New to me Au 777A and restoration questions
- Thread starter toneriderMUC
- Start date
toneriderMUC
Active Member
Mhm, maybe a little brighter. But only a little, the amp was very alive and fresh sounding before the recap, too. Highs now are very liquid and clear while not being harsh. I have experienced burn in when I recapped my Yamaha NS 690 II speakers. Within the first weeks they went from harsh and almost brittle to gorgeous. I thought I had ruined them on my first listen after the recap and now I have seen a pair of NS 1000M's come and go, because I just liked the NS 690s better. On amplifiers the differences have been more subtle so far.
That is good to read!! I have maybe 3-4 hours on it now and will watch how it develops. That textured and rich midrange makes the 777A special to me, I only have heard that from tubes before, and I really hope it returns. At the same time I would not be surprised if the 50 year old and most likely out of spec components brought a certain harmonic distortion to the table that I perceived as nice. The Au 777A actually sonically reminds me a lot of the tubed Au 70 that I once had. I like the 777A better, for what its worth.
toneriderMUC
Active Member
To not see the forest for the trees. Or: sometimes the solution is just too simple.
Careful, rambling ahead!
I gave the amplifier more hours for potential burn in. Still, it sounded just wrong. I noticed that some songs sounded really okay. And some completely off. Something was wrong, mostly with the mids.Sometimes the guitar was only audible faint in the back, sometimes the singing voice very weak. Neil Young & Crazy Horse's majestic "Cortez the killer" sounded like coming out of a tin can. Hard and empty.
I started looking for the fault.
Reflowed solder joints, checked all parts I replaced for correctness.
Took out the tone board again, checked everything here.
Started checking the old wires in case one had cracked from moving around.
Still, the sound remained the same. I felt awful and was afraid that I had ruined a wonderful amplifier.
Read horror stories from people complaining about not liking their recapped amps.
Got more desperate, started reading patfronts much discussed 45 page capacitor shootout in the Marantz forum.
Started suspecting the Nichicon UPWs in the power supply for being accountable for the hard sound.
Started thinking about replacing them for Elna Silmics for that magic fix of my problem.
Started thinking about putting the old grey Elna coupling caps back in.
Started questioning my choice of Y-rated KSC1815 for the C-rated C458LGs throughout the unit.
Actually ordered really hard to find KSC945G from france, which haven't arrived yet.
...and finally had the sadly obvious idea to start checking the channels individually and remembered the stereo/mono selector on the 777A that offers several possiblites of channel composition (in my defense: front plate is still not mounted and just the bare switches are looking at me).
Well. Turns out I was listening to the right channel in MONO all the time.
The pot must have moved one position when pulling the knob.
Back to normal stereo and my 777A is back to his hearty self. The mids are there and they sing! Well, you can tell I am relieved!!
I really need to do some listening now, because now it sounds gorgeous and, as I am still waiting for a few last parts, will hopefully complete the restoration next week. Unless I go back to mono and start reading 45 page threads again...
Careful, rambling ahead!
I gave the amplifier more hours for potential burn in. Still, it sounded just wrong. I noticed that some songs sounded really okay. And some completely off. Something was wrong, mostly with the mids.Sometimes the guitar was only audible faint in the back, sometimes the singing voice very weak. Neil Young & Crazy Horse's majestic "Cortez the killer" sounded like coming out of a tin can. Hard and empty.
I started looking for the fault.
Reflowed solder joints, checked all parts I replaced for correctness.
Took out the tone board again, checked everything here.
Started checking the old wires in case one had cracked from moving around.
Still, the sound remained the same. I felt awful and was afraid that I had ruined a wonderful amplifier.
Read horror stories from people complaining about not liking their recapped amps.
Got more desperate, started reading patfronts much discussed 45 page capacitor shootout in the Marantz forum.
Started suspecting the Nichicon UPWs in the power supply for being accountable for the hard sound.
Started thinking about replacing them for Elna Silmics for that magic fix of my problem.
Started thinking about putting the old grey Elna coupling caps back in.
Started questioning my choice of Y-rated KSC1815 for the C-rated C458LGs throughout the unit.
Actually ordered really hard to find KSC945G from france, which haven't arrived yet.
...and finally had the sadly obvious idea to start checking the channels individually and remembered the stereo/mono selector on the 777A that offers several possiblites of channel composition (in my defense: front plate is still not mounted and just the bare switches are looking at me).
Well. Turns out I was listening to the right channel in MONO all the time.
The pot must have moved one position when pulling the knob.
Back to normal stereo and my 777A is back to his hearty self. The mids are there and they sing! Well, you can tell I am relieved!!
I really need to do some listening now, because now it sounds gorgeous and, as I am still waiting for a few last parts, will hopefully complete the restoration next week. Unless I go back to mono and start reading 45 page threads again...
toneriderMUC
Active Member
Amp clipping - question on centre voltage / clipping setting
I initially thought I'd dialed in the correct setting for the output adjustment using the crude method without scope. When listening yesterday I noticed occasional clipping when a big bass note is coming in after a quiet section. Listening levels were moderate. I went ahead to readjust clipping with VR801 and VR802 and reading the voltage on the coupling caps. Now here is the thing that puzzles me:
-> Voltages on the coupling caps seem to dial themselves in actually, no matter the position of VR801/802. When I turn the pots I can see the voltage rise/drop for a short moment, but then the readings very quickly move back to the same values. The values also arent the beforementioned ~40v, but less.
I am reading:
Main rail voltage: 70v
L/R: 36,5v / 37,5v
(As the values read 37/38 before I did any work on the amplifier, the values itself are probably ok)
The values for L/R, measured on the plus pole of the coupling caps, will always dial themselves in, no matter the postion of the trimpots.
Is that correct behaviour?
And if so, how can I know the correct position of the pots then, when only having the voltage to judge?
I initially thought I'd dialed in the correct setting for the output adjustment using the crude method without scope. When listening yesterday I noticed occasional clipping when a big bass note is coming in after a quiet section. Listening levels were moderate. I went ahead to readjust clipping with VR801 and VR802 and reading the voltage on the coupling caps. Now here is the thing that puzzles me:
-> Voltages on the coupling caps seem to dial themselves in actually, no matter the position of VR801/802. When I turn the pots I can see the voltage rise/drop for a short moment, but then the readings very quickly move back to the same values. The values also arent the beforementioned ~40v, but less.
I am reading:
Main rail voltage: 70v
L/R: 36,5v / 37,5v
(As the values read 37/38 before I did any work on the amplifier, the values itself are probably ok)
The values for L/R, measured on the plus pole of the coupling caps, will always dial themselves in, no matter the postion of the trimpots.
Is that correct behaviour?
And if so, how can I know the correct position of the pots then, when only having the voltage to judge?
Last edited:
The Fuxtor
Addicted Member
They tend to creep up I think. I would be shooting for 40v per side if possible. Getting the same voltages channel to channel is all you can do on this unit if you don't have a scope...
toneriderMUC
Active Member
Thats the thing, I dont have control over these voltages with the trimmers. When I open it up, I can see a short increase in voltage with a max of roughly 40v, but this settles itself back to around 37/38v within a few seconds. Same when I increase resistance; I can see a smaller voltage reading that will also settle itself to the higher value of 37/38v within seconds.
So, no matter what I set on the trimmers, the final reading will be those voltages. Dialing in 40v is impossible this way.
The Fuxtor
Addicted Member
Do the voltages settle in?
toneriderMUC
Active Member
They do, always to ~37v left and ~38v right channel. I had almost the exact same readings before doing any changes in the beginning. So I guess these values are alright for my unit; I just dont know how to set the trimmers, as the result will be always those values after settling.
I have now set the trimmers a bit lower, leading to an inital value of around 36V (before creeping to the final values of 37/38). Also dialed bias down from ~31ma to ~29ma for both channels. This seems to have solved the clipping issue for now; at least I haven't heard it yet.
It still would be nice to be able to set output voltage correctly. Maybe i need a scope here in the end.
The Fuxtor
Addicted Member
Maybe Smurfer77 can chime in....
AnalogueBen
Well-Known Member
This may be helpful for your current situation,
https://audiokarma.org/forums/index.php?threads/au-777-clipping.749280/
https://audiokarma.org/forums/index.php?threads/au-777-clipping.749280/
toneriderMUC
Active Member
Reverting some of the upgrades and finally arriving
In the end I was able to solve the clipping issue by actually just centering the trimpots. This showed the least amount of voltage creep to the relative target voltages described above.
Still, I was not happy with how it sounded. Basically it sounded overly big. Too much bass! Glassy hard highs. In songs with big individual bass notes the bass tended to overshadow the other frequencies. Overall the sonics sounded out of balance and sometimes the amplifier sounded as if it was pushed to the very edges of its design and out of its comfort zone.
I figured that the things that I put in got the amplifier out of tonal balance. I then startet re-evaluating the changes and step by step reverted some of the modifications I implemented following Leestereos restoration thread:
- increased capacitance of coupling and other caps along the signal path (such as increases from 10 to 22 uf and 47 to 100uf). I reverted these back to the original values using new caps of course.
- Changing the KSC1815Y with the low hfe rating to KSC945G (and finally following Leestereo here)
- Changing the huge 6800uf output couplers to 1500uf (still higher than the original 1000uf!)
Each of this step brought back a part of the original sonic character. The by far biggest difference could be observed by the change of the output couplers. This tamed the overblown bass. The 777a still has very decent bass with 1500uf couplers, but the sonic balance has been restored. Changing the 1815s to KSC945 removes a tad of cleaniness and brings back some of the original textured way mids are rendered.
It now sounds really beautiful.
This was a huge learning for me. This is very much a matter of taste, but I will stick to original values a lot more now in future projects.
In the end I was able to solve the clipping issue by actually just centering the trimpots. This showed the least amount of voltage creep to the relative target voltages described above.
Still, I was not happy with how it sounded. Basically it sounded overly big. Too much bass! Glassy hard highs. In songs with big individual bass notes the bass tended to overshadow the other frequencies. Overall the sonics sounded out of balance and sometimes the amplifier sounded as if it was pushed to the very edges of its design and out of its comfort zone.
I figured that the things that I put in got the amplifier out of tonal balance. I then startet re-evaluating the changes and step by step reverted some of the modifications I implemented following Leestereos restoration thread:
- increased capacitance of coupling and other caps along the signal path (such as increases from 10 to 22 uf and 47 to 100uf). I reverted these back to the original values using new caps of course.
- Changing the KSC1815Y with the low hfe rating to KSC945G (and finally following Leestereo here)
- Changing the huge 6800uf output couplers to 1500uf (still higher than the original 1000uf!)
Each of this step brought back a part of the original sonic character. The by far biggest difference could be observed by the change of the output couplers. This tamed the overblown bass. The 777a still has very decent bass with 1500uf couplers, but the sonic balance has been restored. Changing the 1815s to KSC945 removes a tad of cleaniness and brings back some of the original textured way mids are rendered.
It now sounds really beautiful.
This was a huge learning for me. This is very much a matter of taste, but I will stick to original values a lot more now in future projects.
Last edited:
stereofun
Super Member
Interesting observations. I too have learned to respect the coupling cap.
Coupling caps, be it in the form of input or output coupling are quite important. The best ofcourse is to have no cap, but when the design calls for one, the stakes are high since the signal is mirrored across that cap - It never passes through electron by electron, since there is no electrical conductance across the capacitors plates. That is why the inherent design of a capacitor blocks DC because DC needs a physical path - the "circuit" is either conducting or it's not. AC on the other hand can reverb/reflect/echo/mirror across non-conducting plates. This is also referred to as displaced current or a 'time varying electric field".
With that in mind it makes sense how the construction, dielectric, size and other properties of a coupling cap suddenly becomes important for it's ability to "cast" the signal to the other side.
From transistors we know that the more capacitance "COB" the slower they typically conduct. I wonder if that holds true for coupling caps too. At any rate, over the years I've become quite conservative we with no more than doubling output caps. (Supply is an entirely different matter).
Good on you toneriderMUC to "follow" the sound and reverse and tweak until you get the sound that is right.
Coupling caps, be it in the form of input or output coupling are quite important. The best ofcourse is to have no cap, but when the design calls for one, the stakes are high since the signal is mirrored across that cap - It never passes through electron by electron, since there is no electrical conductance across the capacitors plates. That is why the inherent design of a capacitor blocks DC because DC needs a physical path - the "circuit" is either conducting or it's not. AC on the other hand can reverb/reflect/echo/mirror across non-conducting plates. This is also referred to as displaced current or a 'time varying electric field".
With that in mind it makes sense how the construction, dielectric, size and other properties of a coupling cap suddenly becomes important for it's ability to "cast" the signal to the other side.
From transistors we know that the more capacitance "COB" the slower they typically conduct. I wonder if that holds true for coupling caps too. At any rate, over the years I've become quite conservative we with no more than doubling output caps. (Supply is an entirely different matter).
Good on you toneriderMUC to "follow" the sound and reverse and tweak until you get the sound that is right.
toneriderMUC
Active Member
Just a quick update after two months: I had also replaced the Power Supply board caps with Elna Silmics in december, this did bring another increase in sound quality, it removed some harshness/hardness from the sound that still bothered me. Since then within the first 10 hours or so of listening I could further observe some break in, resulting in a more smooth and less edgy and dry sound. My 777A now has gained the highly desireable sound qualities back that I found in him before the restoration and even builds upon that with even more authority and clarity. Its the best of both words of meaty, gorgeous mids, thumping bass and detailed, almost modern sound presentation. Bias is spot on and it seems to punching way above its wattage rating.
I can say the restoration wasnt easy, but in the end a full success. I am amazed everytime how GOOD this old amp sounds.
I can say the restoration wasnt easy, but in the end a full success. I am amazed everytime how GOOD this old amp sounds.
EMenendez
Active Member
To not see the forest for the trees. Or: sometimes the solution is just too simple.
Back to normal stereo and my 777A is back to his hearty self. The mids are there and they sing! Well, you can tell I am relieved!!
I really need to do some listening now, because now it sounds gorgeous and, as I am still waiting for a few last parts, will hopefully complete the restoration next week. Unless I go back to mono and start reading 45 page threads again...
I only a have a question.
Honestly after recap your amp that look fantastic before and after the full recapped like the smurfer's or Leestereo's amps and with some hours of listening and with your experience which is very recent.
Do you think it is worth doing a complete restoration when an amp may still be within its factory parameters even though its caps are 40 years or older, I ask this because several people write that maybe some amplifiers sound sweet because their caps are old and with new caps lose their magic and become more analytical and even sterile amplifiers.
Regards