what are the led's doing in here?

http://www.gyraf.dk/gy_pd/hpamp/HP-Amp.pdf ???

the ones right after the 5534..

The LEDs are biasing the output booster. LEDs are used because the 47R necessitate a larger voltage drop than the usual signal diodes provide. Although you would need to try different LED / emitter resistors to get the right bias. I would not want more than a few mA through the output at the quiescent point. They're probably red LEDs because they have the lowest voltage drop.

Ah cool.  Thanks.

If imputing a ULF sine would we see them alternate?

squarewave said:

....They're probably red LEDs because they have the lowest voltage drop...

Click to expand...

Actually no. Keywords are dynamic impedance and  LED tempco vs. BJT tempco.
Some useful info can be found here
https://groupdiy.com/index.php?topic=9121.0

pucho812 said:

http://www.gyraf.dk/gy_pd/hpamp/HP-Amp.pdf ???

the ones right after the 5534..

Click to expand...

Further evidence of how cheap LEDs got... when they first came out a single red one (their only color) could be tens of dollars. At Peavey we had dual diodes (effectively two diodes in series) in the system for biasing purposes. The dual diodes had a very specific tempco useful for compensating bipolar output devices.

That HP amp sets the class A current for the Class AB output stage at 23 mA which is in the range of some actual power amp output stage bias.  (I expect it sounds very good with no crossover distortion).

JR

thanks for all the info guys... you know what they say, hang with people smarter then you and learn. 8)