Hi!
We know that a constant current source loaded VAS is good.
In fact most amplifiers uses CCS instead of resistor to bias.
This is because CCS puts minimal burdon onto the stage and so reduces distortion.
But if this is true for VAS, then it is true for input stage.
Anything that upsets, loads input stage will have great influence on performance.
I happend to discover the great benefit of using constant current source load for input stage while I was playing with my spice models.
If you look at my first picture, amplifier to the left, you see what we can call a blameless with darlington VAS.
All very normal standard.
Resistor R1 (in red) sets 1 mA in input transistor.
At the right you see this resistor is replaced with 1 mA CCS.
This improves drastically.
When we hunt small things here and there, this fairly simple modification take us almost to another level.
At 16 Watt I could see 18 times lower THD in %.
At 1 Watt 5 times lower THD.
My second picture shows one of my amplifier models using such a current source.
It is a low power amplifier. But it is Hifi.
I don't know why this easy trick not is used more.
In fact I can not remember I have seen it in use.
And I have seen a lot of schematics through the years.
But nothing is new, so for sure it has been done.
We know that a constant current source loaded VAS is good.
In fact most amplifiers uses CCS instead of resistor to bias.
This is because CCS puts minimal burdon onto the stage and so reduces distortion.
But if this is true for VAS, then it is true for input stage.
Anything that upsets, loads input stage will have great influence on performance.
I happend to discover the great benefit of using constant current source load for input stage while I was playing with my spice models.
If you look at my first picture, amplifier to the left, you see what we can call a blameless with darlington VAS.
All very normal standard.
Resistor R1 (in red) sets 1 mA in input transistor.
At the right you see this resistor is replaced with 1 mA CCS.
This improves drastically.
When we hunt small things here and there, this fairly simple modification take us almost to another level.
At 16 Watt I could see 18 times lower THD in %.
At 1 Watt 5 times lower THD.
My second picture shows one of my amplifier models using such a current source.
It is a low power amplifier. But it is Hifi.
I don't know why this easy trick not is used more.
In fact I can not remember I have seen it in use.
And I have seen a lot of schematics through the years.
But nothing is new, so for sure it has been done.
Attachments
Problem is that the differential output can
now swing to almost 0V, that is, U4 in the right picture
can now be supplied one milliamp of base current
and with no more base voltage limitation,that was
provided by the resisive loading, thus, it will
be subject to eventual peak currents of 1mA x hfe,
about 500mA if ever the output is shorted or if the amp clip..
now swing to almost 0V, that is, U4 in the right picture
can now be supplied one milliamp of base current
and with no more base voltage limitation,that was
provided by the resisive loading, thus, it will
be subject to eventual peak currents of 1mA x hfe,
about 500mA if ever the output is shorted or if the amp clip..
The standard "blameless" current mirror has this same major reduction effect on THD. It is not a "pure" CCS , but it does regulate the current , just between the pair. A THD shootout between my "BX" (simple 4 transistor input /VAS) and the "AX" blameless at 20KHZ is drastic , almost .1% for the mirrorless BX versus .001% for the mirrored "blameless".
The difference is the sound and THD harmonic residuals which result by the use of these techniques. Some prefer the simple amp with the very dominant 2nd harmonic (DX , aksa55 ,BX) I listen to these in my shop every day ... they are not offensive in the least , actually very pleasant at normal levels. 🙂
At concert levels is where the ultra low THD amps really "shine" CRANKED UP , as long as what little distortion they do have is H2/3 with little or no higher harmonic content. What you experience with this setup is "headphone" imaging and sound pressure levels, just in a whole room. Some reviews interpret this as "unstrained" or "effortlessness" , it is a different experience. I could never hear this on any of my sony receivers or even my beloved JVC "super A" ... I had to go to the 70-0-70VDC/80K uf /200W+ DIY creations for this "fix".
Don't forget about the VAS , when you combine a "blameless " front end with a "super pair" VAS, I have simulated almost to the PPM level at 1Khz with just 9 cheap trannies 😀 !
OS
The difference is the sound and THD harmonic residuals which result by the use of these techniques. Some prefer the simple amp with the very dominant 2nd harmonic (DX , aksa55 ,BX) I listen to these in my shop every day ... they are not offensive in the least , actually very pleasant at normal levels. 🙂
At concert levels is where the ultra low THD amps really "shine" CRANKED UP , as long as what little distortion they do have is H2/3 with little or no higher harmonic content. What you experience with this setup is "headphone" imaging and sound pressure levels, just in a whole room. Some reviews interpret this as "unstrained" or "effortlessness" , it is a different experience. I could never hear this on any of my sony receivers or even my beloved JVC "super A" ... I had to go to the 70-0-70VDC/80K uf /200W+ DIY creations for this "fix".
Don't forget about the VAS , when you combine a "blameless " front end with a "super pair" VAS, I have simulated almost to the PPM level at 1Khz with just 9 cheap trannies 😀 !
OS
Last edited:
A CCS (or current mirror) here will increase open loop gain, and so reduce closed-loop distortion. However, it will also reduce the open loop bandwidth because the Miller-enhanced Cdom on the VAS is now fed from a higher impedance. Some people worry about this. Others don't.
- Status
- Not open for further replies.