P3A layout

[sakis]

Originally posted by AndrewT
why would two matched BJTs have different dT/dt? I thought this was pretty near fixed by the physics.

Different Dt/dt is not what shows when a matched pair are in the jig.

However, when testing jFETs, then they definitely have different slopes, and one has to find similar Idss and similar slopes, to then select matched pairs. The slope is effectively Gm.

probably nico was refering to my previous post regarding not matched ltp pair and also with no thermal junction .....

a non matched ltp pair will probably have diferent offset behaviour and of course this will be also temperature depending since there is no thermal junction

[mjf]

Originally posted by Nico Ras


then they will perform differently, and worse if glued together.

hello all to you.
hello dear nico ras.
didn't you think of:......,and better of glued together?
greetings...........

[mjf]

hello dear sakis.
foto: nice little substitute for an amp(lification of emotions);
annother theme:i found an old schematic (perhaps late 7ties,mullard?).i think it looks like the p3 a little bit.
greetings all to you.................

[crt]

how about this layout?

layout

unfortunately, have one parts should mount on bottom PCB to keep the track short.

[sakis]

looks very nice and symmetric ( the type of pcb i would like to built )

i will ring a bell though .... here are some thoughts of forum member EVA that took a second and bother with the same issue in another thread .....

******0.1mfd decoupling have no serious effect if located like that IE next to ground taking also in mind that both rails traces are quiet long .... they are too far away fro outs ( evas opinion about this is that it only works good if located next to ground and very next to outputs )

****** then this long trace of out put can aslo cause complications ..... out resistors should be as close possible

****then again signal ground eventhough equiped with 10 R resistor to create a more "clean" ground is quiet long

****** and finally ltp has no option for thermal junction ( that could be very nice also ....

ONE THING i clearly dont like is the resitor tha feeds the ltp second transistor from out put ....this to me looks like a very nice capacitor ...this is something i would never do

eva's ideal pcb was something that had power caps almost in the midle ( we talked about that also ) power caps included on board very possible since this amp will work with 2*10.000 mfd on board ...so caps are on the midle with all power ,out and ground conectors next to them .... just behind the caps the power trs and on the totally other side the small stuf like ltp vas and other stages .....

the goal of all this is to achive the most minimum distortion .....

after studing this for very long ....if you need to fullfill all these specs then this cannot be done in single side pcb ....

and this arize the question if an amp like that worth such an effort IE double side pcb /extreme design....

other wise pcb looks just fine

best regrads sakis ( with very sexy avatar )

[Adi_007]

what do you think about this layout

[Nico Ras]

Originally posted by Adi_007
what do you think about this layout

Nice Layout Now build it.

Nico

[sakis]

this is more or less they lay out i ve done ...with a few minor changes .....

this will work with out any problems at all ....( eva's idea about decoupling and so cannot be done here but we talked about that in a previous post )

only one thing .....

bc transistor that is located top midle is the VBE multiplier CANNOT !!!be attached to the heat sink ( please read carefully rod's recomenation about all that )

though it can be stilllocated there but not attached to the heatsink

nice rounded corners well done !!!!!

[Rafael L]

I build P3A amp, work good and is cheap, I used Q4, Q5, Q6 BD139,140 and output MJE2955/3055 (60W/ 8 ohms)
Some observations:
-Q5 and Q6 Not in heatsink TR's output , Nor need heatsink they operate in class B
-Q4 Need heatsink in +-42V
-Q9 Not is heatsink (according to Elliott),I not put Q9 to heatsink, noticed some instability not colled after work. More low temperature to turn without signal (music),is a little more force signal not back to state low temperature, does not affect more and works well for hours

Originally posted by Havenwood
Originalmente postado por Havenwood [/ i]
[Whit Re1 = RE2 = 100ohms.
E adicionar Re (Q1) = Re (Q2) = 22ohms? (constante degeneração gm)

Olha Audio Power DesignHnadbook por Dougles Self pag84.
Gostaria de saber a sua opinião

This design P3A has many limitations, The first stage is too loaded (gm) to compensate for low Beta the second stage, if improved will be Blameless.
Best build directly Blamelles of book

[Adi_007]

you are right......you layout was my sourse of inspiration......the transistor that you are talking about I will replace it with a BD139