Variable SMPS + Variable Linear PS!

[Shahriar]

Hi

I want to add a linear PS to a SMPS to improve regulation and vary the output voltage between 0-30V. The Vi in the schematic comes from SMPS .In worst condition it has 250mV ripple.
To reduce power dissipation in trasistor I plan to change Vi when I want to
change Vo, But I always have to keep (Vi), 3.5V higher than (Vo) to keep LM317 working proplerly.

{the trasistor turns on when LM317 darws 20mA. the input voltage of LM317 must be 2.5V higher than its output. If Veb=1V then (Vi-Vo = 2.5V + 1V =3.5V) }

when I draw 10A, My SMPS should produce 35W more (10A * 3.5V) and dissipate this power in Linear PS.

I want to know how can I do a trick which enables me to reduce this Differential Voltage(3.5V) Lower. i.e. down to 1V
which components can I use instead pf LM317.

[Guest Alun]

Don't use this circuit, use a low dropout regulator like these:





Can't get hold of the TL431? Don't worry you could build my op-amp version. You might want to use a faster op-amp like the MC34071 I only used the MC33171 because I've got loads of them lying around, just be aware that it uses slightly more power so you'll might have to swap the 10nf capacitor on the output of the 555 for a 100nf. Off course you don't have to use the 555 voltage doubler if you don't want to it was just an idea you could use any voltage that's sufficently greater than the output to bias the MOSFET fully on.

In this case to ensure a good clean output the switcher's output needs to be >250mV + V_DROPOUT at the worst case.
V_DROPOUT = Maximum load current * MOSFET on resistance

These threads might be helpful:
http://www.electronics-lab.com/forum/index.php?topic=3843.0
http://www.electronics-lab.com/forum/index.php?topic=4630.0

[Shahriar]

Dear Alun
Thanx Very much for your help and also thanx to audioguru. ( I read those links), that's great Help.
In the following circuits, Can you Write some  Equations Between Vgate, Vdrain, Vsourse.
to determine How much Vi can change?

Specialy in P-chanel Mosfet

ThanX
shahriar

[Guest Alun]

shahriar,
V_IN ca change as much as it wants providing it doesn't exceed the componant ratings for the current being drawn, however if you want V_OUT to remain constant you V_IN needs to be greater than the dropout voltage and I've already posted a formula for this. The problem with using a p-channel MOSFET (like in my first idea or audioguru's modification to my circuit) is the on resistance is higher the the dropout voltage will be higher, also it's no good for output voltages less than the threashold of the MOSET. The n-channel design might be a bit of a pain because it requires a separate power supply higher than V_IN but it doesn't suffer from any of the afformtntioned problems, by the way the higher input voltage needs to be higher than V_OUT by a factor of the gate threashold voltage.