Offline 400W SMPS for Class D amplifier supply with ON/OFF control?

[treez]

Hello,
Can I do an offline (230VAC) SMPS for Class D amplifier supply (for guitar) with a simple "burst mode" control type, Two Transistor Forward SMPS?

ie, Vout divided down and put into a comparator.....when vout goes below regulation, the comparator trips and sends the tripped signal through a digital isolator to the primary side controller, which simply turns on the gate driver with a fixed duty cycle......as soon as the vout goes below the regulation value, the comparator again trips and the primary side controller simply turns the switching FET(s) off.......and then repeat.

Whats wrong with this?

If its EMC , then I can have a "soft start" so that every time the switching FET is turned on for a "burst", the duty cycle is steadily increased, over say 10 switching cycles, to the maximum duty cycle.

Will this be OK?

I wouldn't have PFC simply because class D amplifiers , on average, never operate above 1/8th of their maximum power.

 

[WimRFP]

Such a scheme can work (from experience). You likely will add some hysteresis in the comparator to have some control over the burst frequency. My concern is in peak current limiting and inrush behavior of your converter. If your two-transistor converter behaves as a current source (seen from the rectified secondary side), the transfer function has a dominant first order behavior.

I used both series inductor (leakage inductance) and series capacitor in the secondary AC circuit to have some current limiing scheme. You can also use a scheme where frequency increases to llimit the charge current of the (large) secondary storage capacitor. Some things to consider: very low burst frequency (large storage capacitor) or above audio burst frequency (small storage capacitor, can have higher order effects that may affect loop stability, larger inrusy loss as they occur more frequent).

If you are familiar with a (spice) simulator, I would first simulate the thing as to make sure everything is finy under all occuring operating conditions. Inrush behavior occurs now may times/s instead of few times/day.

You can also make it based on a voltage type converter and secondary storage inductor, but I never build it as a large storage capacitor is cheaper then a large storage inductor. Also the no-load efficiency for capacitive storage is very good (compared to inductive storage).

 

[treez]

Thanks,
I would have primary peak current limiting in there.
I would also have a soft start from switch on

 

[treez]

Do you think that this kind of switch mode power supply would be more prone to giving out unwanted audible noise, such as buzz from the transformers/inductors?

I was hoping that such noise might only happen when the guitar was being played loud (ie heavy load on the power supply) , in which case it would be drowned out by the guitar playing?

 

[treez]

Hello,

This report (attached) sets out the reasoning for the selection of Burst Mode Control for a 320W Two Transistor Forward converter SMPS for supplying a Full Bridge, Class D guitar amplifier.
(Output of SMPS is 80V & 4A)
Do you agree that Burst Mode control for this power level & application actually makes sense, due to the "pulsey" nature of the load, and the load's very low average power?

Please also find attached the LTspice simulation of the 320W Burst Mode SMPS.

 

[WimRFP]

I didn't go into the detials of the circuit, but such converters can have high efficiency over a large power range.

The converter does operate (at nominal power and high efficiency) or it does not operate. The first time that I suggested a burst mode SMPS (solar powered application) was because of the high efficiency over a large power range. Nowadays we have lots of nice smps ICs available, so one may go for a mixed burst-continuous mode control where burst mode is only activated at a power level where the converter can't operate at high efficiency. I always arrived at larger capacitors (to reduce the ripple due to the on/off control).

Regarding audio noise, when the burst frequency is in the audio range, you can have audio production from magnetics. I had a neighbour playing electric guitar and I am sure he wouldn't notice it...