Voltage Regulator?

[bluefoxicy]

I'm trying to understand figures 6 and 7 here (referenced as 3 and 4 in the documentation on page 9):

http://www.st.com/stonline/books/pdf/docs/2154.pdf

I don't understand the voltage reference. Near as I can tell, the idea here is that the LM317 (I'm using LM217 though...) passes enough voltage to keep the Control Pin at -1.25V from Vout, thus the ground resistor directs some voltage to ground and reduces this reference voltage, and the LMx17 passes more voltage so more voltage goes across the resistor and the Control Pin again reflects 1.25V?

So in effect, Vout will pass +25V if Iadj sees +23.75V, regardless of Vin?

[defaced]

So in effect, Vout will pass +25V if Iadj sees +23.75V, regardless of Vin?

No, a regulator can only knock voltage down. This is located at the top of Table 4, the first row is delta Vo, which states that Vi must be 3 to 40v more than Vo. 3v is a pretty typical minimum loss for regulators, meaning, when you use a regulator your voltage input has to be at least 3 volts higher than your voltage output otherwise the regulator can't do it's job properly.

[bluefoxicy]

No, a regulator can only knock voltage down. This is located at the top of Table 4, the first row is delta Vo, which states that Vi must be 3 to 40v more than Vo. 3v is a pretty typical minimum loss for regulators, meaning, when you use a regulator your voltage input has to be at least 3 volts higher than your voltage output otherwise the regulator can't do it's job properly.

Okay but if I slap 15V higher than I want with an erratic 10V wobble on Vin, it should still come out smoothly on the voltage I expect, maybe with a very little wobble, right?

[defaced]

I don't think a regulator works like that. A regulator keeps constant voltage regardless of the current draw from the regulator. What it sounds like you need is more capacitance before the regulator to filter out the voltage ripple. Do you have a schematic of what you're working on?

[bluefoxicy]

I don't think a regulator works like that. A regulator keeps constant voltage regardless of the current draw from the regulator. What it sounds like you need is more capacitance before the regulator to filter out the voltage ripple. Do you have a schematic of what you're working on?

Ah, okay.

http://www.st.com/stonline/books/pdf/docs/2154.pdf

Page 9 Figure 7 is where I'm looking right now. Pretty much that with a 28VCT transformer, into 4 x 1N4004 rectification stage and a pair of parallel 220uF+/-10% between B+ and Ground off those. That's where Vin comes from.

I don't have the diagram drawn yet.. heh.

[defaced]

With 440uf of capacitance in a no-load condition, I wouldn't expect much voltage ripple. It sounds like something isn't hooked up right between the transformer and the caps. When you get a diagram sketched up, just pen and ink will work and then take a picture of it, we'll see where you're at.