Low Idss.... why did you need to use such high Rs (100ohm)... the resulting CCS current will then be around 1mA.... why so low ?
May be I'm wrong but I think they are CSS for support BC550/BC560 and 1mA run through these transitor was a fine value to get enough Vbe for 2sk170. In that place I use 2SK246GR with 300R to get 2mA and I matched them (not required).
Thank you obscurus...
In my other shunt builds that use similar voltage error amp and vref setup I run the BC5** collector CCS at 4mA and got splendid results.
I was wondering why Dadod chose only 1mA.... maybe it is enough.... Will listen to the rig soon and post here subjective listening impressions 🙂
I checked my shunt and CCS was set to about 2 mA and it's not so critical, so you can use a bit higher Idss 2N5457. 100R will drop 0.2V at 2mA, so how did you com to 1 mA of the CCS current, you have to look at fet Vgs diagram, or better try with one with Idss of 2 mA and see what you get with 100R of the source resistor, and don't judge one design with other one just because there different current. Why do you think that 4 mA is better then 1 mA, it depends of a design?
I am not stating it is better or worst... just trying to understand the reasons that led you to choose these components values.
I realize now I made a mistake because I do not have a 2N5457 model and used a k170 model set to 2mA instead. (In this particular case with the 2mA Idss K170, Rs = 100 drops Id to around 1.2mA)
Your post made me look carefully at the 2N5457 pdf and naturally you are right. (Typical Vgsoff for the 2N5457 is much lower than what I find in a K170)
Please bear in mind I am building your circuit for Learning purposes so you are the Teacher and I am the Student here.
My questions might seem aggressive but that does not mean I am questioning your choices... just trying to understand them.
Anyway, now I see that you use 2mA for the CCS loading the error amp and that will be my goal also.
I realize now I made a mistake because I do not have a 2N5457 model and used a k170 model set to 2mA instead. (In this particular case with the 2mA Idss K170, Rs = 100 drops Id to around 1.2mA)
Your post made me look carefully at the 2N5457 pdf and naturally you are right. (Typical Vgsoff for the 2N5457 is much lower than what I find in a K170)
Please bear in mind I am building your circuit for Learning purposes so you are the Teacher and I am the Student here.
My questions might seem aggressive but that does not mean I am questioning your choices... just trying to understand them.
Anyway, now I see that you use 2mA for the CCS loading the error amp and that will be my goal also.
Just added the following model to my lib:
.MODEL 2N5457 NJF (VTO=-1.6 BETA=1.29M LAMBDA=2M RD=1 RS=1 CGD=6E-12
+ CGS=2.25E-12 KF=6.5E-17 AF=0.5)
Set to Vgsoff = -1.6v and 3.3mA Idss
In reality the 100 ohm does drop Id to 2.5mA
I am really sorry for the misunderstanding and all the trouble I managed to give you.... but now you know this is the way I get to learn
.MODEL 2N5457 NJF (VTO=-1.6 BETA=1.29M LAMBDA=2M RD=1 RS=1 CGD=6E-12
+ CGS=2.25E-12 KF=6.5E-17 AF=0.5)
Set to Vgsoff = -1.6v and 3.3mA Idss
In reality the 100 ohm does drop Id to 2.5mA
I am really sorry for the misunderstanding and all the trouble I managed to give you.... but now you know this is the way I get to learn
Don't worry, everything's OK, just ask any question but don't expect I know all the answers.
Thanks
Q9 Q6 Q8 Q10 are there to lower mirror output impedance ?
R50 // R17 are the resistors that convert current to voltage ?
Q9 Q6 Q8 Q10 are there to lower mirror output impedance ?
R50 // R17 are the resistors that convert current to voltage ?
By nature current conveyor is high impedance. Q9 Q6 and Q8 Q10, Baxandall supper pairs are there to lower distortion not output impedance.
R50//R17 in NGFB and R50 in CFA mode convert current to voltage.
So I was right about the current to voltage conversion...
Now the baxandall super pairs, I am not totally comfortable on how they work.
Would you acre to elaborate a little bit or give me a hint on where to find the right information ?
Now the baxandall super pairs, I am not totally comfortable on how they work.
Would you acre to elaborate a little bit or give me a hint on where to find the right information ?
So the baxandall super pairs are a variant of the darlington or CFP... they form a super transistor.... In this case are they used as followers ?
Amps are ready... now the shunt population will begin 🙂
About the termination... why 47r for phones and 33r for line out ?
What is the reasoning behind these values ?
Amps are ready... now the shunt population will begin 🙂
About the termination... why 47r for phones and 33r for line out ?
What is the reasoning behind these values ?
Baxandall super pair is not a variant of darlington or CFP and could be used in different configuration. Q11 Q3 and Q12 Q4 are Baxandall super pair too, used in similar way as a diamond.
33R for line out is protection from capacitive load, and 47R for phone out to accommodate different impedance headphones, from 32 ohm to 600 ohm.
Everything is in place but the power mosfets on the shunts.
I have two IRF640 but I do not have any IRF9640 .... can I use IRF9530 instead ?
I have two IRF640 but I do not have any IRF9640 .... can I use IRF9530 instead ?
I did not try that combination, it could produce unequal shunt voltage. Try it and decide by yourself. This is one of the question I can't answer with certainty.
Hi Dadod!
many people are using 640/9530 instead of 640/9640 because of better compatibility and almost equal parameters.
- Status
- Not open for further replies.