Adcom GFA-555II

[sakis]

Originally posted by anatech
Hi sakis,
Well, those BJTs will always come in handy. I would use them for something a bit more refined than the Adcom. For the Adcom, I would go with some of the newer On-Semi devices. I used up my Adcom output transistors some time ago, no loss there. I'm thinking of amplifiers like the Nakamichi 620 or similar, those would benefit from the original Japanese parts.

When we got mosfets from anyone, we always checked them. Even if they promised they were matched. Now, if you think that mosfet matching is important with an Adcom, you should see how critical it is with a Counterpoint amplifier. No source resistors at all, so they really did have to be quiet closely matched together. You will not enjoy the experience, trust me on that.

I'll say one thing, at least Adcom was honest about telling us the mosfets were not matched and that they expected us to do that work. I can't say I was happy about it, but at least we knew. Some manufacturers maintained that their parts were matched. Having spent time in the industry, I didn't trust them. That saved me a lot of aggravation on several occasions.

-Chris

i wouldnt say actually

the 5800 series we imported at the time all had hand writen codes on each and every of the transistors regarding matching ....

but also at the same time we think tha adcom started to produce boards in the taiwan so the first problems started there ..... 5800 come in without any notes on the transistor ....indication that either matching was skiped fron default or skiped by mistake ....

the particula amps presented excesive ammounts of tem but otherwise seemed to work ok

voltage drop on the resistors actually was very diferent from transistor to transitor ...

then adcom made the same mistake by sending us again transistors from the box ( still have them ) and at the third time they send matched sets ...installed them tune and and let go till today ...

( nice to hear from you chris ....hope you are doing really well )

[DreadPirate]

I'm probably going to tinker a bit with this unit (GFA-555-II), at the very least do the 22uF bypass on each main cap as suggested by DJK. His mod is using a Bennic NP, any reason why I shouldn't just go with a film cap, other than cost?

I'm also going to add another 30,000uF of computer grade capacitance, bringing total up to 90,000uF. Depending on the outcome of these two mods (if positive), I may even invest in a pair of 800VA toroids and further upgrade the PS.

I'm of a different opinion as to the sound quality, it is somewhat congested, so detail is poor, and the bass seems a bit uncontrolled. No contest compared to my Aragon 8008BB, but perhaps it is not a fair comparison. Bias and offset not tested yet.

Here is a free copy of the service manual, for those interested:

http://www.acousticpsychos.com/Files/

[CBRworm]

So you are going to replace the 4 OE 15,000 caps with 22,500 caps? or are you going to piggyback 4 7,500's?

If I were going to do that I might consider adding a soft start circuit, although I would be hesitant to haphazardly add 50% more capacitance to what should be an adequate supply without figuring out what part will fail, will it be the rectifier bridges?

[anatech]

Hi sakis,
We never did get any of those amps that didn't have matched outputs. I am in the habit of checking current sharing in everything I work on. That way I don't have avoidable failures to worry about. The customer would not know the difference between bad work and a assembly induced fault. Besides, out of warranty I can warn them to see if they want to fix that problem or not. Either way, I am not going to be at fault for something someone else is responsible for.

To be honest with you, I found that most mass production stuff used outputs from the same batches, it was the low production higher end stuff where matching seemed to be problematic. That's probably due to the lower quantities of parts bought, they didn't understand that they may have to match where someone like Yamaha or Sony wouldn't have those problems to begin with.

voltage drop on the resistors actually was very diferent from transistor to transitor ...

A clear indication they are not matched at all, as you've pointed out.

nice to hear from you chris

Thank you! It's always nice to "speak" with you too. We have very similar struggles in the audio market.

Hi DreadPirate,

I'm also going to add another 30,000uF of computer grade capacitance, bringing total up to 90,000uF.

I wouldn't. That seems to be a popular way to stress the rectifiers for no gain. Think about why you think the extra capacitance will help. Are the filtered rails collapsing by chance? I happen to know first hand that they are not. What you are doing though is reducing the charge time and increasing the levels of peak current. That leads to higher I-R losses everywhere, and probably more HF energy being created in the rectifier section.

Understand that as long as your supplies remain above the voltages you are putting out, you will not be clipping. Beyond that, I really don't see any reasons for increasing supply capacitances. Andrew and I disagree very strongly on this point.

I once worked on a large Bedini amplifier that had giant capacitors running from front panel to almost the rear panel. The inductance was so high that the 120 Hz ripple was extreme. Okay, so it played for some time after you turn it off. But that isn't how we generally use our amplifiers. This amp sounded much better with lower values of capacitance, but that did stand in the way of good ad copy!

-Chris

[DreadPirate]

I'll just add a few thousand uF then, see if there is any effect. My Adcom runs on rather the cool side, barely warm even when driven hard.

Originally posted by anatech
Hi sakis,
We never did get any of those amps that didn't have matched outputs. I am in the habit of checking current sharing in everything I work on. That way I don't have avoidable failures to worry about. The customer would not know the difference between bad work and a assembly induced fault. Besides, out of warranty I can warn them to see if they want to fix that problem or not. Either way, I am not going to be at fault for something someone else is responsible for.

To be honest with you, I found that most mass production stuff used outputs from the same batches, it was the low production higher end stuff where matching seemed to be problematic. That's probably due to the lower quantities of parts bought, they didn't understand that they may have to match where someone like Yamaha or Sony wouldn't have those problems to begin with.


A clear indication they are not matched at all, as you've pointed out.


Thank you! It's always nice to "speak" with you too. We have very similar struggles in the audio market.

Hi DreadPirate,

I wouldn't. That seems to be a popular way to stress the rectifiers for no gain. Think about why you think the extra capacitance will help. Are the filtered rails collapsing by chance? I happen to know first hand that they are not. What you are doing though is reducing the charge time and increasing the levels of peak current. That leads to higher I-R losses everywhere, and probably more HF energy being created in the rectifier section.

Understand that as long as your supplies remain above the voltages you are putting out, you will not be clipping. Beyond that, I really don't see any reasons for increasing supply capacitances. Andrew and I disagree very strongly on this point.

I once worked on a large Bedini amplifier that had giant capacitors running from front panel to almost the rear panel. The inductance was so high that the 120 Hz ripple was extreme. Okay, so it played for some time after you turn it off. But that isn't how we generally use our amplifiers. This amp sounded much better with lower values of capacitance, but that did stand in the way of good ad copy!

-Chris

[anatech]

Hi DreadPirate,

My Adcom runs on rather the cool side, barely warm even when driven hard.

That doesn't sound right at all. An Adcom should idle a little warm and heat up with the program material.

Have you checked the bias currents? The procedure involves running the amp hard until it's good and warm - hot. Then you allow it to cool for 1/2 ~ 1 hour until the bias settles in. Then you measure it. Same procedure if you change the bias current.

-Chris

[Jkrider]

Hi,

I'm new to this forum and had some question on the Adcom GFA-555II amp. I've seen a couple of threads mention changing the electrolytic caps. I looked at the amp and none of them appear to be leaking. Should I change them anyway? Thanks for all the info on this thread.

James

[Jkrider]

Also,

I noticed that there were a number of people who talked about adding 22 uf bypass caps to the filter caps on the power supply for the GFA-555. Would this same mod apply to the GFA-555 MkII? I thought I read somewhere that the bypass caps were added to the MkII.

James

[Coyotesx5]

I used to own the GFA-555 and am looking for a used GFA-555II. But some folks have told me to listen to the ADCOM 5500 with MOSFETS before I go "too far back in time." I used to have the 555II into ML CLSes until I stupidly replaced the ADCOM equipment with Krell MDA500s. I sold those when the PS caps dried out(as did a whole bunch of other caps in my KBL and my KSP64). Given the huge number of 555's out there and the age of those models, can anyone in this forum recommend a good source for electrolytic PS caps and polypropylene/mylar/etc. caps to change out old caps, as we get closer and closer to the audio signal path? I can solder and have done mods before. The 5500 has to really sound incredible for me to buy an amp without bridging mode...I now have Thiel CS5i's and they would really benefit from the high damping factor and massive power output of the 555 amps at their inherently low impedence. But I have always felt that the 555 amps were fitted with inadequate PS capacitance...and would double the Farads...esp. given all the extra room "to grow" in the ADCOM amps.

[Coyotesx5]

Hello all,
Does anyone have a schematic for the GFA555MkII? I've looked and looked...to no avail.
I want to reference a few component values and change out some caps. Thanks,
My email is joe85255@gmail.com.