P3A 100W layout - comments?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have made some adjustments to the PCB. At first, the PCB its much smaller now. It was difficult to reposition the component because the two power transistors have a fixed position due the heatsink.

Also I have added two 100uF caps at the +35V and -35V (C9, C10)

pVQy7.jpg

You forgot to foresee heatsink's for those transistors that I have mentioned. And not only that. However, the beginning is promising. Try again.
 
Hi,

The bias loop is Q9, Q5 and Q6. Mounting Q9 on the heat sink wont help
at all in matching Q5 and Q6. Q7 and Q8 temperatures do not affect bias,
unless they are thermally linked to Q9.

Cheapest and simplest is to mount them all on the same heat sink (bracket).

Alternatively Q5 and Q6 have their own or shared small heatsink,
here Q9 should be linked to at least one of the devices, in theory.

This is all theoretical, and the P3A is very liberal in its claimed biasing
requirements, whilst in fact the optimum bias is a very precise affair.
(See Douglas Self and CFP output stage optimal bias.)

If your planning on a nominal 50mA to 100mA bias its already
overbiased and most of the above does not matter. Overbias
can wander around without becoming any sort of real issue.

Just make sure Q5 and Q6 have heatsinks.

rgds, sreten.


Sreten i think that you should take a closer look in Rod's article eventhough the description is brief the principal of operation of especially the output stage is quite clear .

Since it seems that you have quite enough knowledge you may as well check wiki or any other reference about sziklai style amplifiers ...

Your recommendations about the Vbe multiplier and generally the bias approach are wrong .

As about the pcb don't put too much worries on it P3A is very forgiving when it comes to things like that ...it will play music any way ....Though if your target is really hi end performance this PCB and configuration will not make it .

My personal opinion about MJE 15032 drivers is that these are extremely rugged devices and i have been using them as outputs in some small amplifiers but the sound of them is metallic and harsh... Bd139 -140 will sound better but have less kick there is many other real alternatives from Toshiba 2sc4xxx family .

Other than the above there is plenty of other tricks that could lift the P3A to the sky ( and far superior form similar but EFP circuits ) .... Only condition is understanding how sziklai amplifiers work and that can be seen by simply looking at the schematic .

Kind regards
sakis
 
Last edited:
just for the History of it ...
the mother of all transistor amplifiers was EFP .... beyond the first RCA designs and dynaco i think the rest of it was Japanese ...after that all of them start copy or learn from another .

Point that mr Sziklai originates from Hungary while spend most of his life in the states ....There is quiet many Sziklai amplifiers world wide but eventhough i repair 600 amps per year i hardly recall seen a made in Japan Sziklai ...

In a few words i think that these circuits have not been copied so far ( like the above style from American or European designers to Japan or visa verso ) So i think that an sziklai circuit will not be easy to understand for someone that has been working EFP all his life ....

Open the Toshiba book ""Small signal transistor semiconductor data book '83 "" which is the mother book of legendary 2SA 970/ 2SA 1306/2SC 3298 and of course the superior 2SA1302 /2SC3281 notice that you will find plenty of application and circuits but none of them Sziklai .

I still think that these circuits and for the above reasons of not copying has not yet advanced properly and never achieved its full potential like the EFP did ..

That is just my opinion ...

Kind regards
sakis
 
  • Like
Reactions: 1 user
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.