[SOLVED] Feedback Mechanism in Linear Voltage Regulator

[darkseid]

Dear All

In a Linear Voltage Regulator, What happens when the error voltage is zero since the op amp output will be zero and will not be able to drive the series element. Please indicate if you require a circuit diagram.

regards
Darkseid

 

[FvM]

Assuming an ideal I-acting feedback controller, there's an output voltage despite zero error voltage. A real feedback control has finite gain and thus exposes a small error voltage. The load regulation specification of a volateg regulator will tell you about this deviation from ideal behaviour.

 

[darkseid]

So Will it be a good approach to use a potentiometer to adjust the feedback voltage till we get the required no load output and then replace the pot with fixed resistance voltage divider.

 

[enjunear]

It sounds like your linear regulator has an Adjust pin on it. Typically, the datasheet will tell you the desired voltage on that pin. Commonly, I see that at 1.25V. By doing a little bit of math (see resistive voltage divider), you should be able to determine a reasonable set of resistor values to try.

Example:
Desired Vout = 7V, ADJ voltage = 1.25V
Assume we can waste 10 mA of current in the resistive divider (swamps out the bias current flowing into the ADJ pin).

If I = 10 mA and Vadj = 1.25V, then R2 = 1.25 / 0.010 = 125 ohms. Then R1 = (7-1.25) / 0.010 = 575 ohms
For standard resistor values, you can find a 124 ohm 1% resistor. Then recalculate the current (1.25V across a 124 ohm resistor = 10.08 mA), recalculate the R1 resistor, then find a few standard value resistors that would add up to the value you need. It'll take a little fiddling to get spot-on, but you can always test-select the last resistor to get the voltage dead on (use two resistors in series for R1.... one around 550 ohms, then test-select the smaller value, which would be around 10-15 ohms).

 

[darkseid]

Hi enjunear

I am trying to build one using discrete elements i.e. series pass element and error amplifier (for feedback).

 

[FvM]

If the error amplifier has sufficient gain, the error voltage should be low (below mV), so it's no reason to implement an adjustable output voltage. But it depends on your circuit. Did you notice, that error amplifier gain and frequency characteristic raises the question of loop stability and possible need of loop frequency compensation?