An externally hosted image should be here but it was not working when we last tested it.
In the above simplified circuit, what impact on the distortion, gain and soundquality would different choices of the top transistor have? I can imagine using a TIP141 darlington (a really well known audiophile part! 😀 ) , an IRFP054, an IRFP450 or a Semisouth SJEP120R063. Would it have any inpact on the amps performance at all? If not, i could use whatever part that i have or a really cheap part.
I will not build the circuit above, it just serves as a reference for the discussion.
From the discussion I have read, it seems that as long as the cascode device is used within its power limitations, then it does not contribute much in the way of sound quality.
I have been playing around with a few cascoded designs, but I am yet to hear them, so I can't make a comment based on personal experience.
I remember Nelson saying the cascode device has little affect on the sound in his Zen 9 (F3) amp.
I have been playing around with a few cascoded designs, but I am yet to hear them, so I can't make a comment based on personal experience.
I remember Nelson saying the cascode device has little affect on the sound in his Zen 9 (F3) amp.
Last edited:
I have seen that to, but i started thinking about the issue yesterday. The high resistant and low transconductance of a high voltage part like a IRFP450 should modulate the voltages a lot more than a high current low voltage part like an IRFP054. The IRFP054 should give more gain. But i dont know how much difference it will make.
What I am intending to do is compare a Lateral Mosfet vs Vertical Mosfet, vs Jfet, vs Bipolar. Hopefully I will get time to do this over the Christmas Holidays.
why cascode should modulate voltage at all ?
real plain cascode isn't anything else than umbrella for active part .
but - in case where you wish to implement any sort of "modulated" cascode , where cascode element became active part itself , I agree that choice of part is certainly important issue .
The cascode-transistor should keep the voltage fixed across the active transistor, but since a real transistor will have an finite amount of transconductance and some resistance it will not keep that voltage perfect. The upper transistor will add some gain to the circuit and that amount should vary too. Maybe this doesnt matter in the real world, but it seems like there is some potential for improvment?!
Is this soo or am i wrong.
well - put enough current through bias divider , use part which have low enough Rdson (meaning low enough comparing to load impedance ) , and I think everything is OK - in real world ....... usual problem with cascode is wimpy current through bias string .
speaking of that - seems that wimpy current is pretty offten problem in audio gadgets ........ but that's just my opinion .......
same as all opinions of mine ....... are my opinions ......
speaking of that - seems that wimpy current is pretty offten problem in audio gadgets ........ but that's just my opinion .......
same as all opinions of mine ....... are my opinions ......
I've played with cascodes (SK246) on JFET (SK170) input stages (single transistor, diff and complementary). In all these amps cascoding is one of the very few mods that actually changes the tonality of the amp. The sound is "lighter/leaner", cleaner and the mid is experienced as attenuated with a less forward presentation. The difference isn't major, but I believe it's there.
higher transconductance part for cascode result in less miller cap.
lower Cgs par for cascode produce less input referred noise.
Does it result in better sound? ur guess is better than mine.
lower Cgs par for cascode produce less input referred noise.
Does it result in better sound? ur guess is better than mine.
in this case also this thread could be of interest:
http://www.diyaudio.com/forums/pass...adin-no-memory-distortion-ultimate-aleph.html
second stage of figure 6.2 as first stage about
http://peufeu.free.fr/audio/memory/memory-6-test.html
include additional jFET input buffer
and for second stage Andrea Ciuffoli's "Power Follower" about
http://www.audiodesignguide.com/PowerFollower/index.html
must actually present the royal approach
http://www.diyaudio.com/forums/pass...adin-no-memory-distortion-ultimate-aleph.html
second stage of figure 6.2 as first stage about
http://peufeu.free.fr/audio/memory/memory-6-test.html
include additional jFET input buffer
and for second stage Andrea Ciuffoli's "Power Follower" about
http://www.audiodesignguide.com/PowerFollower/index.html
must actually present the royal approach
Last edited:
If you think about it, there is always modulation of the
voltage across the gain device. Cascoding allows you to set
its value in the range from plus to minus, which includes
zero. This is very useful in eliciting the response you want
from the part.
😎
one optimistic premise ( ZM ...... thinking
) , and one category which leads me in at least two paths ;do you mind just another small hint ..... just to ease my leads to "response" ..... is it useful swing , or ......
"Hey , Big Brother (Live) "
If I understood you correctly, you mean choosing a specific Vds (Vce, Vak...) value of the gain device to set the load line ?
Yes. That, coupled with choice of degeneration values on the
gain device, gives quite a variety of characteristics.
Plus you can drive the cascode transistor arbitrarily.
Alas, the Tad monitors went back to Andrew Jones.
Thanks for the loan, Andrew.
😎
- Status
- Not open for further replies.