I was looking into PSU regulation of a linear power supply.
Linear regulator produce massive amounts of heat and need a lot of work to get enough current out of them.
Does anyone have any experiences with step down switching regulators in class AB? It seems like a win, good regulation, low losses and high switching frequency that can be filtered.
Linear regulator produce massive amounts of heat and need a lot of work to get enough current out of them.
Does anyone have any experiences with step down switching regulators in class AB? It seems like a win, good regulation, low losses and high switching frequency that can be filtered.
Linear regulators produce some heat from the voltage drop but they do not need much work to get even 10Amp steady-state out of them (with good dynamic reserves). And, they are easy to design and use standard components. They are more straight forward when adjusting the regulation loop. They basically use the same principle as the class AB amplifier, just with a fixed reference.
Much is called "step down" when you talk about SMPS (Buck, primary switchers, many fly-back, many forward etc.). I have experience with primary switchers both for class D and AB. Their main advantages are the reduced (heat) losses and pricing because they are manufactured in huge amounts for general purposes.
The disadvantages of SMPS for audio is the inherent switching noise and eventually a slow regulation loop you can do little about. The switching noise bothers persons with class D amplifiers less than persons with class AB amplifiers. Class D inherently includes switching noise to be handled by the output filter. Class AB is designed to handle very delicate music signals with very little noise and distortion, and then we want to combine it with a circuit that generates ample noise! It can be done but requires more work to optimize the concept than the linear regulator. The nasty square-wave signals require good filters to bring down the noise and the radiated noise is important.
My personal opinion:
For audio purists - use linear regulators because of the very little noise and often better dynamic properties.
For general audio consumer products - use SMPS because the cost-price will be very important.
For less experienced DIYs - use SMPS because you quickly find one ready-made that should work reasonably well and it may teach you something about the more problematic characteristics of SMPS.
Much is called "step down" when you talk about SMPS (Buck, primary switchers, many fly-back, many forward etc.). I have experience with primary switchers both for class D and AB. Their main advantages are the reduced (heat) losses and pricing because they are manufactured in huge amounts for general purposes.
The disadvantages of SMPS for audio is the inherent switching noise and eventually a slow regulation loop you can do little about. The switching noise bothers persons with class D amplifiers less than persons with class AB amplifiers. Class D inherently includes switching noise to be handled by the output filter. Class AB is designed to handle very delicate music signals with very little noise and distortion, and then we want to combine it with a circuit that generates ample noise! It can be done but requires more work to optimize the concept than the linear regulator. The nasty square-wave signals require good filters to bring down the noise and the radiated noise is important.
My personal opinion:
For audio purists - use linear regulators because of the very little noise and often better dynamic properties.
For general audio consumer products - use SMPS because the cost-price will be very important.
For less experienced DIYs - use SMPS because you quickly find one ready-made that should work reasonably well and it may teach you something about the more problematic characteristics of SMPS.
Last edited:
Do you have any regulator recommendations? Maybe LM338
Figure 30.
Amp is LM3886 based, I need something capable to sustain peak currents.
Well, from the comments on this thread:
Advantage of a regulated power supply
power supply regulation for a power amplifier isn't needed, and may actually limit performance.
Advantage of a regulated power supply
power supply regulation for a power amplifier isn't needed, and may actually limit performance.
I prefer designing my own unless for pre-amplifier circuits or auxilliary circuits with a rather constant and moderate loading. Then, integrated regulators are fine.
For the dynamic loading from a class AB or D power amplifier I prefer an emitter output coupling because the emitter will supply a lot more current with only a very moderate further voltage drop (to the base voltage) when the current is suddenly needed. This is done substantially without reaction of a regulation loop if the base voltage is just held constant. You may then add a further regulation loop to more slowly compensate for the smaller output voltage deviations.
For me, a good audio power-amplifier power-supply is characterized by being "stiff" in voltage with any dynamic loading. With a "stiff" supply voltage the amplifier can rely on always having a stable voltage and the amplifier regulation loop(s) can be designed without bothering about the reaction of the power supply.
For the dynamic loading from a class AB or D power amplifier I prefer an emitter output coupling because the emitter will supply a lot more current with only a very moderate further voltage drop (to the base voltage) when the current is suddenly needed. This is done substantially without reaction of a regulation loop if the base voltage is just held constant. You may then add a further regulation loop to more slowly compensate for the smaller output voltage deviations.
For me, a good audio power-amplifier power-supply is characterized by being "stiff" in voltage with any dynamic loading. With a "stiff" supply voltage the amplifier can rely on always having a stable voltage and the amplifier regulation loop(s) can be designed without bothering about the reaction of the power supply.
After some thinking I have decided to omit the PSU regulation.
This PSU would be for class AB
. And at this point I only care about the ripple.
There are few options:
What are you thought on this?
This PSU would be for class AB
. And at this point I only care about the ripple.There are few options:
RC filter. You get a voltage drop and PSU resistance inceaed by R, also with class AB, voltage will sag with the current pulses.
LC filter, better than RC but big inductors are pricy. To provide attenuation at mains frequency, you need a BIG inductor that will also exhibit big R as well(back to RC filter). There is also potential for ringing with LC.
Cap multipliers. Cool but not for a ugly class AB load.
Stacking up the capacitors or using good SMPS.
What are you thought on this?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.