Removing Shunted Vas Capacitor

[lubbie]

I've looked everywhere. Everyone says that a shunting the Vas is just horible.

My question is.. If i rip it out and change it to Miller Capacitor compensation what can i expect from the changes.

[sreten]

Simply put lower distortion and vastly improved slewrate.

Distortion at high levels at high frequencies is likely to be very
poor due to low slew rate limiting, due to not being able to
provide the current required (>100mA) to slew the shunt
capacitor the required voltage swing.

A miller capacitor will required < 1/100 of the current of a shunt
capacitor, I'd say this will be the major effect, though it will give
some other much less drastic improvements.

sreten.

[padamiecki]

Originally posted by lubbie
I've looked everywhere. Everyone says that a shunting the Vas is just horible.

My question is.. If i rip it out and change it to Miller Capacitor compensation what can i expect from the changes.

hm,
why not to try to put a resistor instead of a capacitor? He would introduca a local nfb makes ths VAS more linear.
But be careful because the working point of the output devices might be change!
And the input sensitivity will be lower.

[sreten]

Originally posted by padamiecki

hm,
why not to try to put a resistor instead of a capacitor? He would introduca a local nfb makes ths VAS more linear.
But be careful because the working point of the output devices might be change!
And the input sensitivity will be lower.

Hmmmmm..... clear as mud.

The Miller capacitor cannot be replaced by a resistor,
niether can be the shunt capacitor.

The Miller capacitor by local negative feedback makes the
VAS more linear, this action increasing with frequency,
another advantage over the shunt capacitor.

The Miller capacitor can be parralleled with a resistor reducing
open loop gain at low frequencies, this does not change input
sensitivity which is determined by feedback, it changes the
amount of effective feedback at low frequencies.

Output working point is also stabilised by feedback.

If you say you are talking about no global feedback then note
that there is no need for a VAS capacitor at all, which is hardly
the point of this thread, global feedback is implicit.

sreten.

[padamiecki]

Originally posted by sreten


Hmmmmm..... clear as mud.

The Miller capacitor cannot be replaced by a resistor,
niether can be the shunt capacitor.

The Miller capacitor by local negative feedback makes the
VAS more linear, this action increasing with frequency,
another advantage over the shunt capacitor.

The Miller capacitor can be parralleled with a resistor reducing
open loop gain at low frequencies, this does not change input
sensitivity which is determined by feedback, it changes the
amount of effective feedback at low frequencies.
depend of its value
Output working point is also stabilised by feedback.
did you mean idle current or 0Vdc at the output?
If you say you are talking about no global feedback then note
that there is no need for a VAS capacitor at all, which is hardly
the point of this thread, global feedback is implicit.
agree
thought about the seroius value of this resistor, for example 100k - 470k. Such a value will impact not only for a low frequencies.
sreten.

This operation will allow to make weaker global nfb.
if you do not change the deepness of the nfb this will impact on the amplifier's input sensivity, what you think?

[sreten]

Originally posted by padamiecki
agree
thought about the seroius value of this resistor, for example 100k - 470k. Such a value will impact not only for a low frequencies.

This operation will allow to make weaker global nfb.
if you do not change the deepness of the nfb this will impact on the amplifier's input sensivity, what you think?

A resistor in parrallel with the Miller capacitor will not effect input sensitivity.

It will reduce low frequency open loop gain and thus the amount
of actual feedback applied closed loop at low frequencies.

Reducing the amount of effective feedback at low frequencies
has the negative effect of reducing negative rail PSRR by the
same amount, as -ve PSRR closely tracks the amount of feedback.

sreten.

[lubbie]

For everyone's info.. This is a working amp and had an extesive production run. I wanted to modify this 1970's amp to squeeze a little more performance out of it. Although this amp is no slouch as it is.. 200W + RMS. 12 deg laging at 20khz but had .1% THD at full rated power.

A comment was made about the current requirement of the shunted VAS stage, but due to the amp being a production unit, I believe that this has been compensated by Bob ********R when he designed it.

I was going to lower the 180pf cap to about 100pf as DSelf's & Randy Slone's book recommended.

I guess my my main question is will this amp have a hissy fit with the mod.... Also if someone has completed such mod, what did you encounter after the mod.. Maybe the possibility of the amp to become a better oscillator then power amp.

[sreten]

180pF is a very unlikely value for a shunt VAS capacitor,
it is a very likely value for a Miller compensation capacitor.

Reducing Cdom is likely to cause oscillation, well at least
reduce the gain and phase margins to oscillation point.

You can reduce Cdom if you also increase closed loop gain.

sreten.

[lubbie]

Well i confirmed that the amp has 180pf shunt capacitor... I might just change it from shunted to miller compensated... Hmmmm..

[AKSA]

Lubbie,

Can you perhaps show us a hand drawn schemat of the VAS and its shunt capacitor?

Cheers,

Hugh