That is a good outlook to have.
I owned a GFA-555II for some 15 years and I rotated it in and out with a couple of other amps in my setup. I liked the sound and only got rid of it because at my age I am downsizing and reducing my audio inventory considerably. The original GFA-555 may sound a bit better than the MkII and it was designed by Nelson Pass (along with the GFA-5800 and GFP-750).
Last edited:
Agree completely that servo should have had bipolar supplies, and with all your comments in general. I do think that the input pair might have reasonably well balanced collector currents; the load networks are different, so the only way to know more accurately is to measure voltage drops across R105 and R106, assume base currents are negligible, and estimate Q101/Q102 currents from resistor values. I think bias dissymmetry was intentional by 100K at input vs. 1k at feedback. Guaranteed positive voltage needed from the servo opamp with accompanying thump. Cheap compromise for BOM, not necessarily bad for steady state listening.
Best regards.
I'm not going to rate it in the lineup. I simply formed that opinion after years of servicing the product under warranty, but I never bothered to rank each amp between them in the Adcom line. I could have maybe 30 years ago.
Keep in mind I have serviced so many different brands and models of amplifiers that keeping a detailed memory of each is futile. I will say that there are some DIY amps I love and are better.
My best three amplifiers of all time? The Bryston 4B3, the Marantz 500 that I have updated and the Marantz 300DC, also updated. This opinion is shared by many who have heard and owned at least two of the above list. Each measures impeccably as well. There are several other amplifiers I really like and can live with.
I also really like some tube amps, certainly not all. But my main listening amplifier is a Marantz 300DC, I can't justify the 4B3. I have owned Lighstars, TFM-75 and others. Also amplifiers from other manufacturers. I have a Luxman M-02 in the bedroom (modified). By the way, the main system speakers are only about 86 dB/watt, PSB Stratus Gold (original models). So they take some grunt to get loud, and I listen to music LOUD! My wife passed away in 2020, so I have no limits. 🙂 The other speakers are 98 and 99 dB/watt. New Klipsch reference series, not the classic stuff.
I understand your position re: many amplifiers. Many clients also tend to collect very good amplifiers. It is a good way to determine what you do like. Having said that, proper changes, and never found on the internet, can really improve the sound of an amplifier or anything else. You might be able to find someone who knows what they are doing close to you. They can transform your equipment. It does not involve "audiophile parts" of any kind. It is about matching (properly) components, selecting newer types, and also using modern circuit techniques to boost performance. Normally this does not involve cutting traces!
-Chris
Keep in mind I have serviced so many different brands and models of amplifiers that keeping a detailed memory of each is futile. I will say that there are some DIY amps I love and are better.
My best three amplifiers of all time? The Bryston 4B3, the Marantz 500 that I have updated and the Marantz 300DC, also updated. This opinion is shared by many who have heard and owned at least two of the above list. Each measures impeccably as well. There are several other amplifiers I really like and can live with.
I also really like some tube amps, certainly not all. But my main listening amplifier is a Marantz 300DC, I can't justify the 4B3. I have owned Lighstars, TFM-75 and others. Also amplifiers from other manufacturers. I have a Luxman M-02 in the bedroom (modified). By the way, the main system speakers are only about 86 dB/watt, PSB Stratus Gold (original models). So they take some grunt to get loud, and I listen to music LOUD! My wife passed away in 2020, so I have no limits. 🙂 The other speakers are 98 and 99 dB/watt. New Klipsch reference series, not the classic stuff.
I understand your position re: many amplifiers. Many clients also tend to collect very good amplifiers. It is a good way to determine what you do like. Having said that, proper changes, and never found on the internet, can really improve the sound of an amplifier or anything else. You might be able to find someone who knows what they are doing close to you. They can transform your equipment. It does not involve "audiophile parts" of any kind. It is about matching (properly) components, selecting newer types, and also using modern circuit techniques to boost performance. Normally this does not involve cutting traces!
-Chris
Glad to see that you removed those 2 large audiophool caps attached to the output stages. The previous owner obviously did not know better.
You can remove the 4 - WIMA MKS polyester film caps as they are not needed for the output stage bypassing. The 47uF electrolytics work just fine and don't need the extra caps across them. Use those caps instead for bypassing (along with electrolytics) for the pc board power supply +/- as described in the upgrade sheet that I sent you the other day.
Finished (I think - LOL!)
Got my order from Mouser and worked until midnight last night replacing all the parts. Fifth time taking it apart, so I'm getting pretty good at it.
!) matched and replaced all TO-92s (Q104/Q154, Q106/Q156 with KSA992, Q103/Q153 with KSA1845)
2) matched and replaced all TO-220s w/ heatsinks (Q105/Q155, Q107/Q157 with KSA1381-E, Q108/Q158 with KSC3503-E)
3) replaced C109/C159 (had Audio Note Caps but replaced with Panasonic FC) - ordered 8mm x 3.5mm LS (size should've been 10mm x 5mm LS)
4) replaced IC101 (prev. replaced IC101) with LT1006
5) replaced C106 (15pF) and C102 (330pF) with CDE 500V mica
6) replaced C201/C251 and C301/C351 (all 68pF) with CDE 500V mica
7) replaced all four LEDs on the front panel (Kingbright WF424EDT-orange, and WF424IDT-red (slightly shorter and a little brighter)
8) went over most of solder joints, especially on the wire connections - no foul smells, so I think I avoided any electrolyte seepage
Possible future change(s) - adding (4) 47uF/100V Nichicon Muse KZ with (4) 0.1uF bypass caps on the pre-amp board and change to an IEC plug
Summary of repairs:
I have changed all pre-drivers, drivers and outputs (KSA1381-E/C3503-E, NJW1302G/3281G and MJ21193/94G), both servo ICs, all electrolytic capacitors (pre-amp and output boards), all film/mica capacitors (pre-amp and output boards), all TO-92 and TO-220 transistors (pre-amp board) and checked all resistor values (replaced 4) and checked all diodes/zeners (all checked out fine).
Testing:
Since it was late when I got it back together, I waited until today to start it up and check bias and functionality. I was tired and didn't want a slip of the probe, or missing connection to give me a repair before it's fully working.
I recorded the start-up behavior yesterday morning, so I could compare the same start-up after replacing parts and compare any improvement.
Before these last repairs it took 2:45mins to reach 4mV of DC offset from a cold start-up (cold start is >3 hours off and unplugged). I ran the same test today and it settles below 4mV in 0:06, so more than 2-1/2mins quicker. Now the left channel mimics the right channel behavior - gets to 5mV in <10secs and tracks at +/-2mV. Not 100% sure if it was the IC replacement, changing the capacitors/transistors or touching up the solder joints. Both channels are working like I thought they should from the beginning and the amp has a really nice sound.
Now I get to spend some time listening and enjoying after spending all this work to bring this amp back from the dead. So far, pretty impressed - it's a fun, lively amp to listen to and I'm impressed with the bass drives my speakers pretty good in my small office. I think the combination of replacement Caps I used has brought it up a notch from what the previous owner had used, I think he was more concerned with the brand and not about the actual location and job to use the best possible capacitor.
Huge thanks to @bigskyaudio (upgrade guide and technical assistance here and via PM), @BSST (explaining the circuit and your fault finding suggestions), @anatech (You know these amps in-and-out, I appreciate your insights and suggestions) and @Phloodpants (read through all your threads/posts and website on Adcom amps, insightful stuff)
Got my order from Mouser and worked until midnight last night replacing all the parts. Fifth time taking it apart, so I'm getting pretty good at it.
!) matched and replaced all TO-92s (Q104/Q154, Q106/Q156 with KSA992, Q103/Q153 with KSA1845)
2) matched and replaced all TO-220s w/ heatsinks (Q105/Q155, Q107/Q157 with KSA1381-E, Q108/Q158 with KSC3503-E)
3) replaced C109/C159 (had Audio Note Caps but replaced with Panasonic FC) - ordered 8mm x 3.5mm LS (size should've been 10mm x 5mm LS)
4) replaced IC101 (prev. replaced IC101) with LT1006
5) replaced C106 (15pF) and C102 (330pF) with CDE 500V mica
6) replaced C201/C251 and C301/C351 (all 68pF) with CDE 500V mica
7) replaced all four LEDs on the front panel (Kingbright WF424EDT-orange, and WF424IDT-red (slightly shorter and a little brighter)
8) went over most of solder joints, especially on the wire connections - no foul smells, so I think I avoided any electrolyte seepage
Possible future change(s) - adding (4) 47uF/100V Nichicon Muse KZ with (4) 0.1uF bypass caps on the pre-amp board and change to an IEC plug
Summary of repairs:
I have changed all pre-drivers, drivers and outputs (KSA1381-E/C3503-E, NJW1302G/3281G and MJ21193/94G), both servo ICs, all electrolytic capacitors (pre-amp and output boards), all film/mica capacitors (pre-amp and output boards), all TO-92 and TO-220 transistors (pre-amp board) and checked all resistor values (replaced 4) and checked all diodes/zeners (all checked out fine).
Testing:
Since it was late when I got it back together, I waited until today to start it up and check bias and functionality. I was tired and didn't want a slip of the probe, or missing connection to give me a repair before it's fully working.
I recorded the start-up behavior yesterday morning, so I could compare the same start-up after replacing parts and compare any improvement.
Before these last repairs it took 2:45mins to reach 4mV of DC offset from a cold start-up (cold start is >3 hours off and unplugged). I ran the same test today and it settles below 4mV in 0:06, so more than 2-1/2mins quicker. Now the left channel mimics the right channel behavior - gets to 5mV in <10secs and tracks at +/-2mV. Not 100% sure if it was the IC replacement, changing the capacitors/transistors or touching up the solder joints. Both channels are working like I thought they should from the beginning and the amp has a really nice sound.
Now I get to spend some time listening and enjoying after spending all this work to bring this amp back from the dead. So far, pretty impressed - it's a fun, lively amp to listen to and I'm impressed with the bass drives my speakers pretty good in my small office. I think the combination of replacement Caps I used has brought it up a notch from what the previous owner had used, I think he was more concerned with the brand and not about the actual location and job to use the best possible capacitor.
Huge thanks to @bigskyaudio (upgrade guide and technical assistance here and via PM), @BSST (explaining the circuit and your fault finding suggestions), @anatech (You know these amps in-and-out, I appreciate your insights and suggestions) and @Phloodpants (read through all your threads/posts and website on Adcom amps, insightful stuff)
Now sit back and enjoy the music and the result of your efforts! The bipolar output classic ADCOM amps have excellent bass reproduction. I had a fair amount of customers that used them as subwoofer amps instead of those cold sterile Class D ones.
Hi bullittstang,
Yep. Specific brands and series of capacitors normally don't matter much as long as you are using a good brand. By good brand, I mean from an industrial perspective and not the really messed up audiophile world. The primary requirement beyond quality is that they fit!
There are circuit locations where capacitors designed for high frequency use perform better. Also to get the right lead spacing you normally need to go higher in voltage - and that is a good thing as dielectric performance seems to be best from 63 to 100V ratings. Certainly going from a 16 volt to 35 volt part will always improve capacitor performance.
Measures of capacitor quality. ESR is flipping useless, in a power supply what you see on an oscilloscope tells you more and is an accurate indicator. D-A, now this is where you can determine early if a capacitor is failing, and how difference capacitors stack up against each other. However in coupling capacitors it isn't that critical.
Your DC servo op amp will not change the sound, but may contribute lower noise. That's only compared to a really noisy op amp, and as long as the output is not filtered (an R-C network between the DC correction voltage and diff pair). Sometimes an op amp can develop a DC offset or drift, replacement is the only cure. The replacement must be unity gain stable.
Matching your diff pair, if done properly, should reduce DC offsets and also drop distortion. In other words, matching these transistors does improve sound quality which you may hear if the originals were not well matched. I certainly measure improvements almost every time. Matching complimentary drivers and pre-drivers sometimes does improve performance, as does matching sets of output transistors. Basically the less distortion the diff pair has to compensate for, the better off you are.
I designed a jig for matching differential transistors decades ago, and later gave it to this community. Chris (Phloodpants) did design a good PCB for it, as did a couple other people I know, and they perform quite well. I think Chris' design is in a GFA-585 thread. This is the only way to get close matches because the pair must be at exactly the same temperature to be compared. Unless to force the exact, same temperature for both parts some how, you cannot measure the Beta and select parts to match that way. Mind you, if you are careful, measuring each and matching that way in the same session is better than nothing.
I'm glad you were able to sort this out. You just performed work well beyond the understanding of most technicians these days.
-Chris
Yep. Specific brands and series of capacitors normally don't matter much as long as you are using a good brand. By good brand, I mean from an industrial perspective and not the really messed up audiophile world. The primary requirement beyond quality is that they fit!
There are circuit locations where capacitors designed for high frequency use perform better. Also to get the right lead spacing you normally need to go higher in voltage - and that is a good thing as dielectric performance seems to be best from 63 to 100V ratings. Certainly going from a 16 volt to 35 volt part will always improve capacitor performance.
Measures of capacitor quality. ESR is flipping useless, in a power supply what you see on an oscilloscope tells you more and is an accurate indicator. D-A, now this is where you can determine early if a capacitor is failing, and how difference capacitors stack up against each other. However in coupling capacitors it isn't that critical.
Your DC servo op amp will not change the sound, but may contribute lower noise. That's only compared to a really noisy op amp, and as long as the output is not filtered (an R-C network between the DC correction voltage and diff pair). Sometimes an op amp can develop a DC offset or drift, replacement is the only cure. The replacement must be unity gain stable.
Matching your diff pair, if done properly, should reduce DC offsets and also drop distortion. In other words, matching these transistors does improve sound quality which you may hear if the originals were not well matched. I certainly measure improvements almost every time. Matching complimentary drivers and pre-drivers sometimes does improve performance, as does matching sets of output transistors. Basically the less distortion the diff pair has to compensate for, the better off you are.
I designed a jig for matching differential transistors decades ago, and later gave it to this community. Chris (Phloodpants) did design a good PCB for it, as did a couple other people I know, and they perform quite well. I think Chris' design is in a GFA-585 thread. This is the only way to get close matches because the pair must be at exactly the same temperature to be compared. Unless to force the exact, same temperature for both parts some how, you cannot measure the Beta and select parts to match that way. Mind you, if you are careful, measuring each and matching that way in the same session is better than nothing.
I'm glad you were able to sort this out. You just performed work well beyond the understanding of most technicians these days.
-Chris