Adcom GFA-565 Bias Adjustment Instructions?


2008-01-14 4:43 am
Hello, all.

I have lurked for some time and learned much, but now have a question (which I have not been able to find answered elsewhere).

What is the procedure for setting the bias on the GFA-565?

I know that with "no input" (I presume this means inputs shorted), you adjust for 24mV +/-1 after 10 minutes, then run for 10 min. at 80W output (I assume the input used to obtain this output isn't critical, though a sine wave might be simpler to manage) and readjust for 24mV (I assume with "no input" for the actual measurement).

What I don't know is which points to test at (apparently it's 301 and 401 for the 585 - are the 565 points labeled the same?), and how many locations to test (I believe it's one for the whole unit, controlled by the pot on the input board, but want to be sure it's not one for each bank of transistors, or some other grouping).

Also, how critical is the 80W figure? Can one ballpark it, or must one confirm it precisely? I'm not a tech and have only good meters (analog and electronic), a freq. generator, and non-latching, single-trace oscilloscopes.

Finally, any advice on obtaining a good schematic? I've gotten the Adcom manual from Adcom (Twice - they seem to send it in response to any question), but the scan is low-res, and some of the labels are illegible.

For those interested (go to next post, or fall into deep stupor here), my story is thus:

I've had a pair of very underutilized Infinity QLS-1's for about 15 years, which I bought at a flea market, along with a lot of other gear, over fifteen years ago. Yes, the price was right.

I ran them for many years with the GFA-1 (low side) and GAS Grandson (high side) that the previous owner had used, in a succession of rooms that were far too small to really hear them.

About two years ago, I spotted a set of four GFA-565's on e-Bay, saw that they were located within blocks of my house, and came to terms with the owner which saved both of us from the horrors of shipping. I ran four separate 20A AC lines and breakers, and used the amps happily but occasionally, as I had mostly moved to another state by then, and was just clearing out the house prior to selling.

Upon arrival at my new home, room arrangements prevented me from setting up right away, and I decided to test DC offset "for the heck of it."
Although I don't have the numbers in front of me, I found three amps with offsets between 3 and 20mV, and one around 25V. Hmmm. That would be the one that always thumped.
Upon disassembly, one had Elna caps, and a mess on the input board and three did not.
I have since bought enough good quality high-temp caps to replace all the electrolytics in all four machines (less than twenty bucks), as well as four OP97's ($2.50 each) and sockets. I was shocked to find that wire-wrap tools start at $30 (and then only for either On or Off, but not both ways) I swear they were only a few bucks the last time I saw them in Byte magazine. I have no intention of replacing any conductors with solid gold spun from Bavarian maidens' hair, or anything else with anything that costs more per gram than Bolivia's finest.

Anyway - I'm primed to rip that bad boy apart, wash up that board in Simple Green (though I wonder about such other solvents as methanol, or industrial concrete-cleaning detergents), replace parts and fire her up. It's my nature, though, to want to confirm such things as the bias adjustment when I mess around with all that stuff, so I haven't started rolling yet.

Oh - and my new venue for the QLS-1's will be our new business location - a 75 by 25 foot space in an old mill with 12 foot ceilings. Finally, a room that's big enough!

Thanks in advance for your help!

- Eric


2007-01-26 9:43 am
Thread should be moved to "Dead Adcom GFA-565 Monobloc"

Hi Eric,

Glad to see that you're attempting to fix the 565.

Your bias adjustment process is different from what Adcom suggests: running 100watts through the 565 with the cover on, then taking a bias measurement and calibrating to 24mV, run the amp for another hour, test again.

The leads to take the bias measurement are on the output stage boards (the boards closest to the input board). The leads are thick metal and pointing inwards, you can't miss them.

Be careful not to touch the metal bridges than run between output stages while the amp is on, they carry current from the filter caps!

Also, when you clean the boards, use an ultrasonic cleaner, otherwise you risk the possibility of leaving some electrolyte on those boards, which causes the offset.

Good luck


2008-01-14 4:43 am
Thanks, MJR.

I finally found the six lines of instructions on bias adjustment in the manual, just under the warning not to use it in the bathtub. I must have glanced over them a dozen times without seeing them. Duh.

Bias was good, but after a (seemingly) thorough series of washings and a new set of 'lytics, the amp still pushes 27VDC. It looks like I'm going to have to bite the bullet and beg borrow or steal an US cleaner.

I've read al of the threads I could find on this, and feel as though I've got the general idea. I know that you did a very thorough job on your 565's. I've asked this question on another thread, but perhaps you are really the person to ask:
Have you got a source or a search term for tiny pin connectors (my input board uses exactly one of them) that solder through the wire holes in the board, and little tunnel-like wire ends that slide over them - it would make the anticipated long series of disassembly and reassembly cycles a bit easier to deal with. Oh, and also those little screw terminals I believe you used on the 3 larger wires.

Wish me luck...


- Eric


2008-01-14 4:43 am

I found the reference in the thread!

MJR, would it be possible for you to tell me the part numbers and sources for the:

"WECO PCB screw-down terminals (epoxied to the PCB with cyanoacrylate, then soldered), gold headers for all LED connection points... [and] ...Copal 100 ohms Cermet variable resistor"

that you used to upgrade your input boards?

I'm very concerned about the possible toll that an as-yet-unknown number of soldering and desoldering cycles may take on my board and components as I troubleshoot my DC offset problem, and would like to be able to get the board in and out without the extensive melting of metal at each iteration.


- Eric