Jump to content

AR Amp Bias - one more time :blink:


Recommended Posts

After several days of fiddling, tweaking, testing, listening, and analyzing, I have come to the conclusion that AR got it right the first time when they installed pots to adjust the bias! Well, let me rephrase that...they got the concept right. Unfortunately they fell down on the execution due to a poor choice of a potentiometer. Or maybe back in '69 it was the only suitable pot available (though I very much doubt that).

Why do I think AR had the concept right? Because over the past week I have found that although the AR factory mod - fixed resistors - eliminates the overheating due to the poor pot's instability, it does NOT allow you to actually achieve the results that pots can deliver. (Note: That difference probably doesn't matter and the factory mod DOES eliminate ever having to fool with it).

Let me explain:

The biggest problem for me was I couldn't determine what the bias voltage/current should be for the amp. I had several suggestions from folks on this site, from emails, and on other sites covering biasing in general. But NO specs from AR. After cleaning the pots I tried various settings to deliver assorted bias voltage across the emitter resistors - everything from a couple of mv to 100mv. I quickly found that anything above 25mv produced a good bit of quiescent heat, though not NEARLY the heat that the amp had been producing before service. Then, based on a post here, I set the pots at 12 ohms, which is within the range AR recommended for the bias mod (12 or 13 Ohm fixed resistor). THis, as I posted previously, produced a cool running amp and a bias voltage across the emitters of around 15-17mv and an offset of around the same figure. So I ASSUMED that if a 12 Ohm resistor produced 15-17MV across the emitter resistor, then that that figure must be within the bias specs for the circuit. Again - that's an ASSUMPTION and it could be wrong since I don't know if other components in the circuit have shifted value and made the current delivered to the transistors other than what it would have been with new components.

Although the offset was quite good, since I had managed to get 0 offset on my Moscode 600 (pots) I decided to continue to fool with the AR. I found that by tweaking the pots very carefully I could get the offset to 0 and keep the bias in the 15-17MV range. However, the oem pots, though as clean as I could get them, are very sensitive to any movement and will not hold that level of tolerance. Any movement - including tapping the amp - can cause the offset to immediately shift randomly as much as 50mv. It can also demonstrate this behavior somewhat randomly, tapping or not. It never gets BADLY out of adjustment but demonstrates again that the pots chosen for the job were not up to it.

So I have ordered some good sealed 15-turn pots to substitute for the eom pots. When I receive them, I will remove the oems from the circuit (though NOT from the amp) and install the new ones in a way that does not permanently modify the chassis or wiring. I expect they will do the job that AR intended originally!

I have to admit that the difference between the fixed resistor mod and the results obtained with good pots is probably not worth quibbling about performance-wise. The resistor mod would seem to get the offset (on MY amp) down to around 17mv. But heck, zero offset and precisely matched bias current is better!

Yes, all this fiddling is more for my own satisfaction than for any expected sound improvement. I have several MUCH better and far more powerful amplifiers than the AR but still, there is something intriguing about it and I can't leave anything alone anyway... :blink:

Link to comment
Share on other sites


This topic is now archived and is closed to further replies.

  • Create New...