SMPS for Class D amplifier.....slow to come out of no load.

[Flyback]

Hello,
As you know, such an SMPS, (for class D supply) must be able to go from no load to full load very quickly.

(This smps is isolated and uses an optocoupler in the feedback loop.)

When you are in no load, how do you stop the opto transistor from saturating?.....as its this saturation which makes the amplifier slow when coming out of no load, and causes overshoot.

 

[ronsimpson]

A schematic helps!!!
If the optocoupler is saturation then the design has problems. (possibly) Where did you get the SMPS?
Many SMPSs do not work well with no load.

 

[Flyback]

schem
https://i47.tinypic.com/inxu1v.jpg

I cannot see how you can stop the opto transistor from saturating in no load?

Many SMPSs do not work well with no load.

..these days, all smps's have to work on no load, due to standby regulations

 

[Nigel Goodwin]

schem
https://i47.tinypic.com/inxu1v.jpg

I cannot see how you can stop the opto transistor from saturating in no load?

With a class-D amp, why would there ever be no load?.

..these days, all smps's have to work on no load, due to standby regulations

Normally they have a tiny separate SMPSU which runs during standby, with the main one switched off.

 

[Flyback]

With a class-D amp, why would there ever be no load?.

when the guitar plugged into the amp was not being strummed

 

[Nigel Goodwin]

when the guitar plugged into the amp was not being strummed

Why would you think there would be no load at that?, the amplifier is still running, and still presenting a load to the PSU.

 

[ronsimpson]

If the audio amp is any good it should have 'supply rejection'. Where supply variations do not show in the output.

 

[Flyback]

Why would you think there would be no load at that?, the amplifier is still running, and still presenting a load to the PSU.

yes but in a class ab its not much load

 

[ronsimpson]

So your opto saturates and you don't like it. I noticed that the base of Q1 is 1.5 volts bellow 5V reference. So the opto should have about 0.9 volts across C to E. If you have 0.5 volts (or less) cross it then Q1-C must be above 2.5V that U1 is looking for. Why? Probably when the supply went form full load to no load the error amplifier is too slow and did not get the duty cycle reduced in time and the voltage over shot. (got too high). This causes the error amplifier to overcompensate and go to zero duty cycle for a while.

Does your supply "ring" when going from no load to full load and to no load?

 

[Nigel Goodwin]

yes but in a class ab its not much load

But it's plenty to make the SMPSU run correctly.

Assuming this is all about you designing a SMPSU for a class-D amplifier?, then you're not making a very good job of it!

 

[ronsimpson]

1) I never did find out what the power supply is doing wrong. (other than saturation)
2) Please post your LTSpice files. (Include everything on IC1,2)

 

[Flyback]

here is ltspice sim

Please tell if there are downsides to this cascoding...because i see no downside....other than a bit of extra cost...howcome nobody does this cascode thing?...optos are a nightmare, and cascoding takes away that nitemare

stability is as easy as when getting stability for a non isolated supply

Assuming this is all about you designing a SMPSU for a class-D amplifier?, then you're not making a very good job of it!

...I dont think anybody is making a good job of it.....you can tell, because every <600W guitar amplifier on the market has a 50Hz mains transformer for a power supply.....and none of these <600W guitar amplifiers have a PFC stage (as you can tell from the fact they use 50Hz trafo's)....i have written confirmation from the biggest guitar amplifier maker in the world that none of their guitar amplifiers have power factor correction.

 

[schmitt trigger]

I don't see any issues with cascoding..........I actually think it is a great idea for an optocoupler......keeping its Vce constant should prevent some of the issues associated wit an optocoupler in the feedback path.

As to how come guitar amplifier manufacturers don't use PFC? I sincerely don't know, but I suspect that they are traditionalists with respect to circuit design.

 

[ronsimpson]

In volume production every part is money.
An error amplifier is a TL431. The opto and a resistor or two gets isolation.

The PWM has the job of keeping 2.5 volts across the isolator. So it sees a constant Vce. (more or less)

 

[Nigel Goodwin]

...I dont think anybody is making a good job of it.....you can tell, because every <600W guitar amplifier on the market has a 50Hz mains transformer for a power supply.....and none of these <600W guitar amplifiers have a PFC stage (as you can tell from the fact they use 50Hz trafo's)....i have written confirmation from the biggest guitar amplifier maker in the world that none of their guitar amplifiers have power factor correction.

There are plenty of perfectly working SMPSU's in guitar and PA amps (my PA mixer/amp has one, and it's less than 600W).

Why are you obsessed by PFC?, what does it matter to you?, and why would you imagine guitar amps would want it?.

 

[ronsimpson]

Without PFC current is only pulled from the power line at very peak of the sign wave. In my house the sine wave is very flat on top because of all the electronics. PFC lowers the peak current greatly.

The noise on the power line, in the audio range, drops when current is pulled in a sign wave.
Without PFC the wires in walls heat more.
The power company likes 'resistive' loads. PFC looks like a resistive load.
Many countries there are PFC requirements on large loads.
Breakers and fuses like sing wave current not sharp spikes of current.

 

[Flyback]

Without PFC the wires in walls heat more.

this is true, but for a non-PFC, 20W offline SMPS, can you say how much extra wattage is lost as "heating in wires in walls" ?

Why are you obsessed by PFC?, what does it matter to you?, and why would you imagine guitar amps would want it?.

....whomsoever knows how to get away with non-PFC'd SMPS's for guitar amplifiers, will make a tidy sum. Mains transformer PSU's dont need PFC, and so thats how guitar amp manufacturers get away with not using it at the moment.

 

[Nigel Goodwin]

....whomsoever knows how to get away with non-PFC'd SMPS's for guitar amplifiers, will make a tidy sum. Mains transformer PSU's dont need PFC, and so thats how guitar amp manufacturers get away with not using it at the moment.

What's to know? - as far as I'm aware only specific items are required to have PFC (TV's probably being the most common).