just like I Said..
two stage...high cut between the legs and active feedback in the second op-amp stage... let the first stage run without the feedback, as the signal levels are so low the distortion would be fairly low, noise sort of highish. (if you use the LSK or SK fets. (interfet would be interesting and lower in noise)
second stage would then have the base shelf as a part of the feedback network..
what you would then have is a floating input. (not balanced) and a single-ended output. Performance would in my objective be good. (One shot is still a more novel circuit.)
two stage...high cut between the legs and active feedback in the second op-amp stage... let the first stage run without the feedback, as the signal levels are so low the distortion would be fairly low, noise sort of highish. (if you use the LSK or SK fets. (interfet would be interesting and lower in noise)second stage would then have the base shelf as a part of the feedback network..
what you would then have is a floating input. (not balanced) and a single-ended output. Performance would in my objective be good. (One shot is still a more novel circuit.)
Thank you guys for your inputs.
I am not a big, big fan of active RIAA, I would like to keep it passive and/or transimpedance if possible.
I know there are benefits by running it active, but for the fact that feedback level changes significantly with the frequency makes me slighltly prefer the passive solution over it.
Joachim if you don't mind, why would you pt transimpedance first?
I will elaborate a desing concept and then post it here soon and maybe we can take it from here.
I am not a big, big fan of active RIAA, I would like to keep it passive and/or transimpedance if possible.
I know there are benefits by running it active, but for the fact that feedback level changes significantly with the frequency makes me slighltly prefer the passive solution over it.
Joachim if you don't mind, why would you pt transimpedance first?
I will elaborate a desing concept and then post it here soon and maybe we can take it from here.
I think the 75usec first, transimpedance and the active bass-mid shelf gives you the best of both worlds. The 75usec after the first stage has the advantage that the second stage sees a much reduced slew rate. Records have scratches and dirt that can bring a spectrum up to 500kHz and that can slew the RIAA. Also the record has a 20dB boost at 20kHz.
That puts a lot of burden on a conventional RIAA. The first stage 75usec transimpedance solves that problem and gives tremendous overload margin.
Also the transimpedance nature lowers distortion in the treble, the higher the less ( because of rising current feedback ) and when you use an open loop class a circuit for the first stage distortion also drops with level. The ear works the same way.
The second stage then has an easy job to make the bass-mid shelf active.
The 6dB dropping nature can be arranged in such a way that the amount of feedback is rather the same over the frequency range because the open loop gain of a VFC is also falling with 6dB/oct. Look at my other phono thread "Nobrainer" for circuits like that.
They sound excellent.
That puts a lot of burden on a conventional RIAA. The first stage 75usec transimpedance solves that problem and gives tremendous overload margin.
Also the transimpedance nature lowers distortion in the treble, the higher the less ( because of rising current feedback ) and when you use an open loop class a circuit for the first stage distortion also drops with level. The ear works the same way.
The second stage then has an easy job to make the bass-mid shelf active.
The 6dB dropping nature can be arranged in such a way that the amount of feedback is rather the same over the frequency range because the open loop gain of a VFC is also falling with 6dB/oct. Look at my other phono thread "Nobrainer" for circuits like that.
They sound excellent.
I like your concept very much.
For the first stage transimpidence, would you use the only N-CH JFET version or the previous versions?
Gotcha MiiB.
We can include the one-shoot on the second stage if you think this has potential.
so basically the transimpedance first stage makes the most sense and I think we all agree with that.
Second stage We were thinking whether to use MAT04 circuit application or one shot or I have another one that I have been using with great success (I will post that next).
The only thing is that the arrangement with the falling -6dB angular coefficient uses OP-AMPS (I will read the post nonbrainer by Joachim, I was in meetings till now so had no time to do that).
One thing I would like to avoid...and I don't want to be a pain on the butt and you guys think I am too picky...but I want a full discrete design.
Said that...how can the calculated 6db fall get done with discrete?
here is my proposed second stage.
That would imply passive RIAA.
any thoughts?
Would this have the mean for a low noise phono design?
I still haven't gotten an answer from you guys if we would like to see transimpidance only N-Channel on or the regular complementary's scheme?
That would imply passive RIAA.
any thoughts?
Would this have the mean for a low noise phono design?
I still haven't gotten an answer from you guys if we would like to see transimpidance only N-Channel on or the regular complementary's scheme?
Attachments
Last edited:
P.S. I am reading the nobrainer's thread now and I see the discrete idea so I am reading that very carefully (I don't know how long the thread is yet )
Anyway, just a small disclaimer, I think that as the statement of work at the beginning we should try to make this phono balance throughout without converting to SE and then re-converting to balance (please feel free to post your thoughts)
)Anyway, just a small disclaimer, I think that as the statement of work at the beginning we should try to make this phono balance throughout without converting to SE and then re-converting to balance (please feel free to post your thoughts)
The only problem i see with discrete is the longer time it needs to design and optimize it.
You have to imagine that we needed 4 !!! years to get to something like the Paradise.
I would definitely like Fets at the input, i think with BJT our team has mostly done everything that can be done. The advantage is DC coupling and immunity to RF but you have to solve the problem of lower conductance ( Gm ) so not to run into noise, gain or linearity problems. If you want to go the N-channel root and balanced you have a 6dB noise penalty any way and then you refuse to parallel many Fets. The only sensible option is then the Interfet part. Both problems can ameliorated with circuit 288, parallel symmetric with P-Channels. The only choice then is the 2SJ74. I have enough, you may have enough plus you do not want to go into group buy plus you want to max it out with the best available parts. That limits the amount of stages that will be build by others anyway. So forget the P-channel problem.
You have to imagine that we needed 4 !!! years to get to something like the Paradise.
I would definitely like Fets at the input, i think with BJT our team has mostly done everything that can be done. The advantage is DC coupling and immunity to RF but you have to solve the problem of lower conductance ( Gm ) so not to run into noise, gain or linearity problems. If you want to go the N-channel root and balanced you have a 6dB noise penalty any way and then you refuse to parallel many Fets. The only sensible option is then the Interfet part. Both problems can ameliorated with circuit 288, parallel symmetric with P-Channels. The only choice then is the 2SJ74. I have enough, you may have enough plus you do not want to go into group buy plus you want to max it out with the best available parts. That limits the amount of stages that will be build by others anyway. So forget the P-channel problem.
I don't exactly understand what you mean. Are you referring to the Mat04 app circuit?
Also when you talk about fet op amp and coupling cap, how would that be better than a discrete stage with little intrusion of a Servo?
I mean I get that a lot of successful design have been done both ways, so I am sure that the solution you are suggesting is pleanty viable, it is just that I would really love to do everything discrete and no coupling cap.
Also, please feel free to post schematics with your ideas, sometime it explains better than 1000words
so if I don't get your post wrong, you are saying that you are liking the circuit I posted on 288?
I mean I have been succesfully using this circuit for years and my first phono had that kind of topology in it sounded nice, just had little noise with a 2 stage design.
would you think then that transimpedance balance with complementary and 75uS time constant, then buffer, then PASSIVE RIAAfor the other 2 time constants then stage with 288 and then buffer could be a very good solution?
What do you guys think? MiiB's suggestion one shot is interesting, but it's hard to convert it to balance.
I leave open discussion hoping that we can seattle for second stage.
P.S. the J74/K170 can still be had at a pretty decent price, so I don't think that it could be impossible for others to build this if they wanted to. Also time is not a problem, We don't have to rush the design, whatever it takes it takes.
LTP= Long Tail Pair, stands for differential input with current source and degeneration if needed.
Have I passed the exame professor?
What do you mean by don't you have a discrete opamp in you gems?
I am sure that once you get back from your trip this post will get much, much more interesting when you can finally get back to your simulation and cranking ideas along with me and Joachim (and hopefully someone else too : )
In Linear Audio 0 i have shown a circuit called FPS ( Flexible Phono Stage ) that is the balanced input version of your post 288. It is parallel symmetric AND balanced.
That input stage i would use ( maybe without feedback ), even today.
It is too complex to describe in words and Jan Didden owns the article so i can not post it here.
That input stage i would use ( maybe without feedback ), even today.
It is too complex to describe in words and Jan Didden owns the article so i can not post it here.
Azre you referring to this?
This is a variant of what I was planning on using for my first version a couple of years ago.
what do you think?
Attachments
Last edited:
Interesting.
Let me shot an email to Jan NOW !!! Before I push it back and back....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.