Calculating inrush current of decoupling caps

[All Ears]

Hi, Anyone know of a method of calculating what the start-up or inrush current would be for the added capacitance of my opamp decoupling caps ?

I just want to find out if my typical LM317/337 reg dual 24v rail supply can handle those currents (and hopefully with a safety margin) without going pop

I know I can add a series resistor (and I am thinking about 10 ohm to 39 ohm) to each of the +/- rails to slow the inrush...but I would still like to know how those inrush currents are calculated and what that means for PSU build and components

Cheers,

All Ears

 

[Walrus]

http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capchg.html

 

[joaquins]

You need to know the ESR that is the R used here, it also may contain the wiring impedance in order to limit the current but if they are small caps go just with ESR that will tell you a good start point.

JS

 

[All Ears]

thanks

cool tool, walrus - i will use that

i see the usual choice for electrolytic decoupling caps for 'audio' opamp rails is 10uF - 100uF .  Are there any performance benefits (eg headroom) of going above this oft quoted 100uF figure ?

 

[JohnRoberts]

The VR probably has current limiting so will self limit the in-rush current if above it's max current.

There is generally little harm from too much capacitance, but likewise little benefit from excessive capacitance. The stiffness of the supply will depend on the ground robustness and length of PCB traces.

In theory a capacitor connected to a DC rail will do nothing. if the DC voltage never changes.

JR

 

[joaquins]

In the other hand the bigger caps at input of PS need some time to get charged too, this time is attenuating the peak current at start up of the last ones.

JS

 

[JohnRoberts]

years ago I experimented with adding large electroilytic caps to the output of then lower performance 3 terminal regulators (think 7815/78915) to improve the HF source impedance. I found the falling impedance with increasing frequency of a thousand uF or so of capacitance cancelled the rising impedance with frequency of the modest performance regulators of the day. When well matched the source impedance was relatively low and relatively flat to above the audio band. 

Modern regulators are better. So I don't expect as much capacitance is needed. if any.

JR