switcher duty cycle - high or low?

[zappedagain]

I'm a bit rusty as it has been about 10 years since I last looked at switcher designs.

I need a 3.6V (+/-10%) 2.5A. I can tap it off a 5V supply (~75% DC) or a 24V supply (~15%). Which would you recommend?

I've read recently in a Linear Tech App note that subharmonics are more of an issue when the duty cycle is over 50%, so that makes me lean toward the 24V solution. Then I suspect I may get into discontinuous operation and have hurdles there too.

Z

 

[rwstowe]

The LT app note is probably referring to subharmonic oscillation present in power supplies with current mode PWM controllers and with duty greater than 50%. Voltage mode controllers do not have this problem. However, current mode controllers have a single pole response in the line and control to output transfer functions which is easier to compensate for with the compensation network. This is as opposed to a double pole transfer function which is present in the voltage mode controller. Current mode controllers can be operated above 50% duty by applying slope compensation to the the current signal.

The 24 volt would probably be the simpler implementation since you could benefit from the single pole response of the current mode controller operated below 50% duty. An additional advantage is that the input current would be much less and therefore much smaller IR drop on the input wires. The only disadvantage I see is the control resolution is less at a low duty cycle.

For tutorial on power supplies please see:

http://www.truepowerresearch.com/2011/02/power-supply-tutorial/

Bob Stowe
President,
True Power Research, LLC

http://www.TruePowerResearch.com

 

[zappedagain]

Thanks Bob! That is some interesting detail. I ended up going with a 24V input solution. I'll have the hardware back in about a week so I can test it out.

Z