On Tue, 16 Jul 2002, Brendan Moran wrote:

>The first is that having fixed frequency (yes, I know, there are TONS
>of harmonics) means that you know what frequencies of radiation
>you're dealing with, which makes various shielding and filtering
>issues easier.

No, the harmonics depend a lot on the regime of the psu.

>The second is that haing a fixed frequency means havign a fixed power
>transfer ratio through the transformer, which is important,
>considering that higher frequencies have transformer efficiencies.

No, transformers have a peak efficiency at a given (by the manufacturer)
frequency.

>Without fixed frequency, I expect that you would see efficiency that
>was worse acorss the operating range, since the power transfer
>through the transformer would not be the same across the operating
>range.

No, the changing efficiency can be used to improve regulation with the
same circuit.

>The third reason is that fixed frequency, variable pulse width
>provides much better regulation by varying the on time.  If there is
>a small load on the output of a switcher that is capable of high
>loads, but uses fixed pulse, the Vr will be terrible in comparison to
>a PWM, since the pwm just turns down the on-time to match the load.

Not if you use flyback. With flyback you lose control over the output for
at least T where T is the period of your switching.

>The fourth reason for me is that PWM is an ideal for switching power
>supplies.  Not necessarily the best for all solutions due to cost
>etc.  But certainly the target to shoot for when possible.

Not if there is a transformer involved. The DC component can saturate the
transformer easier with PWM than with other types of regulation. PWM +
transformer = suck.

>I think that was all coherent, but I'm not too sure.

;-)

Peter

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics