VAS Stage

JaredC79

Banned
2015-09-02 4:32 am
NJ
Can someone please really educate me on Vas/class A stage? First and foremost the math to raise the rails to the desired levels. Or is just the one set of transistors enough and other items are for stability and sound improvement? Secondly - the reason for it existing? If a amp needs 78-80v why not just use those as secondaries off the toroid and skip it all together? Would minimize part count, which is always good increase stability, make a simipler build. Input phase would just go right to the drivers w no increase in voltage, then output devices. Maybe my post makes no sense for those in the know but I question everything try to understand as much of the reasoning as possible and try and do things differently when I can to improve results. Can someone really drop some heavy knowledge about this stage?

Also, if I wanted to create a super fast amp (discrete) what is the best way to draw as much current - is it in the input portion?
 

JaredC79

Banned
2015-09-02 4:32 am
NJ
Well it seems like voltage rails can only really be increased by 1.44x. Other than ICs to lower rails - how far can you push voltage rails w trannies. My second comment was for the purpose of a VAS stage. If an amp if supposed to be 55/55 pumped to 78/78. Why not just skip vas and run at 78v.

Also if I want to create a super fast amp how do I do it? Is it increased current? If I am using a lm6171 as an in w a 15v I get a slew of 3600. I know switching will be slowed by other trannies but how can I get the best slew rate.
 

Cortez

Member
2002-12-26 9:45 am
Hungary
Skip VAS: ??? What does it mean?
And what's that 1.4x rail thing?
Please be more specific in what do you
have as a fundament and what do you
want to achieve and why...
Slew: that's a bit more complex then just
increasing current.
You have to design more thiings to achieve this.
 
Jared, looks like there is some misunderstanding here :confused:

VAS = Voltage Amplification Stage. That's the stage, giving you the full voltage swing you need. How are you goung to live without it? Are we on the same page?

I also don't understand 1.44x. Where does this number come from?

What is "slew of 3600"? Slew rate is measured in V/uS.
 

JaredC79

Banned
2015-09-02 4:32 am
NJ
Yes i know what VAs is voltage amplification stage. lets say you want your amp to run on 78v rails. Instead of using a 55v toroid jump out the vas stage and just use 78v secondaries. Ok maybe the line in needs a little juice so use an opamp with additional windings to run it, skip vas, power drivers right from secondaries at 78v.

1.4x seems to be the commonality of optimized designs in the VAS 55v gets 78v, 40 get you 56v. It is just a trend that i noticed by backing into the numbers from what people seem to talk about when discussing some of the better/best solid state designs. I am a finance geek by trade and my head just works weird with numbers and trends.

The LM6171 is am opamp - some people use it for audio but its mostly made for super high speed switching applications and doesn't have enough gain on its own. With 15v the slew is 3600v/us and the slew is dictated by the voltage coming in. So basically i am wondering if i got a 1kva tranny with 78v and 15v secondaries, i run the LM6171 as a line in the 0 carries the sound the 78v powers the lfets directly. i am over simplifying things but that is the gist, i would add additional lm 6171s for zero offset and possibly dc servo bias it...

hope i explained it correctly - i am new to building amps but making progress everyday and as i understand things more and more i have more questions and more questions and ideas about how to do things differently or change things to make them easier, better, more efficient, etc.
 
I think, I start getting your point. Are you talking about amplification by the means of a transformer? If yes, this approach is accompanied with a number of serious drawbacks - that's why it's not used widely in the solid state amplification (although transformers were used a lot in the tube amps). In solid state designs there are simply better ways of providing the voltage gain.

It is much easier to discuss something if there is a schematic diagram. Otherwise - too many variables, so whatever we are talking about - is too general. However, the devil - in details.

Cheers,
Valery
 

JaredC79

Banned
2015-09-02 4:32 am
NJ
I guess in essence yes it would be a 2 transformer design or 1 w additional windings. 15v for an opamp in - vas would be omitted and correct voltage for the outputs would be another winding. Of course both would need to be rectified and smoother. Now can we take it one step further, cutting the juice to the opamp in running it discrete or going with a cfa input phase - I say that because it seems to use less components and a larger cap can be use post input to push the small signal thru with more juice behind it.
 

Jay

Banned
2003-02-11 9:02 am
Jakarta
Secondly - the reason for it existing? If a amp needs 78-80v why not just use those as secondaries off the toroid and skip it all together?

The power supply is a bit irrelevant because it is DC. The music signal is AC, so we need a gain stage to increase small signal into large signal (which is the VAS)...

But this stage is very complex as large voltage (or gain) is not the only thing that matter. HF linearity is the primary issue. We need to design local feedback (OLG) for HF linearisation (and this also related with stability)...

Then the VAS (Cdom) need to be driven by sufficient current from previous stage...

Then the VAS output need to be converted into current, so we need to design the load or buffer for good power transfer... (this might be relevant to what you are looking for).

And high supply voltage has implication to what transistor that can be used in VAS. We want adequate Vce, Pd, beta, etc... and not all transistors have what we need.
 
The power supply is a bit irrelevant because it is DC. The music signal is AC, so we need a gain stage to increase small signal into large signal (which is the VAS)...

But this stage is very complex as large voltage (or gain) is not the only thing that matter. HF linearity is the primary issue. We need to design local feedback (OLG) for HF linearisation (and this also related with stability)...

Then the VAS (Cdom) need to be driven by sufficient current from previous stage...

Then the VAS output need to be converted into current, so we need to design the load or buffer for good power transfer... (this might be relevant to what you are looking for).

And high supply voltage has implication to what transistor that can be used in VAS. We want adequate Vce, Pd, beta, etc... and not all transistors have what we need.

I think, Jared is trying to think in the direction of replacing the VAS stage with a step-up transformer. Well, you know what that means ;) Possible, but... you get more issues than advantages :rolleyes: Bandwidth, frequency response linearity, impedance alignment, etc.
 

JaredC79

Banned
2015-09-02 4:32 am
NJ
I think, Jared is trying to think in the direction of replacing the VAS stage with a step-up transformer. Well, you know what that means ;) Possible, but... you get more issues than advantages :rolleyes: Bandwidth, frequency response linearity, impedance alignment, etc.

Exactly. Not a step up transformer as much as the same transformer with an other secondary winding. Using 1kva toroid with 78v and 15v secondaries both with FWB by schottsky diodes, You will still have the 0 rail for AC single to travel thru because the 78v would be the grounded center tap. LM6171 input with 15w to it, the single will haul a$$ at a 3600 slew rate, biased A, be fully linear and fully differential, with a huge bandwidth. Impedance alignment, i have to put a pen to paper and play around a bit or perhaps someone could assist. 78v would go to 4 or 5 pairs of LFets. The 6171 doesn't increase voltage much which isn't bad and the lfets will get protected by 14v Zeners, just need 50MA to the gates, and it should work.
 
Hi,

A rather clueless attempt to bypass common sense in amplifier design.

The output stage would need to be wired for voltage gain, rather than the
typical current gain, it can be done, but your completely missing the point.

All you will ever achieve using the LM6171 in an amplifier is not its specs,
but a high powered oscillator, if you ignore the amplifiers stability criteria.

rgds, sreten.
 

Ian Finch

Member
Paid Member
2010-04-11 4:22 am
Coffs Harbour, NSW
Exactly. Not a step up transformer as much as the same transformer with an other secondary winding. Using 1kva toroid with 78v and 15v secondaries both with FWB by schottsky diodes, You will still have the 0 rail for AC single to travel thru because the 78v would be the grounded center tap. LM6171 input with 15w to it, the single will haul a$$ at a 3600 slew rate, biased A, be fully linear and fully differential, with a huge bandwidth. Impedance alignment, i have to put a pen to paper and play around a bit or perhaps someone could assist. 78v would go to 4 or 5 pairs of LFets. The 6171 doesn't increase voltage much which isn't bad and the lfets will get protected by 14v Zeners, just need 50MA to the gates, and it should work.
I wish I could understand what you are saying :xeye: Perhaps I should translate into Russian and back to English to get the gist of it. That may be how it works for Valery :D

The LM6171 output only swings 12V from a specified supply of +/-15VDC - less into low impedance loads. http://www.ti.com/lit/ds/symlink/lm6171.pdf
Your 78VAC power transformer windings will give DC power rails of about +/-110V and the signal voltage will need to swing >200V to keep up with this. The slew rate will not be limited only by the VAS but who needs 3,600V/uS in an audio amplifier anyway?
 
RCA High Power Application Note

You need a voltage amplification stage between the op.amp and the output driver and power output stage combination.

Transistors 2N3440 and 2N5416 in the attachment illustrate the point.

Note the transformer secondaries are 60 volts a.c. while yours are 78 volts a.c. so as it is the circuit is not suitable for your purpose. In either case without a lot of experience in building amplifiers something like this would be like dealing with dynamite.
 

Attachments

  • IMG_20160302_0001_NEW.pdf
    144.9 KB · Views: 62

JaredC79

Banned
2015-09-02 4:32 am
NJ
Hi,

A rather clueless attempt to bypass common sense in amplifier design.

The output stage would need to be wired for voltage gain, rather than the
typical current gain, it can be done, but your completely missing the point.

All you will ever achieve using the LM6171 in an amplifier is not its specs,
but a high powered oscillator, if you ignore the amplifiers stability criteria.

rgds, sreten.

Val pointed the issues out and and how it can be done - the lm6171 solves a lot of them if the initial power supply is decoupled from the larger power rails - as per his comment impedance would be the final issue. Clueless and unconventional are two different things - if common sense is following the leader only changing a transistor Or cap value or bootstraping something rather than developing something new and eliminating a stage that is problematic and a source of distortion well I guess I am his plain old stupid - if I do things I try to be innovative and figure this would be the place to gain knowledge and questioned the need in the first post. It was already discussed that 2 transformers can do it. Feel free to do what you want in your spare time, I'll build a powered oscillator if that's the only outcome.

6171 can be stable to 15v if I am not mistaken. I have to go over the white pages but knew them in and out just learned a ton somethings get blurred, If not then it's 12. Regardless of what it is, then that's what I will use, if it needs more juice After the lm I can buck boost it but what I put thru it ill make sure it won't oscillate.

How does skipping the vas give 78v windings a gain in voltage? If the jump is still there the other windings can be 55v to get to 78, once the impedance issues are compensated for an mpsa350 can drive the lfets.
 

JaredC79

Banned
2015-09-02 4:32 am
NJ
I wish I could understand what you are saying :xeye: Perhaps I should translate into Russian and back to English to get the gist of it. That may be how it works for Valery :D

The LM6171 output only swings 12V from a specified supply of +/-15VDC - less into low impedance loads. http://www.ti.com/lit/ds/symlink/lm6171.pdf
Your 78VAC power transformer windings will give DC power rails of about +/-110V and the signal voltage will need to swing >200V to keep up with this. The slew rate will not be limited only by the VAS but who needs 3,600V/uS in an audio amplifier anyway?

Wish I could describe it a little better but maybe he will get the idea. Part of my last post was for you. So if voltage is getting kicked up even without a true vas stage, the other secondaries would have 55v to get it to roughly 78 v. Slew will be limited to driver and output devices as well - the idea is that it isn't limited by input stage or vas. Minimalist approach sort of thing.
 
Wish I could describe it a little better but maybe he will get the idea. Part of my last post was for you. So if voltage is getting kicked up even without a true vas stage, the other secondaries would have 55v to get it to roughly 78 v. Slew will be limited to driver and output devices as well - the idea is that it isn't limited by input stage or vas. Minimalist approach sort of thing.

Can you please draw what it's going to look like.

I'm also just sort of... guessing what you mean :D Some of your statements don't make sense to me (Ian, you're not alone ;)). Maybe it's a matter of misunderstanding from my side.

A simple diagram will bring us closer to being on the same page.