How to protect against Voltage Dip Test

[ku637]

Hi,

I'm testing a prototype Board having a 24V to 5V & 3.3V converter . Microcontroller circuits works the 5V output .To make the board fool proof against voltage dip tests ( typically 10ms on 24V line)

Shall i give more priority to 24V input bulk capacitance or to the 5V output side of buck regulator capacitance..

Thanks,

 

[barry]

Well, since you're testing it, why don't you see what happens with different caps in different places? It depends on what your load is, what your regulator looks like, etc.

 

[emresel]

Hi;
I just think that a cap on the 24V side, and consider 24V source is off for a moment, probably voltage accross bulk cap will flow not only to your load side but also the switched off generator side. So I think it is better to apply 5V side.
And since voltage is lower, you are able to use bigger bulk caps in small physical sizes (comparing to 24V)

 

[chuckey]

In principle the converter supplying the 5V only needs 6V for it to continue to work. So if you use a cap on the 24V line it can discharge to 6V, before the 5V fails. If you use a cap on the 5V line only, a failure is normally reckoned as 4.75V, so this is the minimum the cap can discharge to. It should have a diode in series with the +24V feed, so the storage cap is not back powering the rest of any other equipment connected to the +24V line.
Frank

 

[emresel]

Hi;
Diode isolation might be a solution but need to check current on 24V line, if it is high, diode might heat up. Consider schottky diodes for this purpose for reasobale current ratings.

 

[ku637]

In principle the converter supplying the 5V only needs 6V for it to continue to work. So if you use a cap on the 24V line it can discharge to 6V, before the 5V fails. If you use a cap on the 5V line only, a failure is normally reckoned as 4.75V, so this is the minimum the cap can discharge to. It should have a diode in series with the +24V feed, so the storage cap is not back powering the rest of any other equipment connected to the +24V line.
Frank

Exactly this is the conclusion i ve reached after testing..adding cap at 5V require large values to hold the voltage steady (i.e say 4.75V as you noted) even though the voltage rating is less..
but at 24V input its lesser cap value required since it can tolerate to a lower discharge levels..though voltage rating should be higher..

Thanks