Trnsfmr
Well-Known Member
Hi all,
I been following the developments with the AU999 here. I think that many are beginning to realize this is Sansui's best sounding amp.... ever. I recently modded two of them with Kev's bias mod, or anti clipping mod, or... preamp mod, I think he called it. Wow, very impressive. It raised the clean output level of this amp by 3-6 db. They just sound so clean and detailed at any volume level, now. Fantastic.
There's a mod I have been doing to this amp for a while, that is really simple and results in a better low end. Customers who bring me their AU999's, I tell them that they will be able to turn off the low EQ switch, and still have plenty of low end, and not need the 2-4db boost everybody runs most of the time. They don't believe me until they get the amp back and try it.
First, a little background. I don't know if you've ever heard a DC amp. A DC amplifier supposedly has response all the way down to DC, with no blocking capacitors anywhere in the chain. If you've had the chance to listen to one, on a speaker system with good low end response, you know how good they can sound on low end.
My first experience with a DC amp, was with a Crown DC300A in 1977. I had a pair of corner Klipschorns at the time and the low end just amazed me. I just had to figure out why it was so good. It certainly wasn't the added power, with the super efficient Klipsch's.
It turns out that when low end rolls off, from a tone control, amplifier passband, or whatever, there are phase effects that occur high above the -3db rolloff point. These phase effects happen all the way up to 2-4 octaves above the cutoff freq, depending on the nature of the rolloff.
These phase effects actually change the frequency and amplitude of the first few cycles of any bass note. A DC amp passes them without change. There was an AES paper about this years ago, and I still have a copy of it somewhere.
Turns out that you can show in SPICE, an electronics design simulator program, that adding a 2-4db boost to the low end will counteract these phase effects to a large degree. So maybe what we're doing when we boost the low end a little is to fix the phase response.
Even better, of course, would be to not have the problem to begin with.
In the 70's there was hardly any music with content below 40hz. Records just didn't work well below 40hz. Most of the HiFi gear started their rolloff somewhere between 20 and 40hz.
The thing is that every coupling cap between every stage of an amp or preamp has a rolloff point. They all cascade together to make an often very nasty 4-8 pole filter, with no engineered-in desirable characteristics. In fact the low end phase response is often awful. The earlier amps are usually worse, as the low end got extended as time went on and engineers became more aware of this outcome. The AU999 is from 1972 or earlier, and was not immune to the problem.
So how about this. Let's increase the size of the coupling caps between stages, so as to get these effects down as low as possible. The math would predict that if you got each insterstage rolloff point down really low, down to 2-4hz, then the phase effects in the passband would be minimized.
The coupling caps on the AU999 are these really cool Metalized Polyester (I think?) caps, the big greenies, .47-1uf. By adding larger electrolytics in parallel with these caps, this goal can be realized. I am currently adding 4.7uf caps across the big greenies, and it seems to work.
Sounds like a DC amp to me.
It's really easy, just add the little electrolytics on the back side of the boards, like you see in the picture. Don't add them on the phono board. Or, do it if you want, if you have a really quiet turntable.
First, with the amp on, use your meter to measure across each of the coupling caps to see how they are biased. Solder in the caps in with appropriate polarity.
I'd really like to hear what others think about this simple mod. Of course, you can do it to any amplifier. The possible problem is one of low frequency instability, and most SS amps do not have a problem in this regard. I've done it to quite a few amps now, with great results. I do this to the AU505/555 and it is amazing how that little amp can sound on low end.
Pretty soon, there will be two kinds of AU999. The Sansui factory model, and the AK model.
On the f1208
C701-02
C711-12
C715-16
On the F1209
C721
C725
C727
C737
On the F1159
C801
You can increase the size of the added caps across C715-16 and C737 to 10uf@50v. instead of the 4.7uf@50v for the others. These caps already have an electrolytic in parallel, and bigger will be better here as the object is to increase the coupliing cap value by several times. I know someone will ask, so, the caps I used here are Nichicon UKW.
Now, don't go crazy and go to super high values on these caps. It's not as if a little is good and a lot is better, for this. Huge caps will cause all sorts of slow turn on transients, that could cause problems.
Hope this works for you.
I been following the developments with the AU999 here. I think that many are beginning to realize this is Sansui's best sounding amp.... ever. I recently modded two of them with Kev's bias mod, or anti clipping mod, or... preamp mod, I think he called it. Wow, very impressive. It raised the clean output level of this amp by 3-6 db. They just sound so clean and detailed at any volume level, now. Fantastic.
There's a mod I have been doing to this amp for a while, that is really simple and results in a better low end. Customers who bring me their AU999's, I tell them that they will be able to turn off the low EQ switch, and still have plenty of low end, and not need the 2-4db boost everybody runs most of the time. They don't believe me until they get the amp back and try it.
First, a little background. I don't know if you've ever heard a DC amp. A DC amplifier supposedly has response all the way down to DC, with no blocking capacitors anywhere in the chain. If you've had the chance to listen to one, on a speaker system with good low end response, you know how good they can sound on low end.
My first experience with a DC amp, was with a Crown DC300A in 1977. I had a pair of corner Klipschorns at the time and the low end just amazed me. I just had to figure out why it was so good. It certainly wasn't the added power, with the super efficient Klipsch's.
It turns out that when low end rolls off, from a tone control, amplifier passband, or whatever, there are phase effects that occur high above the -3db rolloff point. These phase effects happen all the way up to 2-4 octaves above the cutoff freq, depending on the nature of the rolloff.
These phase effects actually change the frequency and amplitude of the first few cycles of any bass note. A DC amp passes them without change. There was an AES paper about this years ago, and I still have a copy of it somewhere.
Turns out that you can show in SPICE, an electronics design simulator program, that adding a 2-4db boost to the low end will counteract these phase effects to a large degree. So maybe what we're doing when we boost the low end a little is to fix the phase response.
Even better, of course, would be to not have the problem to begin with.
In the 70's there was hardly any music with content below 40hz. Records just didn't work well below 40hz. Most of the HiFi gear started their rolloff somewhere between 20 and 40hz.
The thing is that every coupling cap between every stage of an amp or preamp has a rolloff point. They all cascade together to make an often very nasty 4-8 pole filter, with no engineered-in desirable characteristics. In fact the low end phase response is often awful. The earlier amps are usually worse, as the low end got extended as time went on and engineers became more aware of this outcome. The AU999 is from 1972 or earlier, and was not immune to the problem.
So how about this. Let's increase the size of the coupling caps between stages, so as to get these effects down as low as possible. The math would predict that if you got each insterstage rolloff point down really low, down to 2-4hz, then the phase effects in the passband would be minimized.
The coupling caps on the AU999 are these really cool Metalized Polyester (I think?) caps, the big greenies, .47-1uf. By adding larger electrolytics in parallel with these caps, this goal can be realized. I am currently adding 4.7uf caps across the big greenies, and it seems to work.
Sounds like a DC amp to me.
It's really easy, just add the little electrolytics on the back side of the boards, like you see in the picture. Don't add them on the phono board. Or, do it if you want, if you have a really quiet turntable.
First, with the amp on, use your meter to measure across each of the coupling caps to see how they are biased. Solder in the caps in with appropriate polarity.
I'd really like to hear what others think about this simple mod. Of course, you can do it to any amplifier. The possible problem is one of low frequency instability, and most SS amps do not have a problem in this regard. I've done it to quite a few amps now, with great results. I do this to the AU505/555 and it is amazing how that little amp can sound on low end.
Pretty soon, there will be two kinds of AU999. The Sansui factory model, and the AK model.
On the f1208
C701-02
C711-12
C715-16
On the F1209
C721
C725
C727
C737
On the F1159
C801
You can increase the size of the added caps across C715-16 and C737 to 10uf@50v. instead of the 4.7uf@50v for the others. These caps already have an electrolytic in parallel, and bigger will be better here as the object is to increase the coupliing cap value by several times. I know someone will ask, so, the caps I used here are Nichicon UKW.
Now, don't go crazy and go to super high values on these caps. It's not as if a little is good and a lot is better, for this. Huge caps will cause all sorts of slow turn on transients, that could cause problems.
Hope this works for you.
Attachments
Last edited: