since the scarcity of p channel how about using N channel Jfets in complimentary input stage?
well its stated in cordell book Im sure we will not have serious matching issues but atleast we can get jfet input amp and not in uni polar style. How will be the THD at 10k and 20k?
well its stated in cordell book Im sure we will not have serious matching issues but atleast we can get jfet input amp and not in uni polar style. How will be the THD at 10k and 20k?
No way to do a fully symmetric input stage. The best compromise is
the sansui way , just N-channel Jfets. (below). Single ended driving
a complimentary second stage/VAS.
I'm not going to develop that posted "kypton V" further , use it as you wish.
Member thimios was quite impressed by it's SQ , BTW .
OS
the sansui way , just N-channel Jfets. (below). Single ended driving
a complimentary second stage/VAS.
I'm not going to develop that posted "kypton V" further , use it as you wish.
Member thimios was quite impressed by it's SQ , BTW .
OS
Attachments
OStripper.
LTspice simulation of your Krypton schematic generated an output FFT that has higher THD on Odd_Harmonics than the THD level of the preceeding even order harmonic. I had expected the single JFET Diff stage to lower the progressive odd harmonics.
The Symetri topology with both complementary diff inputs and complementary diff VAS circuits has a similar FFT harmonic profile and is more robust to clipping. Naturally, you will need to send a "Dear Santa" letter to Nelson Pass to get PJFETs before Christmas.
LTspice simulation of your Krypton schematic generated an output FFT that has higher THD on Odd_Harmonics than the THD level of the preceeding even order harmonic. I had expected the single JFET Diff stage to lower the progressive odd harmonics.
The Symetri topology with both complementary diff inputs and complementary diff VAS circuits has a similar FFT harmonic profile and is more robust to clipping. Naturally, you will need to send a "Dear Santa" letter to Nelson Pass to get PJFETs before Christmas.
Attachments
OStripper.
LTspice simulation of your Krypton schematic generated an output FFT that has higher THD on Odd_Harmonics than the THD level of the preceeding even order harmonic. I had expected the single JFET Diff stage to lower the progressive odd harmonics.
The Symetri topology with both complementary diff inputs and complementary diff VAS circuits has a similar FFT harmonic profile and is more robust to clipping. Naturally, you will need to send a "Dear Santa" letter to Nelson Pass to get PJFETs before Christmas.
The FFT not looking good. I think its better to go complimentary jfet input than using one.
Even with that horrible looking fft , you are still 95+db down.
That is 30ppm , BTW.
My 1k sound card FFT shows a more realistic harmonic plot.
edit line ... that "spray" of odd harmonics does not show in real life - model generated.
get rid of the input cap , input filter cap ... disconnect the servo ... it goes away.
OS
That is 30ppm , BTW.
My 1k sound card FFT shows a more realistic harmonic plot.
edit line ... that "spray" of odd harmonics does not show in real life - model generated.
get rid of the input cap , input filter cap ... disconnect the servo ... it goes away.
OS
Last edited:
The FFT not looking good. I think its better to go complimentary jfet input than using one.
Full complimentary cancels even order .
A single ended cancels all.
OS
For comparison. No input cap, no DC cap, no servo.
Kypton FFT of JFET input and BJT VAS
Symetri FFT of JFET input and BJT VAS
Yes the symetri is better since they both even and odd are kind of closer.
But my question is about compared to a circuit like this does the Kypton THD better?
http://m.eet.com/media/1196048/F11large.jpg
The only reason why i see jfet for input for being less noisier and i believe jfet amp would give better transparency. I have listened to Amp camp from pass diy despite of having single jfet input the natural tones were not good and all together and no hint of transparent sound. I read somewhere that single differential input is always better than just using a single jfet.
The main problem comes with the cost of the jfets which is whooping expensive and the worst part is to match them hence to buy alot of them.
so single differential input pair would be a better option available at this point only worried about the distortion.
The main problem comes with the cost of the jfets which is whooping expensive and the worst part is to match them hence to buy alot of them.
so single differential input pair would be a better option available at this point only worried about the distortion.
Im trying to simulate the circuit but something is wrong on the negative side of the input
https://dl.dropboxusercontent.com/u/7421969/0013_lsk_489%20JFET_KSA_30V.asc
https://dl.dropboxusercontent.com/u/7421969/0013_lsk_489%20JFET_KSA_30V.asc
Last edited:
Yes the symetri is better since they both even and odd are kind of closer.
But my question is about compared to a circuit like this does the Kypton THD better?
http://m.eet.com/media/1196048/F11large.jpg
Linesources plots seemed a bit off.
(below ) are my chosen 3 designs - the better ones ...
-60ma AB bias
-20K 50W
-Keentoken-Cordell BJT models
- same jumpered servo/caps
PS - 3rd one is the spooky leach with a symetri VAS , BTW.
Edit - ND beats em' all at 5ppm (first plot).
OS
Attachments
Im trying to simulate the circuit but something is wrong on the negative side of the input
https://dl.dropboxusercontent.com/u/7421969/0013_lsk_489%20JFET_KSA_30V.asc
can you tell me what needs to be fixed in the above circuit?
Just reworked with the current mirrors but the output current is in the order of nano and pico amps.
https://dl.dropboxusercontent.com/u/7421969/0015_lsk_489 JFET_KSA_50V.asc
https://dl.dropboxusercontent.com/u/7421969/0015_lsk_489 JFET_KSA_50V.asc
Just reworked with the current mirrors but the output current is in the order of nano and pico amps.
https://dl.dropboxusercontent.com/u/7421969/0015_lsk_489 JFET_KSA_50V.asc
This is the second circuit which I tried to simulate from bob cordell book and didnt work. I dont know why its not working.
Bob's circuits were written for the book - just a general guideline.
Here , some .asc's - a VFA and a CFA + cordell/keantoken models.
PS - I listening to both currently. Just click and run.
OS
Here , some .asc's - a VFA and a CFA + cordell/keantoken models.
PS - I listening to both currently. Just click and run.
OS
Attachments
Bob's circuits were written for the book - just a general guideline.
Here , some .asc's - a VFA and a CFA + cordell/keantoken models.
PS - I listening to both currently. Just click and run.
OS
Im looking for something which can work with N channel dual jfet input with pushpull vas. so I just need to know how to implement the input section with those so that the vas can be in pushpull. Can you suggest one?
which one do you like the VFA or CFA?
Im looking for something which can work with N channel dual jfet input with pushpull vas. so I just need to know how to implement the input section with those so that the vas can be in pushpull. Can you suggest one?
which one do you like the VFA or CFA?
Both are good , I've only run BJT's on either.
Here is another that I actually have used e-waste 70's Jfets with.
A very common 70-80's Japanese Jfet input stage.
The 1'st Jfet stage run +/-12V , any FET will work.
OS
Attachments
Im thinking of this input stage..
http://www.diyaudio.com/forums/atta...osfet-vas-lateral-output-perfect-perfect1.png
Reg kypton v2 im sure it would sound great as its harmonic structure is excellent for good sounding amp. Im not talking about THD here but the ratios of the even and odd harmonics as gain increases..
http://www.diyaudio.com/forums/atta...osfet-vas-lateral-output-perfect-perfect1.png
Reg kypton v2 im sure it would sound great as its harmonic structure is excellent for good sounding amp. Im not talking about THD here but the ratios of the even and odd harmonics as gain increases..
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Solid State
- Bob Cordell complementary input stage with N channel Jfets