high performance 25W PowerAmp

I know that fact 😢.

The problem is more of a academic nature, for this mini circuit you /we can certainly evaluate any layout program. But merging D into X should not actually be a hurdle at all, after all, this arrangement has been functioning for a good half century.

The scenario reminds me more of the fact of German (or international) incompetence in matters of repair service for our automobiles:
An example could be, "since our diagnostic connector or the latest software update is no longer compatible with your SuperDuper DSG, we are unable to change your wheels."

Sorry, this is offtopic.
 
Last edited:
Tim's first circuit board

Everything your heart desires: converted and stored here as SprintLayout6 files. The *.zip now contains the PCB in modern form as *.lay, *.ps, *.gbr, *.drl and there should now be no problems with further editing.

Perhaps someone would like to create an assembly print 😉.

Have fun,

HBt.
 

Attachments

hbt
please cross-check schematic...

yellowsubmarine_sch.jpg


Idle currents
yellowsubmarine_idle.jpg
 
1718645290816.png



Q1,Q2,Q3,Q4 plus N1,N2 should (finally) all be immersed together in a small aluminum body (to be made), add some thermally conductive paste.

Under no circumstances should the quiescent current through Q1 & Q2 be greater than 10mAdc ... the common RE= (10Ohm/3) is far too small to have a stabilizing effect ... This is one of the imponderables of this stage. It may be necessary to replace N1 and N2 with other types or, or, or ...

The power supply to the output stage is unregulated and separate from the pre-stage (in reality!).


The 5mAdc sink is already rudimentarily implemented on the circuit board.

#
Dear Bernhard,
I think it's great that you decided to feed MicroCap12 with the circuit after all - brilliant.


Thank you
 
Last edited:
So the 5mA now should be a ccs?
I removed it in my schematic and the resistor was as good as the ccs.
Also I simulated your schematic already. 4mA, 5mA, 6mA - doesn´t matter.
And, the ccs doesn´t help either.
Same with 3.3k and 3.3 ohms.
Different idle currents still show the same result.

yellowsubmarine_sch2.jpg
yellowsubmarine_idle2.jpg
 
Note:

Under no circumstances
  • should it be assumed at this stage that the circuit shown here has already been tested in practice and finally dimensioned
  • will I be able to etch or mill the two test boards before the coming weekend, i.e. 14 days will certainly pass before I can contribute real results to the test setup.
Please be patient.

For me personally, a little PowerAmp has to follow the KISS or economy principle.
I hope many readers will feel encouraged to think through the circuit and size it successfully. Practical results from my side will not be available for a few weeks.

Anyone who would like to ask questions now is encouraged to do so.

Now that a model for MC12 is available, @Bernhard can already put the design proposal through its paces in the computer and sound the alarm if something could go seriously wrong.


regards,
HBt.
 
Last edited:
This is by far not the only problem. a spice simulation confirmed my initial bad feeling: This circuit will not work at all.
The culprit is the coupling from Q2.collector to the VAS - prohibiting a stable DC operating point.
 

Attachments

  • ampsimu.png
    ampsimu.png
    34.6 KB · Views: 52