Thrifty regulator modification as a capacitor multiplier

[neazoi]

Hi, this is a more powerful version of the well known thrifty regulator.
I was wondering if I could add a base capacitor (shown in red circle) to make it also a capacitor multiplier (https://www.radio-electronics.com/info/circuits/transistor/capacitance-multiplier-circuit.php), so that some ripple can be filtered out?
Can I do that?
Also, is the capacitor polarity ok?

 

[schmitt trigger]

Yes, the capacitor in that position, will act as a capacitance multiplier. The polarity is also correct.
A word of caution, since this will also cause a lag network, you may want to check stability and settling time with/without the capacitor.

Why is it called a "thrifty" regulator?? It is linear, it will waste power.

 

[neazoi]

A word of caution, since this will also cause a lag network, you may want to check stability and settling time with/without the capacitor.

Why is it called a "thrifty" regulator?? It is linear, it will waste power.

I was called so on an elektor article, maybe he refers to the cheap meaning.

This lag could be used in benefit in some cases, if for example a slow start up is required at the load?

 

[schmitt trigger]

Indeed the lag *could* be beneficial in certain aplications.

But since you are modifying the original concept, you must check it to ensure you have not introduced an anomalous behavior.

 

[neazoi]

Indeed the lag *could* be beneficial in certain aplications.

But since you are modifying the original concept, you must check it to ensure you have not introduced an anomalous behavior.

You mean to check if the addition of the capacitor affects the feedback loop, so that it's effect is degrated?

- - - Updated - - -

You mean to check if the addition of the capacitor affects the feedback loop, so that it's effect is degrated?

Hm... you are right. Untill the capacitor get's charged the output voltage is higher. I have tested it. This might be a problem indeed if the loag cannot tollerate a higher voltage.

 

[Orson Cart]

There really is no advantage to adding the 100uF to the circuit in that position, if you want a soft start add a cap (47uF say) and a resistor (1k) across the zener....

 

[neazoi]

There really is no advantage to adding the 100uF to the circuit in that position, if you want a soft start add a cap (47uF say) and a resistor (1k) across the zener....

Parallel to the zener, right... Does it matter which way the resistor should be? I am thinking of trying it towards the Vout end.

 

[neazoi]

Parallel to the zener, right... Does it matter which way the resistor should be? I am thinking of trying it towards the Vout end.

This worked like a charm!
Thank you

Any ideas how a slow shut down could be implemented with it as well? Maybe a capacitor from the base of the 2222 to the gnd?

 

[schmitt trigger]

The slow shutdown will be provided by the reservoir in the output capacitor.....the 4.7uf is a quite small cap, you may increase it to 100 or perhaps 220 uF. It also helps in reducing the output impedance.

 

[neazoi]

The slow shutdown will be provided by the reservoir in the output capacitor.....the 4.7uf is a quite small cap, you may increase it to 100 or perhaps 220 uF. It also helps in reducing the output impedance.

Thanks!
I have just noticed that the circuit proposed in #8 works, but only if the cap is fully discharged. Maybe I will use a dpdt switch to connect a 5R resistor at parallel to the cap when the circuit is off. This will discharge the cap rapidly and the cycle will begin again dusing start up.

Any other ideas are welcome.

 

[FvM]

Any circuit modification that adds a pole to the feedback loop, e.g. capacitor from 2N2222 base to ground will not only bring up a risk of instability but also affect the regulation quality. You don't want this.

 

[neazoi]

Any circuit modification that adds a pole to the feedback loop, e.g. capacitor from 2N2222 base to ground will not only bring up a risk of instability but also affect the regulation quality. You don't want this.

Thanks, I finally did the soft shut down with a 1F supercap at the output paralleled with a 120R. It works satisfactorily.