It sounds clear and lush. It drives both low impedance and high impedance headphones.
The DT-880 600 ohms, which is a bit of monster all its own sounds fantastic with the buffer.
In the 20 dB of gain configuration I have in place, it can drive most headphones with 1V out of a DAC so far.
I will post measurements later.
With a 330 ohm load, I already say the FR is very good. -2 dB at 20 Hz and flat after 30 Hz or so. THD+Noise sits at 0.1% with 1Vrms of input. H2 being the biggest contributor.
The DT-880 600 ohms, which is a bit of monster all its own sounds fantastic with the buffer.
In the 20 dB of gain configuration I have in place, it can drive most headphones with 1V out of a DAC so far.
I will post measurements later.
With a 330 ohm load, I already say the FR is very good. -2 dB at 20 Hz and flat after 30 Hz or so. THD+Noise sits at 0.1% with 1Vrms of input. H2 being the biggest contributor.
@sov, don’t do part 2 yet. And I’ll post the cap change, which I found to be very useful.
I’ve been enjoying this beauty recently. That’s a Type 85 tube, which has an absolutely glorious sound. The load is a CCS, similar to the SCG. Power supply has a Maida reg. I built this just before starting down the SCG journey, so it is maybe three years old now. It has a truly lush sound where you can forget about everything else and just enjoy the music. If I wasn’t a slave to building new stuff and experimenting, I could just stop here and be happy forever.
When I hooked it up recently, I remembered why I fell in love with this tube. The SCG is cleaner than this, it doesn’t editorialize as much, and it shows you more without ever being close to harsh. But sometimes, you just want to relax and not really want to notice everything. That’s why we have more toys than we need.
I’ve been enjoying this beauty recently. That’s a Type 85 tube, which has an absolutely glorious sound. The load is a CCS, similar to the SCG. Power supply has a Maida reg. I built this just before starting down the SCG journey, so it is maybe three years old now. It has a truly lush sound where you can forget about everything else and just enjoy the music. If I wasn’t a slave to building new stuff and experimenting, I could just stop here and be happy forever.
When I hooked it up recently, I remembered why I fell in love with this tube. The SCG is cleaner than this, it doesn’t editorialize as much, and it shows you more without ever being close to harsh. But sometimes, you just want to relax and not really want to notice everything. That’s why we have more toys than we need.
I just listened to Flamenco Sketches on my build and it was truly excellent!
I do love tubes, though, particularly the 300B, so now I have to wonder about the Type 85.
Tubes are not superior, but the way they work is so close to fundamental research in physics, it’s as magical as science can get.
I do love tubes, though, particularly the 300B, so now I have to wonder about the Type 85.
Tubes are not superior, but the way they work is so close to fundamental research in physics, it’s as magical as science can get.
Ra7, still using the SCG?
I read the whole thread after finding a post where you compared SCG with tube preamps I know (01A, 10Y) --this got me interested and I might want to try it myself .
Your comment about the type of amplifier that should follow the SCG is the problem. What would you suggest that would work well and doesn't have a high dissipation? I bi-amplify, so low dissipation is (even more) important.
I read the whole thread after finding a post where you compared SCG with tube preamps I know (01A, 10Y) --this got me interested and I might want to try it myself .
Your comment about the type of amplifier that should follow the SCG is the problem. What would you suggest that would work well and doesn't have a high dissipation? I bi-amplify, so low dissipation is (even more) important.
Not to speak for ra7 but I can tell you he has continued with the project. More important things, like his young family, have taken precedence.
I made mine using perf board and have been enjoying it immensely for the last six months, if not more.
I am using mine to drive Rahul's TDV amplifier idea using the THF51. I made four of these since I, too, biamplify.
I think the only precautions are that you need an amplifier that needs 18-20 dB of gain. If your amplifier already has 20 to 30 dB of gain this will not work well since you will have to attenuate much of that gain.
I am a fan of Rahul's designs so I am very biased.
I look forward to Rahul getting time to return to this thread.
I made mine using perf board and have been enjoying it immensely for the last six months, if not more.
I am using mine to drive Rahul's TDV amplifier idea using the THF51. I made four of these since I, too, biamplify.
I think the only precautions are that you need an amplifier that needs 18-20 dB of gain. If your amplifier already has 20 to 30 dB of gain this will not work well since you will have to attenuate much of that gain.
I am a fan of Rahul's designs so I am very biased.
I look forward to Rahul getting time to return to this thread.
Thank you for the support, Rick! You have helped me keep going. Yes, life has been busy but I have been working on this design. Rick has been on the experimental journey and we have made quite a bit of progress. I also learned a number of things along the way. The next update will introduce JFETs for the input and gain transistor. In the past, I tried the good old 2sk170, but it didn't sound that great. Now I know the reasons why.
I'm waiting to finish developing a JFET headphone buffer amp before posting an update to this thread. There is already a MOSFET buffer but I want to hear what JFETs can do, especially after hearing their potential in the SCG. Testing on that is not done yet. I just ordered 200 JFETs from Digikey, the dirt cheap kinds, of course. Hope to have that wrapped up soon.
If interested, I can post a couple of schematics.
I'm waiting to finish developing a JFET headphone buffer amp before posting an update to this thread. There is already a MOSFET buffer but I want to hear what JFETs can do, especially after hearing their potential in the SCG. Testing on that is not done yet. I just ordered 200 JFETs from Digikey, the dirt cheap kinds, of course. Hope to have that wrapped up soon.
If interested, I can post a couple of schematics.
I have been using the F4 to do all the testing. VFET followers are great also. Basically, if you have 2Vrms from your DAC or other sources, then you can clip the F4 using the SCG preamp. I would lean towards low distortion, follower type, output stages. F4 or the BA output stages are good a good fit. Then you can play with the sound using the preamp.
For low dissipation, I would look at the MOFO-style amps, which use a choke to increase efficiency. Still Class-A of course. But more like 20V x 2A = about 40W per channel. If you have four channels, it comes to 160W, not too bad. I have a Parasound, and the Hypex amps, and they come out from time to time, to see if something changed, but they have always gone back into hiding when the big boys come out to play
How efficient are your speakers? If you need lower power, the Zenductor could be a good fit as well.
Hope this inspires you to post some new ideas, Rahul.
I have to say again - Rahul's amplification ideas are simple as all good sounding circuits are but are the result of careful listening and refinements to his premise. His ideas evolve - he is a build/listen guy not a simulations expert - not to say he does not know his way with simulations but he knows better than to trust them.
I have to say again - Rahul's amplification ideas are simple as all good sounding circuits are but are the result of careful listening and refinements to his premise. His ideas evolve - he is a build/listen guy not a simulations expert - not to say he does not know his way with simulations but he knows better than to trust them.
Thanks for your reply!
I don't need much power (horn loaded highs (>500Hz) and a 15 inch below (96 db/1m/1w)). I also have (but not always use) three small sealed subs (distributed in the room), with SS amplification. So, sometimes I am actually tri-amping.
SIT/VFET amps are new to me, I have been reading a lot here lately and Iean towards a choke loaded follower if going this route. I even have the Hammond chokes....
Two things have made me consider exploring SS. One is to have a (secondary?) amp system which I can leave powered most of the day without worrying about the tube$. Another, to see how a digital crossover would work, which would mean having a few preamps (unpractical with tubes). In this context your SCG preamp caught my attention, but regarding amplifiers I have, for long, been interested to try one of the Kaneda designs (finding parts is headache).
Thanks for your comment too, I will keep an eye on the next refinement of the SCG, and perhaps build one just for fun. You mention you are using the THF51. How would you compare them to 2SJ28 or even simulated SITs?
@ra7 Have you tried to connect Rsch1 to the source of Q103 instead of the drain of Q101?
I played with the same kind of circuit ( https://www.diyaudio.com/community/...es-from-a-n-channel-dmos-and-not-only.376683/ ) inspired by the UNSET and that was a solution to have real CCS load on the "solid-state-triode", otherwise it will see the CCS in parallel with the 11k load.
Then a dedicated load can be set to fix the loadline of the "solid-state-triode", so its harmonic content.
I played with the same kind of circuit ( https://www.diyaudio.com/community/...es-from-a-n-channel-dmos-and-not-only.376683/ ) inspired by the UNSET and that was a solution to have real CCS load on the "solid-state-triode", otherwise it will see the CCS in parallel with the 11k load.
Then a dedicated load can be set to fix the loadline of the "solid-state-triode", so its harmonic content.
EDIT from previous post: I tried an output buffered version where I took the voltage divider, and in that case the 10k+1k load was not applied to the solid-state-triode, but it was loaded with pure CCS. Connecting as I described in the previous post, is always in parallel with the CCS. But IIRC there was an improvement.