yup
though - proper ratio of shunt /load currents starts with 1:1 , but rougher/wilder the circ, we must increase headroom
however , that roughness is limited with A class circs having moderate output currents ... and going with mentioned ratio above 3:1 is likely nothing else than waste of heat
btw. mentioning shunt regs - I'm not messing with PL construction decisions .......... I know that you boyz know how to make shunt reg , same as anything else
edit : what you are obviously lacking of , on that picture , is proper amount of mess on bench
that's too tidy .......
though - proper ratio of shunt /load currents starts with 1:1 , but rougher/wilder the circ, we must increase headroom
however , that roughness is limited with A class circs having moderate output currents ... and going with mentioned ratio above 3:1 is likely nothing else than waste of heat
btw. mentioning shunt regs - I'm not messing with PL construction decisions .......... I know that you boyz know how to make shunt reg , same as anything else
edit : what you are obviously lacking of , on that picture , is proper amount of mess on bench
that's too tidy .......
Last edited:
Hi Zen Mod,
I just happen to be a big fan of shunts ...that is a personal opinion and does not necessarily reflect the thoughts of other people at Pass Labs.
I have always been a fan of your work because you build stuff before you come to a conclusion.
As far as the mess goes I did clean the bench up a little for the picture. I will try to follow proper bench etiquette in future.
I just happen to be a big fan of shunts ...that is a personal opinion and does not necessarily reflect the thoughts of other people at Pass Labs.
I have always been a fan of your work because you build stuff before you come to a conclusion.
As far as the mess goes I did clean the bench up a little for the picture. I will try to follow proper bench etiquette in future.
It's a basic circuit and values are easy to calculate. The fun is in the additions: different types of current mirrors, cascodes, shortened feedback loop, no feedback loop, etc.
The playground is wide here so everyone can find what he likes the best...
Yeah Pico,
.........it would be Pico on the barbie!
I just had a vision.
https://www.youtube.com/watch?v=C1nxaQhsaaw
Juma, I could not agree with you more, I have been following your posts for some time now, keep up the good work.
Current mirrors are great building blocks and get you around some of the problems of a traditional VAS as long as some care is taken, but in the end it is how you use a given topology to meet your design objectives.
jam
Current mirrors are great building blocks and get you around some of the problems of a traditional VAS as long as some care is taken, but in the end it is how you use a given topology to meet your design objectives.
jam
Hi Hikari,
Juma is essentially correct about the topology but I use a different type of mirror in the HPA-1.
One thing I need to add is that most modern headphones, especially planers are a lot harder to drive than the traditional headphones so you have to adjust your designs accordingly. I am not married to this topology as each design is different, that being said I am a big fan of current mirrors in the right application.
Juma is essentially correct about the topology but I use a different type of mirror in the HPA-1.
One thing I need to add is that most modern headphones, especially planers are a lot harder to drive than the traditional headphones so you have to adjust your designs accordingly. I am not married to this topology as each design is different, that being said I am a big fan of current mirrors in the right application.
A question for you guys:
If there's a servo to null out the DC offset, then why are the coupling caps needed on the pre-amp out?
Does a servo work differently with a current feedback design? (My go-to book only covers voltage feedback, so I'm a little out of my depth here.) Or is there something else in play on the pre-amp side of things?
Thanks,
Jeff
If there's a servo to null out the DC offset, then why are the coupling caps needed on the pre-amp out?
Does a servo work differently with a current feedback design? (My go-to book only covers voltage feedback, so I'm a little out of my depth here.) Or is there something else in play on the pre-amp side of things?
Thanks,
Jeff
Servo works same for current feedback. Perhaps the preamp amp output is before the heavy biased mosfet output stage...less drive needed for line amp duty than headphones. And since headphone impedances are so low compared to power amp input impedances...servo makes more sense than cap.
Maybe?
Interesting theory, Hikari1.
The SILMIC line tops out at 3,300uF, which gets the corner down to 3Hz. So going with a cap for the headphones wouldn't be *hugely* different to what's in there.
Looking at it from the other angle, once you do have the servo (for whatever reason), is there any reason *not* to use it at the higher impedance of the preamp out? Is it less effective or something?
Cheers,
Jeff.
The SILMIC line tops out at 3,300uF, which gets the corner down to 3Hz. So going with a cap for the headphones wouldn't be *hugely* different to what's in there.
Looking at it from the other angle, once you do have the servo (for whatever reason), is there any reason *not* to use it at the higher impedance of the preamp out? Is it less effective or something?
Cheers,
Jeff.
Silmics are only available in low voltages up that high.
This is all purely speculation let's assume HPA-1 is 3 stages, with the last stage being a heavily biased MOSFET source follower. 1st stage is JFET input buffer, 2nd stage BJT VAS.
It's possible the line amp output is from the 2nd stage and the headphone output is from the 3rd. It's possible that only the 3rd stage is servoed.
This is all purely speculation let's assume HPA-1 is 3 stages, with the last stage being a heavily biased MOSFET source follower. 1st stage is JFET input buffer, 2nd stage BJT VAS.
It's possible the line amp output is from the 2nd stage and the headphone output is from the 3rd. It's possible that only the 3rd stage is servoed.