> In this case, the load of the 12V output is higher than the
> load on the 5V output.... are you suggesting that this is bad?
> The 5V o/p is still giving me 5V, and the 12V o/p should be
> "proportional" assuming the windings are in the right "phase".
> But this is where I was getting just a few volts out.

Referring to
http://www.narwani.org/neil/stuff/dc-dc-converter-schematic.jpg

The load on the isolated winding should ideally be smaller than the non
isolated UNLESS you can design it just right and this will quite probably
require special conditions.

BECAUSE

The non isolated (here 5 volt) output is supplied with current both when the
switch is on (via the switch) and when the switch is off (via inductor
flyback).

BUT the isolated winding is supplied with energy ONLY during one or other
half of the cycle. With the dotting shown this will be when the switch is
on. You have a forward converter. If the input voltage is close to the
non-isolated output voltage then the switch will be on for most of the cycle
and the isolated output will see large duty cycle drive pulses. It acts as a
forward converter so the energy taken is not "stored" in the core but still
contributes to core magnetisation and possible saturation on the forward
stroke. Now, as Vin rises the switch turns off for greater portions of the
cycle and the isolated voltage sees a decreasing average voltage.

Approximately Voi = Von/Vin x N / Z

    Von = non isolated output voltage
    Voi = isolated output voltage
    Z = efficiency
    Vin is input voltage.

If Voi = 12 volts when Vin is slightly above Von,  then Voi will drop to
about 6 volts when Vin = 2 x Von !!!!!!!!!!!

In any case you will have a pulsating output driving the Voi output. The
"proper" way to do this is to drive a series inductor with its own flyback
diode - effectively a second isolated buck converter.

To design this properly you MUST know the range of Vin and design so Voi is
always at least as great as the minimum desired value. If Vin varies by 2:1
say then Voi will also vary by this amount, whereas Von will remain stable
!!!

This means you need a post regulator which tends to negate what you are
trying to do with the buck converter.

Making Von the main energy output does not cure the above problem but means
that you are post regulating the smaller of the two outputs.

One great advantage of the Voi system is that it can be higher than Vin
while Von is lower.
eg Vin = 12, Von = 5, Voi = 15 say.

I have run the above system with good success but with Voi < Von and with a
Voi linear post regulator.


________________________


If you reverse the dotting the isolated winding will draw its energy solely
from the flyback cycle and energy for it must be stored in the core. As Vin
decreases the off time decreases and available energy for the isolated
winding decreases PLUS it "robs" flyback energy from Von. The system
responds by providing more forward energy (switch on longer) as Von is now
not getting the flyback portion it was expecting and AFAICS this drives it
towards starving Voi completely (haven't tried this in practice).


            Russell McMahon

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu