If you parallel R908 with 150k, Vcc will drop to ~17.5V.
Now that you've done that, consider using OPA1641 and OPA1642 opamps -- will cut noise density in half.
IC203, IC204 were LT1010 buffers, while the pinout differs from the BUF634A, the latter would offer a significant improvement (of the already excellent) THD+N%
Now that you've done that, consider using OPA1641 and OPA1642 opamps -- will cut noise density in half.
IC203, IC204 were LT1010 buffers, while the pinout differs from the BUF634A, the latter would offer a significant improvement (of the already excellent) THD+N%
Just replaced those two IC’s (201,202) with OPA134’s (what I had on hand).
Not blown away, but pretty amazed at how much more “open” (wider soundstage), “even”, the sound is. Some drumming/bass still seems a bit muffled/blurry, but mids and highs are clearer, vocals more forward.
Will try some 604’s for fun.
I’ve been on a couple of other boards and the common conversation regarding OPAMPS is “huh, so you want to change them just for the sake of changing them”… and “LOOK AT THE SPECS” vs. the argument which says it’s mostly subjective, and why not.
To which I’ll add: a LOT of research, technology has come along in the 30 years since this preamp was designed/built.
Of course newer OPAMPS are going to change the sound.
Contemplating a pair of Burson V5i-S OPAMPS too.
Thanks to all in this thread.
Not blown away, but pretty amazed at how much more “open” (wider soundstage), “even”, the sound is. Some drumming/bass still seems a bit muffled/blurry, but mids and highs are clearer, vocals more forward.
Will try some 604’s for fun.
I’ve been on a couple of other boards and the common conversation regarding OPAMPS is “huh, so you want to change them just for the sake of changing them”… and “LOOK AT THE SPECS” vs. the argument which says it’s mostly subjective, and why not.
To which I’ll add: a LOT of research, technology has come along in the 30 years since this preamp was designed/built.
Of course newer OPAMPS are going to change the sound.
Contemplating a pair of Burson V5i-S OPAMPS too.
Thanks to all in this thread.
My power supply with a somewhat beefier EI transformer and Jung-Didden regulators is attached via an umbilical. If one were very creative they could instead use an Elvee "de-noiser" circuit which has the extremely low noise and output impedance of the Jung. See his thread elsewhere.
At some point I would like to compare the hotted-up GFP565 with some of the high end preamps discussed in the hi-fi mags.
At some point I would like to compare the hotted-up GFP565 with some of the high end preamps discussed in the hi-fi mags.
Here’s a question for the gurus here.
”Install .1uF/200V KEMET C330C104K2R5TA from IC201 (IC202) pin 4 (V-) and pin 7 (V+) to ground.”
Most OPAMP mfrs show some kind of (isolation?) cacitors from pins 4 and 7 to ground. OPA2228 data sheet from TI says 10 nF (not 100nF as above), BUT…
In looking at IC201 and IC202 (gain section), there are already TWO capacitors in parallel linked to those pins, e.g. - C119 and 120, then C121, 122, a 1 uF stacked metallized polyester cap and an electrolytic 470uF/35V cap.
What, if anything will be gained by placing a 10 or 100 nF (0.1, 0.01 uF) cap in parallel to those two?
Or was the original comment pertaining to an earlier model?
Thanks in advance.
Since I own two of the GFP-565’s, I’ve done updating of the PS on both, replaced PE with PP capacitors where ADCOM hadn’t already done so, one unit I’m using Nichicon UFG caps, another UKA. Might be my aging ears but the UKA’s seem to bring more “presence”, expanded soundstage BEFORE changing from OEM 6A IC’s
The UKA preamp I’ve put IC sockets in place and have been playing with diferent OP-amps. I like the 134/2134 set a LOT, am awaiting delivery of 604/2604 IC’s.
If/when I get a GFP-915 I’ll likely sell the UFG one.
”Install .1uF/200V KEMET C330C104K2R5TA from IC201 (IC202) pin 4 (V-) and pin 7 (V+) to ground.”
Most OPAMP mfrs show some kind of (isolation?) cacitors from pins 4 and 7 to ground. OPA2228 data sheet from TI says 10 nF (not 100nF as above), BUT…
In looking at IC201 and IC202 (gain section), there are already TWO capacitors in parallel linked to those pins, e.g. - C119 and 120, then C121, 122, a 1 uF stacked metallized polyester cap and an electrolytic 470uF/35V cap.
What, if anything will be gained by placing a 10 or 100 nF (0.1, 0.01 uF) cap in parallel to those two?
Or was the original comment pertaining to an earlier model?
Thanks in advance.
Since I own two of the GFP-565’s, I’ve done updating of the PS on both, replaced PE with PP capacitors where ADCOM hadn’t already done so, one unit I’m using Nichicon UFG caps, another UKA. Might be my aging ears but the UKA’s seem to bring more “presence”, expanded soundstage BEFORE changing from OEM 6A IC’s
The UKA preamp I’ve put IC sockets in place and have been playing with diferent OP-amps. I like the 134/2134 set a LOT, am awaiting delivery of 604/2604 IC’s.
If/when I get a GFP-915 I’ll likely sell the UFG one.
This thread inspired me to pull my 565 off the shelf, where its been sitting for the past two years. My GFP-555 sounded far better so I knew there was something wrong and I benched the 565 until I had time to plot an upgrade path. I put the 565 with various preamps, amps, speakers, sources, cables - it sounded way worse than I remember. I'm looking forward to putting in the parts order laid out from Big Sky, but I'm going to tweak it a little bit. The LKG caps are on back-order nationwide so I'm going to splurge on Mundorfs. They're 4-5x the cost but I had really good results with them in a Sumo rebuild. I'll probably go with the Nichicon UKZ for the rest of the cans.
JackinNJ mentioned the transformer was a weak link - any recommendations for installing a better one? Maybe a decent toroid? (Something that will fit in the stock case, hopefully. I know the outboard power is the right way to reduce noise in many cases but having another box to deal with isn't optimal for me)
JackinNJ mentioned the transformer was a weak link - any recommendations for installing a better one? Maybe a decent toroid? (Something that will fit in the stock case, hopefully. I know the outboard power is the right way to reduce noise in many cases but having another box to deal with isn't optimal for me)
You could mount the transformer on the exterior of the chassis, but you'll need one which is enclosed.
I prefer an EI transformer to a toroid...but I used an Amveco !
You could use a switcher with a pair of high quality regulators AND a "pseudo-ground". Too bad the DIYAUDIOSTORE is out of Jan Didden's "Silent Switcher".
That is not a bad idea. Build a 18 volt bipolar supply in an external case. The BA2018 by nelson pass uses those voltages. Leave the stock transformer in the case but unconnected. I was going to suggest against messing with the upgraded power supply as I have a BA2018 which I appreciate much more than the Adcom GFP565 (that isn't to say the Adcom was bad).
I would prefer to keep everything inside the case. I have nothing against those folks who want 10/10 out of their audio equipment and do remote power supplies, but the only time I ever consider heavy customization is when the shell/chassis are toast. I have one more 565 sitting at a friend's house and its pretty beat up, so I'll experiment on that one. The case is rusted out so I'll probably pick something neat to put the chassis in, like an old slimline PC. The volume knob is bent at an angle, so I'll look into doing a remote vol. If I go ahead with all that I'll try the outboard PS.
I'm curious, why do you prefer the EI transformer to a toroid? I've never had any kind of tf buzz out of toroids, but I have several units here with buzzing frame transformers. Most of them are NAD / Proton, but I've run into others.
You lost me there. Sorry, I'm still a rookie, 4 years into this hobby.
I had a "fun" time sourcing the parts for this. Parts shortages everywhere, so I had to split things up between two vendors and I still had trouble finding everything I needed. To get the project moving required changing a few things. Once I have all the parts are installed and I'm sure I'm not sending people down the wrong path I'll post a list of what I changed.
One package showed up a couple of hours ago, but the lion's share of parts just shipped while I was typing this post. In todays package were my power supply caps - the 6800uF Nichicon KG and the 1.0uF film bypass caps, along with the 4.7uF films at C207/C208. I pulled all six caps and tested them as I went. The main caps were still good but they were 6180uF each, with ESR approaching 3 ohms, quite a bit higher than I like to see in those caps. No leaks though, so I'd say they still have some years left in them, and I doubt they were affecting the sound much.
The bypass caps for those two were a different story. According to the Adcom schematic they're supposed to be stacked metal film but the package looks like a standard plastic film. They tested out as bad, both of them. In went the 1.0uF poly films. Last up were the 4.7uF films for the outputs. The original parts tested out fine, but the ESR was higher than I've ever seen in a metaled poly cap, and also uneven. One channel was 1.8 ohms, the other was 3.6 ohms.
Before this the preamp was extremely fatiguing. I couldn't listen to it for very long before I got a headache, and I'm not kidding about that. I've got the thing up and running now and at first it appeared that just this little bit of work solved the issue with the sound. Even with the meager 60wpc from my GFA-535 the sound was quite good. I went through an entire vintage Sheffield Labs disc and the sound started clean and crisp almost to a fault, then the soundstage got deeper, and by the end of the album the sound had sweetened up quite a bit. I switched to a Cat Stevens playlist, and I heard some distortion in the bass. I don't know if this is an artifact of the caps breaking in, it could also be the amp thumbing its nose at me and demanding new caps for its own power supply.
The rest of todays package is just ceramics and a few resistors. When the rest of the electrolytics arrive I'll get back to work on this.
One package showed up a couple of hours ago, but the lion's share of parts just shipped while I was typing this post. In todays package were my power supply caps - the 6800uF Nichicon KG and the 1.0uF film bypass caps, along with the 4.7uF films at C207/C208. I pulled all six caps and tested them as I went. The main caps were still good but they were 6180uF each, with ESR approaching 3 ohms, quite a bit higher than I like to see in those caps. No leaks though, so I'd say they still have some years left in them, and I doubt they were affecting the sound much.
The bypass caps for those two were a different story. According to the Adcom schematic they're supposed to be stacked metal film but the package looks like a standard plastic film. They tested out as bad, both of them. In went the 1.0uF poly films. Last up were the 4.7uF films for the outputs. The original parts tested out fine, but the ESR was higher than I've ever seen in a metaled poly cap, and also uneven. One channel was 1.8 ohms, the other was 3.6 ohms.
Before this the preamp was extremely fatiguing. I couldn't listen to it for very long before I got a headache, and I'm not kidding about that. I've got the thing up and running now and at first it appeared that just this little bit of work solved the issue with the sound. Even with the meager 60wpc from my GFA-535 the sound was quite good. I went through an entire vintage Sheffield Labs disc and the sound started clean and crisp almost to a fault, then the soundstage got deeper, and by the end of the album the sound had sweetened up quite a bit. I switched to a Cat Stevens playlist, and I heard some distortion in the bass. I don't know if this is an artifact of the caps breaking in, it could also be the amp thumbing its nose at me and demanding new caps for its own power supply.
The rest of todays package is just ceramics and a few resistors. When the rest of the electrolytics arrive I'll get back to work on this.
Some of the caps I wound up buying from reseller in Germany by way of ePay. Not just for the 565 but for other projects.
When Mouser Diigikey Allied West coast and other guys here in the US, UK even, are back ordered for months or years, it’s not just us hobbyists who are suffering.
keep us posted.
When Mouser Diigikey Allied West coast and other guys here in the US, UK even, are back ordered for months or years, it’s not just us hobbyists who are suffering.
keep us posted.
C901~C904 = .1uF/500V KEMET monolithic ceramic C330C104KCR5TA
I only had three spots for the .1uf caps. I couldn't find a C902 so I suspect that there isn't one. There was only one cap in place out of the three originally
That is really odd. Your 565 is a different revision from mine then. C901-C904 all show up on my board as large blue mylar 472pF parts. If I've learned anything over the past few years these caps are bypassing the rectifying diodes to prevent (reduce) high frequency oscillation. Lots of power circuits forgo those so they're not really necessary, but in a preamp or CD player they're desirable. And you have to have the 4th diode or your preamp isn't going to work too well, so why bypass three out of four diodes and forget the fourth?
Of course the schematic is no help because C901 shows up in the parts list as a 472pF, which is the small teal colored disc on the transformer configuration board, and there's no mention of the diode caps. The pictorial of the board is even more useless, because it shows C901 and R901, but not the other three caps.
FWIW you can add that 4th diode yourself : put the cap on top of the diode, wrap the cap leads around the diode leads, and solder it in place. Its very easy and I see no harm in doing it. If you want to be fancy, leave an air gap between the two and put a drop of hot melt glue in between.
One final note on this - you have an R901, but you're missing C901. I have all four caps, C901-C904, but my board is missing R901.
Install .01uF/250V Vishay WYO103MCMCF0KR Class X2 EMI suppression capacitor across AC5 to AC6
Mine didn't have this capacitor so I went ahead and installed it. I just wrapped the leads around the AC5 and AC6 and soldered them.
Thats one way to do it. Take a look at the transformer config board I mentioned above.
As long as the jumper for 120V operation is in place, you can delete the cap in the 120V position and install that new suppression disc capacitor in the 250V position. They're equivalent as far as circuit position and no different than jumping AC5 and AC6, just a little more elegant. Regardless of whichever you choose, I think its safe to delete the cap in the 120v location.
And there are people here with a lot more experience than me, so I hope they jump in and correct me if any of the points I've brought up aren't right. I'm still learning myself.
Well this is interesting.
I've gone through all the electrolytics so far and gotten great results. The sound is exceeding my expectations and I'll have a full writeup as soon as I get the resistors installed.
However...
I've just finished the replacement on the C901-904 caps I mentioned in my last post. Those are 472pF parts, blue mylar discs. All four of mine were defective. Using my Peak ESR70 meter I saw faults with each of them. Usually the part shows up as soon as I have the test clips attached but with these I had to force it using the 'test' button. All showed "open circuit / low capacitance". Following up with my Fluke DMM and checking for resistance I got "OL" on each part. Using my BK 815 meter, I finally got a result out of them. All were in the high 500pF to low 600pF range.
So maybe they're bad, maybe not. I shouldn't have to dig equipment out to confirm these things when my first two tools have worked in every other case since I've bought them.
And one more "however"...
Now I have a faint transformer buzz, where this thing was completely quiet before. I can't exactly blame C901-C904 though. I tested this preamp at every step of the way on the electrolytic replacement - EXCEPT - when I did C913 and C921. I wish I had.
So, more experienced peoples, did any of these parts replacements affect this transformer?
I've gone through all the electrolytics so far and gotten great results. The sound is exceeding my expectations and I'll have a full writeup as soon as I get the resistors installed.
However...
I've just finished the replacement on the C901-904 caps I mentioned in my last post. Those are 472pF parts, blue mylar discs. All four of mine were defective. Using my Peak ESR70 meter I saw faults with each of them. Usually the part shows up as soon as I have the test clips attached but with these I had to force it using the 'test' button. All showed "open circuit / low capacitance". Following up with my Fluke DMM and checking for resistance I got "OL" on each part. Using my BK 815 meter, I finally got a result out of them. All were in the high 500pF to low 600pF range.
So maybe they're bad, maybe not. I shouldn't have to dig equipment out to confirm these things when my first two tools have worked in every other case since I've bought them.
And one more "however"...
Now I have a faint transformer buzz, where this thing was completely quiet before. I can't exactly blame C901-C904 though. I tested this preamp at every step of the way on the electrolytic replacement - EXCEPT - when I did C913 and C921. I wish I had.
So, more experienced peoples, did any of these parts replacements affect this transformer?
Not a bad idea to just bypass the diode using the leads themselves. It has been a while since I worked on the GFP-565 The preamp did sound nice. I don't know why there would be three instead of four caps on my board. I do not have the preamp anymore. It was a great preamp but I have noticed that I tend to not use a lot of the features on these types of preamps. I now use a BA2018 with three inputs and some nice components.
What frequency is the buzz? A stupid check, I have tested preamps by setting them on my amplifier before. Last I remember, the GFP-565 is fairly silent and
well-shielded. However, I did chase a buzz problem on a Hafler DH-101 that I rebuilt only to figure out that setting it near my amp was the reason for the buzzing. This was after a couple of hours of scratching my head. Are all your solder joints good? Maybe go around and tap things with a chopstick to see if there is a change in the buzzing sound. It is unlikely that a new component is bad although it does happen. Are all the grounds making good contact etc.?
What frequency is the buzz? A stupid check, I have tested preamps by setting them on my amplifier before. Last I remember, the GFP-565 is fairly silent and
well-shielded. However, I did chase a buzz problem on a Hafler DH-101 that I rebuilt only to figure out that setting it near my amp was the reason for the buzzing. This was after a couple of hours of scratching my head. Are all your solder joints good? Maybe go around and tap things with a chopstick to see if there is a change in the buzzing sound. It is unlikely that a new component is bad although it does happen. Are all the grounds making good contact etc.?
The Results are in and the verdict is: do this mod.
I did this upgrade in several stages, evaluating the sound in between stages. I also checked each part that was removed to see how much drift there was from the expected values. I'll put those notes in a later post. I didn't do any of the optional steps, as I don't use headphones or tape decks, but I'll probably do those at a later date just for the sake of completeness.
Listening equipment: Adcom GFA-535 amplifier, original condition. CD input supplied with digital source from late model iMac. All files are 16/44 ALAC. Speakers are Canton LE102, unmodified, in near-field position approximately 6 feet apart, and 4 feet off my seating position, on Sanus wood stands.
Listening criteria: I already have the Hafler IRIS if I want a preamp without tone controls. I want my Adcom preamps because they can offer a nice bass EQ when I want it. Also, since non-audiophiles expect bass/treble controls, I want to see just how much this improves my EQ. Therefore, I used the Normal outputs and Tone was enabled. Treble straight up, bass at the 2:30 position. I also selected the Contour button. These steps compensate for the lower frequency deficit endemic to bookshelf speakers.
Suppliers: Mouser, PartsConnexion.
Stage One:
C207, C208 = 4.7uF/250V Panasonic ECW-F2475JA polypropylene film
These were out of stock. I used the ECW-FD2475J instead, the main difference being the higher 450V rating. There is nothing wrong with going much higher on any film cap voltage, in my experience, other than size constraints. These installed easily. Since the originals were oddball Wima-style boxes these new Panasonics match every other PP cap on the board for a better visual presentation as well.
C907, C908 = 1.0uF/450V Panasonic ECW-FD2W105KB polypropylene film
I used the ECW-FD2W105JB instead. The difference is the JB is a 5 percent tolerance vs the 10 percent of the KB from the OP. These films bypass C905 and C906, and were bad on my GFP-565. As Mike mentioned above, make sure these films are replaced while the much larger 905/906 pair are out of the board. I've installed mine from the bottom due to size constraints.
C905, C906 = 6800uF/50V Nichicon KG Gold Tune electrolytic LKG1H682MESBAK (30x35mm)
Both Mouser and PC were out of stock on these, but PC had stock of the 6800/63V, p/n LKG1J682MESBAK. The existing 50V caps are already more than double the rail voltage but the higher 63v part isn't going to hurt this project. They were also less than half the price of the Mundorf's I originally wanted to use.
Result: the preamp is now listenable.The worst of the dirt is gone from the sound, and I attribute that mainly to the dead 1.0uF films. If they weren't doing their job then there was high frequency trash coming out of the power supply. Or at least that's what I think was happening. All I know is that I could have stopped right there and this preamp would have been adequate. I'm never happy with 'adequate' so I kept going.
Stage Two:
C917, C918 = 1000uF/50V Nichicon Muse KZ electrolytic
Installed.
C919, C920 = 1000uF/25V Nichicon Muse KZ electrolytic
Installed. But afterwards I realized I should have just bought the same 50V parts used in C917 and C918. The extra voltage room has no effect on the surface, and it simplifies ordering. I'd also have been a total of 4 parts closer to a 10 part discount, which is important if you want to have stock on hand for future products.
C914 = 1000uF/25V Panasonic FC electrolytic (short style)
I couldn't find that part so I ordered the rated value in a Nichicon KA. This was a mistake from a physical POV. The KA is a tiny cap and not only would have looked ridiculous in C914, it also had a much smaller lead spacing and wouldn't have sat on the board, which meant hot glue would be needed. I'm not a fan, so instead I picked a 2200uF 35V Nichicon FG "Fine Gold" out of my miscellaneous pile. It fits perfectly and looks appropriate between the KZ caps.
Install Wakefield Kapton 175-6-220P thermal insulators on Q901 and Q902
Installed while I had C917 and C918 out of the board.
Result: What remained of the distortion has been cleaned up. Things are clear, but the soundstage isn't that good. Right now its like listening to the usual compromises you might expect from an Adcom GTP-350 or GTP-400 tuner preamp. Clear but not engaging. The music also seems a little dysfunctional, as if I'm listening to all the instruments slightly out of sync with each other.
After a few minutes the coherence has improved. Everything seems lined up properly. Is this the "PRAT" nonsense I see in reviews from time to time?
Now I can evaluate the sound based on familiar tracks, so I'm using a favorite jazz test CD. Snare is slightly distant and has a shimmer to it, but the bass is tight and controlled. . Right now, this preamp is competitive with my GFP-555, which is still in unmodified factory condition. Excellent but not what I expect from this supposed top of the line Adcom preamp. Lets keep going.
Stage Three:
C121B, C122B = 470uF/50V Nichicon Muse KZ electrolytic
Installed.
Result: FINALLY things are coming together. Glistening highs. Rim shots, strings, cymbals, all are very present and very detailed yet pleasant. This is what I expected from this preamp.
Stage Four:
C121, C122, C401, C403 = 470uF/50V Nichicon Muse KZ electrolytic
Installed.
Result: Further refinement is evident all around, but not as easily identified as previous steps. Slightly less boomy in the mid-bass, perhaps. The most noticeable effects are an improvement in three dimensions, and there is far more sense of "air" in the material.
Stage Five:
C107, C108 = 2200uF/25V Nichicon FG electrolytic
These went out of stock at PC before my order went through. My rep contacted me to ask what substitute I would want. I suggested getting the same value in Nichicon KA. This turned out to be sort of a mistake, as the footprint for the KA is so much smaller. I should have asked for KZ. It'll work though. This pair are part of the phono section so the sound should be excellent. I'm short on extra turntables here and I don't want to tear apart my main system, so I'll have to leave this for now.
C913 = 22uF/50V Panasonic FC electrolytic
I used the Nichicon KZ 22uf/50V
C921 = 100uF/25V Panasonic FC electrolytic
I used the Nichicon KA 100uF/25V.
C901~C904 = .1uF/500V KEMET monolithic ceramic C330C104KCR5TA
Installed.
Result: Slight transformer buzz afterwards. Sound is still excellent but that buzz really ticks me off. I'll figure that out later.
Phase Six:
Install .01uF/250V Vishay WYO103MCMCF0KR Class X2 EMI suppression capacitor across AC5 to AC6
As I mentioned in my earlier post, I deleted the existing buzz-killer mylar disc from the transformer configuration board, and put the new Kemet disc in the nearby "250v" position. I installed it from beneath for a cleaner appearance. Either side will work.
R117 (R118) = 470Ω, ¼W, 1% PRP PR9372 metal film (phono)
PC carries these. Installed.
R121 (R122), R217A (R218A), R237 (R238), R239 (R240) = 100Ω, ¼W, 1% PRP PR9372 metal film
PC was out of stock, and Mouser doesn't carry PRP, so I bought TE Holco H8100RFZA from Mouser. They fit, and specs are equivalent.
R213 (R214) = 22.1k, ½W, 1% Holco H4 metal film (feedback)
Mouser had stock of these. No problems installing them.
Change IC201 (IC202) to OPA134 (Change R908 to 12.1k) or OPA604.
I skipped this.
Install .1uF/200V KEMET C330C104K2R5TA from IC201 (IC202) pin 4 (V-) and pin 7 (V+) to ground.
Lol... Mike said he didn't do this step because he only bought one cap. I bought two. Then I realized it was two caps per IC, and I had two ICs. So mine isn't done either.
Result: No change in the sound, and I didn't expect one. Each prior step led to immediate results and seemed to skip "break-in" for the most part. Yet as I sit here and fine tune this report for you all, the sound does seem to have matured. I'm listening to the Jeff Beck/ Rod Stewart "Truth" album. Excellent sound. Switch to Curtis Mayfield, "Superfly". This is amazing. The instruments are so distinct. Mayfield's vocals are like silk.
Conclusion: The Big Sky mod is an excellent upgrade. I was so disappointed in this preamp before but right now I don't want to stop listening through it. If I had a remote on this I'd be using it in my main system. I'm looking forward to putting the GFP-565 with a far more powerful amp and larger speakers.
I did this upgrade in several stages, evaluating the sound in between stages. I also checked each part that was removed to see how much drift there was from the expected values. I'll put those notes in a later post. I didn't do any of the optional steps, as I don't use headphones or tape decks, but I'll probably do those at a later date just for the sake of completeness.
Listening equipment: Adcom GFA-535 amplifier, original condition. CD input supplied with digital source from late model iMac. All files are 16/44 ALAC. Speakers are Canton LE102, unmodified, in near-field position approximately 6 feet apart, and 4 feet off my seating position, on Sanus wood stands.
Listening criteria: I already have the Hafler IRIS if I want a preamp without tone controls. I want my Adcom preamps because they can offer a nice bass EQ when I want it. Also, since non-audiophiles expect bass/treble controls, I want to see just how much this improves my EQ. Therefore, I used the Normal outputs and Tone was enabled. Treble straight up, bass at the 2:30 position. I also selected the Contour button. These steps compensate for the lower frequency deficit endemic to bookshelf speakers.
Suppliers: Mouser, PartsConnexion.
Stage One:
C207, C208 = 4.7uF/250V Panasonic ECW-F2475JA polypropylene film
These were out of stock. I used the ECW-FD2475J instead, the main difference being the higher 450V rating. There is nothing wrong with going much higher on any film cap voltage, in my experience, other than size constraints. These installed easily. Since the originals were oddball Wima-style boxes these new Panasonics match every other PP cap on the board for a better visual presentation as well.
C907, C908 = 1.0uF/450V Panasonic ECW-FD2W105KB polypropylene film
I used the ECW-FD2W105JB instead. The difference is the JB is a 5 percent tolerance vs the 10 percent of the KB from the OP. These films bypass C905 and C906, and were bad on my GFP-565. As Mike mentioned above, make sure these films are replaced while the much larger 905/906 pair are out of the board. I've installed mine from the bottom due to size constraints.
C905, C906 = 6800uF/50V Nichicon KG Gold Tune electrolytic LKG1H682MESBAK (30x35mm)
Both Mouser and PC were out of stock on these, but PC had stock of the 6800/63V, p/n LKG1J682MESBAK. The existing 50V caps are already more than double the rail voltage but the higher 63v part isn't going to hurt this project. They were also less than half the price of the Mundorf's I originally wanted to use.
Result: the preamp is now listenable.The worst of the dirt is gone from the sound, and I attribute that mainly to the dead 1.0uF films. If they weren't doing their job then there was high frequency trash coming out of the power supply. Or at least that's what I think was happening. All I know is that I could have stopped right there and this preamp would have been adequate. I'm never happy with 'adequate' so I kept going.
Stage Two:
C917, C918 = 1000uF/50V Nichicon Muse KZ electrolytic
Installed.
C919, C920 = 1000uF/25V Nichicon Muse KZ electrolytic
Installed. But afterwards I realized I should have just bought the same 50V parts used in C917 and C918. The extra voltage room has no effect on the surface, and it simplifies ordering. I'd also have been a total of 4 parts closer to a 10 part discount, which is important if you want to have stock on hand for future products.
C914 = 1000uF/25V Panasonic FC electrolytic (short style)
I couldn't find that part so I ordered the rated value in a Nichicon KA. This was a mistake from a physical POV. The KA is a tiny cap and not only would have looked ridiculous in C914, it also had a much smaller lead spacing and wouldn't have sat on the board, which meant hot glue would be needed. I'm not a fan, so instead I picked a 2200uF 35V Nichicon FG "Fine Gold" out of my miscellaneous pile. It fits perfectly and looks appropriate between the KZ caps.
Install Wakefield Kapton 175-6-220P thermal insulators on Q901 and Q902
Installed while I had C917 and C918 out of the board.
Result: What remained of the distortion has been cleaned up. Things are clear, but the soundstage isn't that good. Right now its like listening to the usual compromises you might expect from an Adcom GTP-350 or GTP-400 tuner preamp. Clear but not engaging. The music also seems a little dysfunctional, as if I'm listening to all the instruments slightly out of sync with each other.
After a few minutes the coherence has improved. Everything seems lined up properly. Is this the "PRAT" nonsense I see in reviews from time to time?
Now I can evaluate the sound based on familiar tracks, so I'm using a favorite jazz test CD. Snare is slightly distant and has a shimmer to it, but the bass is tight and controlled. . Right now, this preamp is competitive with my GFP-555, which is still in unmodified factory condition. Excellent but not what I expect from this supposed top of the line Adcom preamp. Lets keep going.
Stage Three:
C121B, C122B = 470uF/50V Nichicon Muse KZ electrolytic
Installed.
Result: FINALLY things are coming together. Glistening highs. Rim shots, strings, cymbals, all are very present and very detailed yet pleasant. This is what I expected from this preamp.
Stage Four:
C121, C122, C401, C403 = 470uF/50V Nichicon Muse KZ electrolytic
Installed.
Result: Further refinement is evident all around, but not as easily identified as previous steps. Slightly less boomy in the mid-bass, perhaps. The most noticeable effects are an improvement in three dimensions, and there is far more sense of "air" in the material.
Stage Five:
C107, C108 = 2200uF/25V Nichicon FG electrolytic
These went out of stock at PC before my order went through. My rep contacted me to ask what substitute I would want. I suggested getting the same value in Nichicon KA. This turned out to be sort of a mistake, as the footprint for the KA is so much smaller. I should have asked for KZ. It'll work though. This pair are part of the phono section so the sound should be excellent. I'm short on extra turntables here and I don't want to tear apart my main system, so I'll have to leave this for now.
C913 = 22uF/50V Panasonic FC electrolytic
I used the Nichicon KZ 22uf/50V
C921 = 100uF/25V Panasonic FC electrolytic
I used the Nichicon KA 100uF/25V.
C901~C904 = .1uF/500V KEMET monolithic ceramic C330C104KCR5TA
Installed.
Result: Slight transformer buzz afterwards. Sound is still excellent but that buzz really ticks me off. I'll figure that out later.
Phase Six:
Install .01uF/250V Vishay WYO103MCMCF0KR Class X2 EMI suppression capacitor across AC5 to AC6
As I mentioned in my earlier post, I deleted the existing buzz-killer mylar disc from the transformer configuration board, and put the new Kemet disc in the nearby "250v" position. I installed it from beneath for a cleaner appearance. Either side will work.
R117 (R118) = 470Ω, ¼W, 1% PRP PR9372 metal film (phono)
PC carries these. Installed.
R121 (R122), R217A (R218A), R237 (R238), R239 (R240) = 100Ω, ¼W, 1% PRP PR9372 metal film
PC was out of stock, and Mouser doesn't carry PRP, so I bought TE Holco H8100RFZA from Mouser. They fit, and specs are equivalent.
R213 (R214) = 22.1k, ½W, 1% Holco H4 metal film (feedback)
Mouser had stock of these. No problems installing them.
Change IC201 (IC202) to OPA134 (Change R908 to 12.1k) or OPA604.
I skipped this.
Install .1uF/200V KEMET C330C104K2R5TA from IC201 (IC202) pin 4 (V-) and pin 7 (V+) to ground.
Lol... Mike said he didn't do this step because he only bought one cap. I bought two. Then I realized it was two caps per IC, and I had two ICs. So mine isn't done either.
Result: No change in the sound, and I didn't expect one. Each prior step led to immediate results and seemed to skip "break-in" for the most part. Yet as I sit here and fine tune this report for you all, the sound does seem to have matured. I'm listening to the Jeff Beck/ Rod Stewart "Truth" album. Excellent sound. Switch to Curtis Mayfield, "Superfly". This is amazing. The instruments are so distinct. Mayfield's vocals are like silk.
Conclusion: The Big Sky mod is an excellent upgrade. I was so disappointed in this preamp before but right now I don't want to stop listening through it. If I had a remote on this I'd be using it in my main system. I'm looking forward to putting the GFP-565 with a far more powerful amp and larger speakers.
This preamp does exactly what I want, I just wish it had a remote control for volume. Thats the only thing I would change now.
What chassis did you use for your BA2018?
The buzz seems the same as every other frame transformer buzz I've ever experienced. This one is shielded and I'm assuming its potted under that shield, so I doubt the transformer is the trouble. I cleaned every pad prior to installing new parts. All the new joints were checked and in some cases I reflowed them after I trimmed the leads just because I can't stand how they look if I accidentally nip the solder when I'm doing the trim. Its ridiculous, I know, but I can't put the cover on knowing there's flat spots on the joints. OCD I guess.
I'm wondering if its possible that the diodes themselves could be contributing to the buzz. I only have UF4003 on hand, not the IN4003, and I don't know if that would be an appropriate replacement. There's always so much more to learn.
I made a chassis out of walnut. I was worried it would make some noise not being in a completely metal chassis but it is pretty quiet. It has a Salas I-select input switcher which I like because of the sealed relays and it also has a Khozmo stepped attenuator. I have a helios bipolar power supply which has worked well and has an adjustable voltage up to about 19vdc. It is all manual with no tone control. I use a DAC as well as Deezer and control volume from my computer.
I used to do the whole CD thing but really like the convenience of not having to get up. The only thing is it tends to require a little discipline to sit through a whole album when you have the option to listen to anything at your fingertips. Are you using the iMac for the digital to analog conversion?
It looks like there is a ALPS-style volume pot on the GFP-565. I know that ALPS makes motor-controlled pots but I don't suspect that would be an easy project. Some people will have the volume inside of a separate box from the linestage which is actually pretty neat (great for DIY preamps) but then you have more signal leads and complexity. Not very practical with your preamp.
Your write-up is excellent. I can appreciate your OCD. I have done that a few times as well. My OCD thing is to lift resistors slightly off of the board and use a lead bending index. Silly and no real practical advantage other than clamping leads on for testing values, voltages.
Maybe it would be worth rebuilding the Adcom GFA-535? I rebuilt a first gen GFA-555 and it performed very well. I didn't spend much time with that amp before the rebuild but I can say that it was very accurate and fast sounding afterward. I sold it to a friend as I am always looking for more projects. I believe Big Sky has an upgrade posted here for the 555 that should translate to the 535 in some ways.
I used to do the whole CD thing but really like the convenience of not having to get up. The only thing is it tends to require a little discipline to sit through a whole album when you have the option to listen to anything at your fingertips. Are you using the iMac for the digital to analog conversion?
It looks like there is a ALPS-style volume pot on the GFP-565. I know that ALPS makes motor-controlled pots but I don't suspect that would be an easy project. Some people will have the volume inside of a separate box from the linestage which is actually pretty neat (great for DIY preamps) but then you have more signal leads and complexity. Not very practical with your preamp.
Your write-up is excellent. I can appreciate your OCD. I have done that a few times as well. My OCD thing is to lift resistors slightly off of the board and use a lead bending index. Silly and no real practical advantage other than clamping leads on for testing values, voltages.
Maybe it would be worth rebuilding the Adcom GFA-535? I rebuilt a first gen GFA-555 and it performed very well. I didn't spend much time with that amp before the rebuild but I can say that it was very accurate and fast sounding afterward. I sold it to a friend as I am always looking for more projects. I believe Big Sky has an upgrade posted here for the 555 that should translate to the 535 in some ways.
That seems like a nice setup. If you have any remaining concerns about noise perhaps you could look into some RF shielding, maybe copper sheet?
The iMac is doing the conversion. I wouldn't mind having a DAC but I can't see spending so much on a single component. I can buy an Oppo Blu-ray player here with an excellent chipset for under $350, or I can spend $1k for an equiivalent chipset in a separate chassis. It just doesn't seem right to me. Are there any deals out there that you know of?
I was hoping the TauDAC would be more widely available by now, but apparently its still in the hands of just a few people who are evaluating it.
ALPS makes a direct, motorized replacement for the Adcom pot, but as you said its not an easy job. Someone did it on a Carver preamp, if I remember right, and it had an LED in the volume knob. Very slick looking job when it was done. I need to chase that down.
Thanks for the review. I tried to be detailed without boring people, and I think it was the right thing for me to do the work in stages, along with my perceptions. Too many times have I just pulled all the parts out of a project and then replaced everything at once. I have no idea what affected the sound then.
I hadn't planned on upgrading the 535 right now because in stock form it actually combines really well with the GFP-555. I'd planned to just leave that as a matched pair, and I'll put the GFP-565 with one of my bigger Adcom amps. I'll look for the Big Sky 555 upgrade, thanks for mentioning that.
Last edited: