Adcom GFA555mk2 Optimum Bias

[jackocleebrown]

Hi All,

I have an Adcom GFA55mk2 amp.

I've got a copy of both the service manual and the schematic, links below.

In the service manual it recommends setting the bias to 10mV measured at the test points. The emitter resistors are 0.22 Ohm and the test point bridges across one pair (total resistance 0.44). So that gives only 22mA quiescent current for each pair of output transistors.

This seems very low and must be below the optimum to reduce crossover distortion? I wonder if it is a mistake in the service manual?

Has anyone tried running with a bit more.... say 30mV (~66mA)??

Cheers,
Jack.

Service manual: https://dl.dropboxusercontent.com/u/37121092/hfe_adcom_gfa-555_ii_service.pdf

Schematic: https://dl.dropboxusercontent.com/u/37121092/hfe_adcom_gfa-555_ii_schematic.pdf

 

[JonSnell Electronic]

0.02mV will remove crossover distortion. Output transistors need to be only just on otherwise you are changing from Class B to Class AB or Class A, destroying the quality of the amplifiers design. 10mV is ample to remove any trace of distortion.

 

[jackocleebrown]

0.02mV will remove crossover distortion. Output transistors need to be only just on otherwise you are changing from Class B to Class AB or Class A, destroying the quality of the amplifiers design. 10mV is ample to remove any trace of distortion.

Thanks for the reply, John.

I agree with you that I'm likely to change the character of the amp. However, I still think 22ma is quite low and a higher bias would normally give lower measured distortion, which is why I am curious to try it.

I don't have any data for the 2SD424/2SB554 pair used in the 555mk2, perhaps these devices are different, but I've just simulated a wingspread diagram for a mjl21194/mjl21193 pair into 8 Ohms. With 23ma there is a significant droop in gain around 0V. 63ma is about right to give the flattest wingspread. I believe this is pretty typical.

An externally hosted image should be here but it was not working when we last tested it.