vzaichenko
I was using that kind of Yamaha VAS (recommended by mr Mile) in few my amplifiers and must say that is very very good singing voltage stage, I am not using EF VAS since.
I even putted it in very popular AX11 amplifier with very good reults.
I was using that kind of Yamaha VAS (recommended by mr Mile) in few my amplifiers and must say that is very very good singing voltage stage, I am not using EF VAS since.
I even putted it in very popular AX11 amplifier with very good reults.
Those are super pairs in the VAS and CCS. Similar looking to a CFP but different arrangement.
Enter the "evil" IC (eyesee) ...
One of the {different} implementations of a op-amp V to I converter .
(below 1).
U2 is a unity gain follower ... and a current sink. R8 and 7 are
the current feedback loop. U1 is a current sinking inverting servo.
I've gotten crazy good operation at <5ppm using some op-amp
models and some ringing with other models.
Since this is a true current FB/IC "abomination" , slew is stellar ...
and compensation is easy (C6 and 5).
Only a couple Arcam IPS's have a similar circuit (prior art).
Any opinions are welcome. An IC based amp would be a nice
addition to the "lineup" 😀 .
OS
One of the {different} implementations of a op-amp V to I converter .
(below 1).
U2 is a unity gain follower ... and a current sink. R8 and 7 are
the current feedback loop. U1 is a current sinking inverting servo.
I've gotten crazy good operation at <5ppm using some op-amp
models and some ringing with other models.
Since this is a true current FB/IC "abomination" , slew is stellar ...
and compensation is easy (C6 and 5).
Only a couple Arcam IPS's have a similar circuit (prior art).
Any opinions are welcome. An IC based amp would be a nice
addition to the "lineup" 😀 .
OS
Attachments
Boris, right - Yamaha arrangement also gives very good results, it's just a different nature and transistos are placed somewhat vice-versa.
Cheers,
Valery
Same deal (superpair) ... but one is a "super" CCS and the other is active.
At clipping the clamping would be similar to my symmetrical versions.
clipping might not be as "anally" symmetrical (within a mV).
Still ,a nice design !
OS
You're right. The bottom of the sine clips slightly before the top and is slightly more rounded. I thought something was wrong but square waves look good and it sounds lovely. As I said, I also used the front end of this as an IPS with the Slewmaster and it performs very well.
Attachments
I sim super pair have higher THD than enhanced beta (2EF) VAS. I sim several topology amp using super pair VAS.
My friend who built AX-14 complain about the quality of high frequency.
My friend who built AX-14 complain about the quality of high frequency.
I built AX-14 3 ways. Stock BJT, with vertical MOSFET and N channel only MOSFET. Actually 4 ways counting SLewmaster. The highs sound good on mine but the soundstage is a little less broad than the VSSA. The best one is the slewmaster version.
Last edited:
There is an important detail here. First transistor in the superpair has to work at extremely low Vce (and rather low Ic at the same time). Not every transistor works good in such conditions. BC550/560 are the best ones I'm aware of for low voltage. Did you try them? Like BC550+KSA992 for example.
Also, in some cases superpair tends to oscillate at some MHz (due to the positive feedback of the base currents) - requires appropriate compensation...
TubaSui v2.0 - ultra-linear symmetric low-TIM hybrid
Hi All,
Another contribution to front-ends collection. Still a little bit under construction - no prototype yet. This is a revised version of my initial hybrid. I was working on it, mostly focusing on keeping the balance during start-up, when OS inspired me by super-pair VAS. So, this is what I came up with in the end.
THD20 is not measurable by Multisim even at 40V RMS output swing, low TIM by design, nice clipping. Much better behavior at start-up, thanks to a-la-Sansui symmetric mirrored LTP approach.
This is the next one I will prototype - working on PCB...
Comments are welcome 😉
Cheers,
Valery
Hi All,
Another contribution to front-ends collection. Still a little bit under construction - no prototype yet. This is a revised version of my initial hybrid. I was working on it, mostly focusing on keeping the balance during start-up, when OS inspired me by super-pair VAS. So, this is what I came up with in the end.
THD20 is not measurable by Multisim even at 40V RMS output swing, low TIM by design, nice clipping. Much better behavior at start-up, thanks to a-la-Sansui symmetric mirrored LTP approach.
This is the next one I will prototype - working on PCB...
Comments are welcome 😉
Cheers,
Valery
Attachments
I'm wondering, the IPS for the slewmaster are set at 4.5mA. Can the KSA992 handle that?
Terry, good point. They can if used at relatively low voltage (with cascode for example). Otherwise, if used at full swing, 2SA1381 would be a good option.
