AP System One Front End

This is where we talk about testing, measuring and repairing things. Sometimes we have to repair the equipment we use to test, measure and repair other things. It's an endless cycle of fixing the broken things we need to fix other broken things.
Post Reply
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

AP System One Front End

Post by mediatechnology »

There are some interesting circuit tricks in the AP System One front end.

Image
https://proaudiodesignforum.com/images/ ... ematic.pdf

Image
https://proaudiodesignforum.com/images/ ... ematic.pdf

Of particular interest at the moment is the "bias current assistance" role of R8601 and R8604.

R8602 and R8603 establish an 11V filtered reference for 22MΩ resistors R8601 and R8604.
R8601 and R8604 then act as voltage dividers with the 106KΩ input attenuator networks and their parallel 1.6MΩ compensating resistors.
The parallel combination of 106KΩ attenuator network and the 1.6MΩ is almost exactly 100KΩ per leg.

When the attenuator is off, the 22MΩ and 100KΩ voltage divider network biases the NE5534 inputs to approximately +50mV...
Or does it?

Add the NE5534 back into the picture.
An NE5534 has NPN inputs and it's bias current pulls the bias resistors slightly negative.
It's sinking current into its inputs.
What this network does is provide crude "bias current assistance" for the NE5534s.

By biasing the inputs in an opposite polarity, some, if not most, of the bias current is cancelled.

For this approach to work assumptions about what the bias current is have to be made.
The NE5534 has a published bias current specification of 500 nA typical (1500 nA max) at room temp.

The attenuator is off.
-500nA * 100KΩ is -50mV.
If the bias current compensation scheme were disabled, the -500 nA typical bias current would pull the NE5534 inputs negative by about -50 mV.
To correct this R8601 and R8604 source 500 nA current into the inputs. (11V/22MΩ.)
By sourcing an offsetting positive current into the inputs, the bulk of Ibias is cancelled.
The +50mV positive bias offsets the typical NE5534 Ibias * R offset of -50 mV.

In the AP, the amount of correction voltage is lowered as the attenuator is engaged but the current remains constant at 500 nA.

This is not "bias current compensation" by any means but it does provide the bulk of the required bias current and reduces the Ib * Rbias error significantly.

Its a nice trick and one that has obviously stood the test of time.
Audio1Man
Posts: 18
Joined: Mon Feb 04, 2019 2:09 pm
Location: Thousand Oaks, CA
Contact:

Re: AP System One Front End

Post by Audio1Man »

OK, not many persons got this. Bruce spent lots of time correcting several small errors that make small improvements that give the total system performance. There are others tricks and when another Mfg copied the special tricks this stopped all schematics/service information being released. The System Two and beyond have no detailed service info.
Duke
Post Reply