Whatever works out us good enough. I will try to parallel 10 BFs and give it a try and if works out obviously the interfet part will be the part I am looking for.
Overall we are talking about 40 jfets paralleled for balance right? If they are biased at 6mA each 240mA for front stage seems crazy though.
Here is the other big pro of the if part which gets up to that gm with only 10mA am I wrong?
Last edited:
MiiBs one shot will need 20 Bfs. Make that 16 because when you parallel more then 8 per side nothing happens any more. Yes, run the input stage hot, there is no other way.
This guy is actually not my friend but his noise calculations are hard to criticize.
Here you can learn everything you have to know about designing for low noise at low impedances with BJT and JFet.Low Noise Designs
When you have swallowed that pill we can proceed. Noise, hum and distortion is simply not allowed when you design for SOTA.
Here you can learn everything you have to know about designing for low noise at low impedances with BJT and JFet.Low Noise Designs
When you have swallowed that pill we can proceed. Noise, hum and distortion is simply not allowed when you design for SOTA.
I had already read that and talked a bit to the guy too.
However I will re-read it again.
The good thing about BF862 is that they need less current to sound good than the SK170's so you can run them at may 3 ma each... so in total you get around 50 mA for a LTP input.
that is pretty good.
Do you guys think I can make a SE version with BFs to compare against Paradise?
Making the balance version and then changing cables and topology running the preamp balance wouldn't give me a straight idea whether this topology has potential to sound better or not.
I know that with balance I gain 6dB gain.
What do you guys think? Would this be a logical approach?
good. But will I get enough gain out of it paralleling let's say 8 BFs in Single End configuration?
here is the simulation.
I can't get anymore than this gain out of it and apparently I need 100mA at the input.
Beside switchin on the simulation from K170 to BF862 didn't bring any gain up.
Please help with the simulation (MiiB?!?) -models attached here-
I can't get anymore than this gain out of it and apparently I need 100mA at the input.
Beside switchin on the simulation from K170 to BF862 didn't bring any gain up.
Please help with the simulation (MiiB?!?) -models attached here-
Attachments
Ok, i see. You are loosing 6dB gain because there is no P-channel compliment.
The other way is using double value resistors and half value caps in the RIAA.
That gives you 6dB of gain.
The other way is using double value resistors and half value caps in the RIAA.
That gives you 6dB of gain.
yes you are correct, no P-channel.
Is change value on the RIAA the only option? I had those custom values and change will completely change the set up.
how am I going to raise the output impedance on the current mirror? Isn't that a function of rbe of the BJTs?
Simulation shows no increase on gain paralleling more ..it actually collapses (I think I loose control over the DC offset)
If anybody here wants to help with the simulation, please go ahead.
Yes, more then 8 Bfs at the input brings trouble. To raise the output impedance of the mirror you can try to raise the value of the 120 Ohm emitter resistors. The other is "cascoding the cascode", so put an additional cascode at the output of the mirror.
all that inner section of mirrors is not needed....there are simpler and better ways to shield the servo...
You can raise the resistors to 47 ohms each..and you can scale the mirror by using 220/110 or 220/47 ohms that will give you more current so you can get the gain you need by scaling the riaa components.
You can raise the resistors to 47 ohms each..and you can scale the mirror by using 220/110 or 220/47 ohms that will give you more current so you can get the gain you need by scaling the riaa components.
how would you shield the servo differently from this way?
Something I said I would have liked not to do is changing the RIAA values to fit the custom capacitors I have.
But it looks like there is no other way around to get more gain.
you could shield the current output by a jfet cascode... that will have substantial higer impedance than the mirrors
funney you want to save something so insignificant, and compromise in other vital areas... make the circuit using cheap parts to eval it...
if good and stable then you can go down the exotic route....
But I must say it again... in this kind of riaa the components matter a lot less.... and teflon caps are for many reasons vastly inferior to other industrial types of caps....
funney you want to save something so insignificant, and compromise in other vital areas... make the circuit using cheap parts to eval it...
if good and stable then you can go down the exotic route....
But I must say it again... in this kind of riaa the components matter a lot less.... and teflon caps are for many reasons vastly inferior to other industrial types of caps....
Last edited:
- Status
- Not open for further replies.