I have previously designed BJT complimentary output stage based
amplifier. It is first time i am going to do a mosfet Class AB amplifier.
I want it to be Quasi Complimentary NMOS output stage, preferably IRFP240
as Models are easily available
How to protect mosfets from blowing up due to oscillation as the
amplifier will work up to 300Khz?
Design Target:
*2W class A region before moving to AB
*300Khz Bandwidth
*At least 100V/us Slew-rate
*Low phase shift [ Less than 6degree]
Ripple eater are taken from Wolverine Amplifier
Please help me to setup spice to find out THD, and output stage.
How to measure slew rate and check out phase margin under Lt-spice?
Preliminary Input and VAS stage , i have attached
Edit: Removed *Low Phase Margin [Preferably Under 6d]
amplifier. It is first time i am going to do a mosfet Class AB amplifier.
I want it to be Quasi Complimentary NMOS output stage, preferably IRFP240
as Models are easily available
How to protect mosfets from blowing up due to oscillation as the
amplifier will work up to 300Khz?
Design Target:
*2W class A region before moving to AB
*300Khz Bandwidth
*At least 100V/us Slew-rate
*Low phase shift [ Less than 6degree]
Ripple eater are taken from Wolverine Amplifier
Please help me to setup spice to find out THD, and output stage.
How to measure slew rate and check out phase margin under Lt-spice?
Preliminary Input and VAS stage , i have attached
Edit: Removed *Low Phase Margin [Preferably Under 6d]
Last edited:
"Design Target:
(...)
*Low Phase Margin [Preferably Under 6d]"
Assuming d stands for degrees: so it's a requirement that it has to be at the edge of oscillating?
(...)
*Low Phase Margin [Preferably Under 6d]"
Assuming d stands for degrees: so it's a requirement that it has to be at the edge of oscillating?
Did you post the wrong spice file? What I see is a ~20% finished circuit with the current mirror messed up.
Why do you want a quasi output? A quasi circuit makes life difficult and a fast slew rate near impossible.
Class A up to 2W for 8 Ohms just means a ~700mA idle current.
Why do you want a quasi output? A quasi circuit makes life difficult and a fast slew rate near impossible.
Class A up to 2W for 8 Ohms just means a ~700mA idle current.
Being your first attempt why set such high targets?
Looking at the specs listed and at the attempt of a schematic it is clear you have a very long way to go before you can really design a power amplifier.
Because copying segments of other people's work and stitching them together without proper knowledge of the theory involved, isn't designing at all.
I am sorry, it is not my intention to discourage any new enthusiast in electronics, but the thing is you must respect electronics enough to follow the proper steps in your evolution, you cannot jump from crawling on 4 limbs directly to running a 100m dash course...
I do not think you understand what it means to design a amplifier with more then 100V/uS, i do not even think you really know what it means to have such low phase shift ( not even sure why you would want it ), as for the high bandwidth, that would require either high OL amplitude or bandwidth ( or both ), or even low enough CL, and very tight controlled compensation...
All of them involves high level of audio electronics, which forgive me if it doesn't sound to your liking, seams unlikely for you at the moment.
Maybe you should read more theory before you start practical work, and start with easy stuff...
All the best.
Looking at the specs listed and at the attempt of a schematic it is clear you have a very long way to go before you can really design a power amplifier.
Because copying segments of other people's work and stitching them together without proper knowledge of the theory involved, isn't designing at all.
I am sorry, it is not my intention to discourage any new enthusiast in electronics, but the thing is you must respect electronics enough to follow the proper steps in your evolution, you cannot jump from crawling on 4 limbs directly to running a 100m dash course...
I do not think you understand what it means to design a amplifier with more then 100V/uS, i do not even think you really know what it means to have such low phase shift ( not even sure why you would want it ), as for the high bandwidth, that would require either high OL amplitude or bandwidth ( or both ), or even low enough CL, and very tight controlled compensation...
All of them involves high level of audio electronics, which forgive me if it doesn't sound to your liking, seams unlikely for you at the moment.
Maybe you should read more theory before you start practical work, and start with easy stuff...
All the best.
Sorry i inverted the current mirror ,
Quasi complimentary MOSFET output stage gives a dominant even order harmonic profile.
Quasi complimentary MOSFET output stage gives a dominant even order harmonic profile.
That could be better accomplished by proper degeneration.
PS: No decent power amplifier schematic ever needs that cap multiplier crap...
Maybe it would be needed for tube amplifiers which inherently have low supply rejection, but good quality transistor amplifier have more than enough PSRR, to handle a few volts of ripple.
PS: No decent power amplifier schematic ever needs that cap multiplier crap...
Maybe it would be needed for tube amplifiers which inherently have low supply rejection, but good quality transistor amplifier have more than enough PSRR, to handle a few volts of ripple.
I don't know if it will be enough
BJT amp of mine with same input stage and VAS had PSSR of 92db at 1KHz
BJT amp of mine with same input stage and VAS had PSSR of 92db at 1KHz
They say anything higher than 70db at 1Khz is good for 20Khz bandwidth amp , but how much is good for 300Khz?
I haven't decided whether i will do miller or TPC for compensation ,
OPS to begin with
I haven't decided whether i will do miller or TPC for compensation ,
OPS to begin with
Improve PSRR more than it in audio band freq.
Try using Miller Comp + Lead Comp. Reach 70deg phase margin & 14dB gain margin with tian probe
Try using Miller Comp + Lead Comp. Reach 70deg phase margin & 14dB gain margin with tian probe
with respect to your work,
you have reading to much Hi-Fi Guru nonsens,
such design does not make any audible sense except to increase the BOM costs to get nothings
Hi Dibya,
I am still working on the Xmas amp along with the FH9, hope to finish off in 2022 if I can find the time. I am going to build the "Very Simple Quasi MFet" amp that Christian and Hugh designed. Would this not suit your purposes, or at least give you a base to modify from, depending on your needs.
MM
I am still working on the Xmas amp along with the FH9, hope to finish off in 2022 if I can find the time. I am going to build the "Very Simple Quasi MFet" amp that Christian and Hugh designed. Would this not suit your purposes, or at least give you a base to modify from, depending on your needs.
MM
Hi Dibya,
Good luck on your new amp design. Most stuff under the sun has already been done already and all of it is really a rearrangement of previous blocks of technology. How many amps here have an LTP in the input stage? How many have complementary N and P MOS on the output stage with emitter resistors? Start with a working amp and modifying it for a goal of something improved is much easier than designing from scratch. I wish I had this much curiosity and motivation to design and simulate an amplifier when I was a teenager! Don’t let people throw water on your fire of a desire to learn.
Cheers,
X
Good luck on your new amp design. Most stuff under the sun has already been done already and all of it is really a rearrangement of previous blocks of technology. How many amps here have an LTP in the input stage? How many have complementary N and P MOS on the output stage with emitter resistors? Start with a working amp and modifying it for a goal of something improved is much easier than designing from scratch. I wish I had this much curiosity and motivation to design and simulate an amplifier when I was a teenager! Don’t let people throw water on your fire of a desire to learn.
Cheers,
X