In a VAS stage, I often see a small capacitor from collector to base at the transistor of the negative side (NPN).
Typical value is 10 to 100pF.
Why is it on the negative rail side only and never at the positive side where the PNP is?
What does it do? How do I calculate its value?
Regards
Typical value is 10 to 100pF.
Why is it on the negative rail side only and never at the positive side where the PNP is?
What does it do? How do I calculate its value?
Regards
The cap is there to slow down the VAS stage so it have no speed available to compensate errors of output emitter followers. As the result, preceding differential stage trying to catch up jumps out of trousers creating new problems that are solved by different means that in turn create own problems, that are solved again, creating new problems, and so on... It is called a steady technical progress. As the result, we have an amp that measures according to today's standard measurement procedures as a straight amplifying wire.
Now for a serious reply: It's the Miller capacitor, often labelled Cdom because its function is dominant pole compensation.
Almost all amplifiers will require frequency compensation of some sort to prevent oscillation, and dominant pole compensation is the easiest. Calculating the optimum value is hard; the easiest way is to build the amplifier and test it with various values. 100pF is a pretty good starting point because it will ensure stability in almost all amplifiers.
You only see one because it is connected across the voltage amplifying transistor; the PNP transistor on the other side is part of a current source. That is in a standard amplifier anyway - other possibilities exist.
Almost all amplifiers will require frequency compensation of some sort to prevent oscillation, and dominant pole compensation is the easiest. Calculating the optimum value is hard; the easiest way is to build the amplifier and test it with various values. 100pF is a pretty good starting point because it will ensure stability in almost all amplifiers.
You only see one because it is connected across the voltage amplifying transistor; the PNP transistor on the other side is part of a current source. That is in a standard amplifier anyway - other possibilities exist.
Stops oscilation.
this is what the one does...without it you may have oscilations... this is the main practical result... practical daily use.
More than that is impedance, transference, waveshape, lag compensation and a lot of interesting names and effects..but the main effect is to stop oscilation.
If it is too much big..your amplifier distorts high frequencies....if too small or absent you will be endangering your amplifier... can oscilate, overheat and burn.
regards,
Carlos
this is what the one does...without it you may have oscilations... this is the main practical result... practical daily use.
More than that is impedance, transference, waveshape, lag compensation and a lot of interesting names and effects..but the main effect is to stop oscilation.
If it is too much big..your amplifier distorts high frequencies....if too small or absent you will be endangering your amplifier... can oscilate, overheat and burn.
regards,
Carlos
Ok I think I remember the term "Miller cap" from somewhere. Thank you all.
My amp is fully symmetrical, positive and negative sides are identical in parts and values, only NPN and PNP are swapped of course. I don't see a current source in the VAS, it is just a resistor, cascoded PNP, PNP, outputstage-bias-regulator-thingie, NPN, cascoded NPN, resistor, negative rail. That's it.
The Miller cap (and the feedback network) are the only things of the circuit which are not symmetrical. So I wondered if there is an error in my schematic.
My amp is fully symmetrical, positive and negative sides are identical in parts and values, only NPN and PNP are swapped of course. I don't see a current source in the VAS, it is just a resistor, cascoded PNP, PNP, outputstage-bias-regulator-thingie, NPN, cascoded NPN, resistor, negative rail. That's it.
The Miller cap (and the feedback network) are the only things of the circuit which are not symmetrical. So I wondered if there is an error in my schematic.
EE said:
Besides what others said ...used for freq compensation ie loop stabilty, it's also usually the limiting factor in the slew rate of an amp.
from I=Cm*dV/dt solve for slew rate ie dV/dt where roughly I = collector current Vas stage/Cm = miller cap.
The "high side" transitor is a current source for the Vas stage.
millers capacitor
Hi guys, my first post so go easy lol
lets get this right then, to stop an amplifying transistor from oscillating due to the miller effect, which i understand to be capacitance signals being sent from collector to base, and being amplified over & over, you'd place a suitable capacitor, call this X across the collector & base, which would offer very high resistance at high frequencies, and virtually no feedback signal, filtering in effect high frequency signals caused by the miller effect away from being amplified over & over, when the frequency of these unwanted miller capacitor signals lets call them that are low, then the capacitor X charges, and discharges when the transistor is in its off state, that way the transistor doesn't amplify any unwanted miller signal. As a diy guy is my understanding correct?
Thanks
Hi guys, my first post so go easy lol
lets get this right then, to stop an amplifying transistor from oscillating due to the miller effect, which i understand to be capacitance signals being sent from collector to base, and being amplified over & over, you'd place a suitable capacitor, call this X across the collector & base, which would offer very high resistance at high frequencies, and virtually no feedback signal, filtering in effect high frequency signals caused by the miller effect away from being amplified over & over, when the frequency of these unwanted miller capacitor signals lets call them that are low, then the capacitor X charges, and discharges when the transistor is in its off state, that way the transistor doesn't amplify any unwanted miller signal. As a diy guy is my understanding correct?
Thanks
Last edited:
I recently picked up a Maplin 50 watt amp and that was oscillating a bit becasue that capacitor wasnt in the design. I used slowly increasing values until I found 1000pf stopped the oscillation.
1000pf sounds a bit high as i have used 220pf in my own class AB designs and had good results.
1000pf sounds a bit high as i have used 220pf in my own class AB designs and had good results.
download Leach's Lo Tim papers. He explains in a lot of detail how the various sections of an amplifier work.
He also explains how to join the pieces together to create a whole.
You description really only looks at the little part around the Compensation capacitor (Ccomp) and as a result comes out with wrong conclusions.
It's the effect the Ccomp has on the frequency response of the whole amplifier that makes Ccomp work well.
Yes, it is referred to variously as C Dom, Miller (comp) cap. Is this also lag compensation?
A google on these could find some simple explanations.
- Status
- Not open for further replies.