact in the sound ....Ground loops may also introduced from a badly made pcb , or worst from an average PCB and with a bad wiring or arrangement.
None of the above can be predicted by simulators and PCB software tools ... real life testings will only tell
Happy regards
Sakis
Why do that ...as about the ground issues ...learn to produce and install audio circuits properly and you will never see this problem in your life again ...Learn it the hard way which is only one and its called PA applications where every Tom Dick and Harry will mix match and connect anything with everything .
Sorry G can i be mean for a second ?
Ok thanks
This one is for Alex MM with no hard feelings but with a meaning that produces most of his pcb in the same model ...The topology and equal distribution symmetries as i like to call them can and will have an impact in the sound ....Ground loops may also introduced from a badly made pcb , or worst from an average PCB and with a bad wiring or arrangement.
None of the above can be predicted by simulators and PCB software tools ... real life testings will only tell
Happy regards
Sakis
Member
Joined 2009
Paid Member
I didn't check specifications but I believe high quality input transformers can have very wide bandwidth if amplifier input has a suitable impedance. I'm not arguing in favour of it, but only suggesting that there may be a solution that is the least harmful to the sound for somebody who insists on balanced input.
But with pre-amp, I find I am always considering a preamp and amplifier as one-system and therefore, the P3a with pre-amp is no longer a P3a but has become something new. So when you have found the optimal system set up over all these years, you must have found what kind of preamp is needed. It says to me that you can't run P3a with higher gain without preamp or it changes the distortion profile away from the optimum. What if you could make a P3a with higher gain that still sounded good and then a preamp is not needed ?
But with pre-amp, I find I am always considering a preamp and amplifier as one-system and therefore, the P3a with pre-amp is no longer a P3a but has become something new. So when you have found the optimal system set up over all these years, you must have found what kind of preamp is needed. It says to me that you can't run P3a with higher gain without preamp or it changes the distortion profile away from the optimum. What if you could make a P3a with higher gain that still sounded good and then a preamp is not needed ?
This is done in the past in classic vintage amplifiers but the principal is to preserve a very high input Z so there is going to be no loss in dynamics in low listening levels ...Or at lest less problems with that kenwood KA5700 is made like that KA 305 also but it should be more ...
Problem was and always been the pot in the input
One of the best amplifiers ever made the Sansui AU 717 has a unique pot and preamplifier configuration ... look at it it might explain a lot ...
been trying a P3A modified for input Z of 47K it didn't made any stability issues but still the change in dynamics was marginal .
I may as well check the original schematic of the P3a and find out what the original input R was ...as far as you know there is no parthenogenesis for Class AB circuits everybody was thinking alike or copy from each other .
as about the preamp s you are correct each and everyone of them will make a different pair with the P3A with plus and minus
Kind regards
Sakis
Problem was and always been the pot in the input
One of the best amplifiers ever made the Sansui AU 717 has a unique pot and preamplifier configuration ... look at it it might explain a lot ...
been trying a P3A modified for input Z of 47K it didn't made any stability issues but still the change in dynamics was marginal .
I may as well check the original schematic of the P3a and find out what the original input R was ...as far as you know there is no parthenogenesis for Class AB circuits everybody was thinking alike or copy from each other .
as about the preamp s you are correct each and everyone of them will make a different pair with the P3A with plus and minus
Kind regards
Sakis
Last edited:
Member
Joined 2009
Paid Member
I have them all plus extremely deep knowledge around 317-517-717 problems let me know if you need a hand
The original basis of the P3A or a similar circuit that was presented in the 70's had an input of aprox 38K but it was driven by an also very "'familiar "" buffer ....
So seems that someone has been there before ...
The original basis of the P3A or a similar circuit that was presented in the 70's had an input of aprox 38K but it was driven by an also very "'familiar "" buffer ....
So seems that someone has been there before ...
list is getting populated it seems that Sony made also an amplifier that is also szikla based in the output its the sony TA -F3A
Its a lookalike the P3a and the circuit is quite similar ( schematic available )
It seems that almost every of the known "big " manufacturers made his effort with a CFP output stage
Kind regards
Sakis
Its a lookalike the P3a and the circuit is quite similar ( schematic available )
It seems that almost every of the known "big " manufacturers made his effort with a CFP output stage
Kind regards
Sakis
Hi all!
While working with my VSSA amp project, my enthusiasm has brought me to do a revision of my earlier attempts on designing my P3A boards, attached are two of my recent designs, one for BCs and the other for 2S (input diff trannies). Question though, do they look 0K?
Many thanks!
While working with my VSSA amp project, my enthusiasm has brought me to do a revision of my earlier attempts on designing my P3A boards, attached are two of my recent designs, one for BCs and the other for 2S (input diff trannies). Question though, do they look 0K?
Many thanks!
Attachments
Member
Joined 2009
Paid Member
Member
Joined 2009
Paid Member
so it is - I missed that at first look !
Well I didn't check out if it's all correct, doesn't your pcb layout software do that for you ? - so long as the schematic is correctly entered it should be fine. Just for fun, you could give yourself the option of adding a lead compensation capacitor, a small (e.g. 5pF to 15pF) high quality cap between the VAS collector and the feedback node. In other words, a silva mica (or np0) cap between base of Q5 (or Q6) and the base of Q2. This allows you to reduce the value of C4 (e.g. to 47pF to 68pF) which you may find on listening tests can improve the sound.
Strange that Rod didn't put a cap (e.g. 22uF electrolytic) across the Vbe multiplier, this is usually to be seen.
I didn't build the original P3A but instead made my own version (TGM8) which has some differences but I think the sound you will have from your design will be very nice!
Last edited:
^ I'm using Sprint Layout so no schematic capture function. My earlier builds (different pcb design) had <25dcmv offset, no turn on and off at power on bass was good but I find it a bit too much on the midrange. Thanks, for the lead compensation tip and yes I've seen a few local builds in my area with an additional small e-cap across the vbe multiplier. I've seen your works on TGM8 it must've sounded really good, the circuit looks sophisticated. 
My experience with the original P3A was that VAS transistor gets much warmer than the driver transistor, hence I have provided a small space for a small heatsink to be attached on board.
Many thanks!
and yes I've seen a few local builds in my area with an additional small e-cap across the vbe multiplier. I've seen your works on TGM8 it must've sounded really good, the circuit looks sophisticated. My experience with the original P3A was that VAS transistor gets much warmer than the driver transistor, hence I have provided a small space for a small heatsink to be attached on board.
Many thanks!
Last edited:
Though particular builds and layouts don't have much to do with Sakis' thread, I would like to add some support for the layout on the right. It has the potential improvement of keeping the supply rails and fuses away from the input and low level stages, which is not ideal in the left side example.
Too often, the fuses and rails are the dominant feature of the PCB, taking long paths around the PCB on small amplifiers and this is not good for rail-induced distortion and noise, particularly for a design that has low PSRR to begin with. If I started a small new design, the fuses would be on the power supply board, since on the forum amplifiers I've assembled so far, that makes at least a measurable improvement.
Too often, the fuses and rails are the dominant feature of the PCB, taking long paths around the PCB on small amplifiers and this is not good for rail-induced distortion and noise, particularly for a design that has low PSRR to begin with. If I started a small new design, the fuses would be on the power supply board, since on the forum amplifiers I've assembled so far, that makes at least a measurable improvement.
I hope they become useful for you but I'm sure you will work out the layout better than I.
Could I suggest though, that you rearrange the output stage with the emitter resistors above,
rather than below the drivers/VAS. You have a very nice, compact arrangement with great
thermal coupling and short leads for the driver stage there. It would be good to retain that, I think.
Could I suggest though, that you rearrange the output stage with the emitter resistors above,
rather than below the drivers/VAS. You have a very nice, compact arrangement with great
thermal coupling and short leads for the driver stage there. It would be good to retain that, I think.
Member
Joined 2009
Paid Member
If you really want to re-layout the pcb then you could try something where the power rails are kept close together and close to power ground so that there are only small current loops. This is what I did with my very first layout for TGM and all the small signal circuitry was kept to one side. I didn't know at the time, but I think Rod does this on his P3A pcb too.
http://www.diyaudio.com/forums/newreply.php?do=postreply&t=164756
In my TGM layout I also kept the fuses at the power supply. The colours of the traces help show the idea - the red and blue traces at the top left are the main power rails and they keep fairly close together so that the large currents that flow in them are kept away from the sensitive circuits. The red and blue traces in the lower half are the small signal power rails. The lower right corner has all the small signal circuits in it. The power resistors (Re) are mounted under the board and shown with dotted lines. It's a very compact board (in TGM2 I managed to fit more parts on it!). I don't suggest you follow this method and I prefer the all-in-one board I made for TGM8 with the power supply caps and speaker protection altogether, but it gives you some more ideas to think about.
http://www.diyaudio.com/forums/newreply.php?do=postreply&t=164756
In my TGM layout I also kept the fuses at the power supply. The colours of the traces help show the idea - the red and blue traces at the top left are the main power rails and they keep fairly close together so that the large currents that flow in them are kept away from the sensitive circuits. The red and blue traces in the lower half are the small signal power rails. The lower right corner has all the small signal circuits in it. The power resistors (Re) are mounted under the board and shown with dotted lines. It's a very compact board (in TGM2 I managed to fit more parts on it!). I don't suggest you follow this method and I prefer the all-in-one board I made for TGM8 with the power supply caps and speaker protection altogether, but it gives you some more ideas to think about.
Attachments
Ian Finch's suggestion about placement of fuses is correct. I have similar experience, though I am a novice. I have another doubt- can I use a common heatsink for both drivers and thermally couple the VBE multiplier to this? on Rod's board, it will need some special design heatsink sheet since the powwer resistors are in betwwen the drivers. expect comments
Last edited:
Rather than hijack this thread for yet another P3A layout experiment thread, have a look at this thread, for example, where most guys have exactly the same issues. There are more threads about P3A layouts too.
http://www.diyaudio.com/forums/solid-state/201446-rod-elliot-p3a-layout-critics.html
http://www.diyaudio.com/forums/solid-state/201446-rod-elliot-p3a-layout-critics.html