High VOLTAGE Switching input Tantalum Capacitor

[jimmykk]

HI

I am going to build a High voltage Switching Power supply using IC MAX1771. I have a doubt regarding the decoupling tantalum capacitor which is placed at the point where 12V is coming in to the board before it goes to inductor. The recommended value is 100uF,35v tantalum but what if i use a 100uf, 16v TANTALUM CAPACITOR in its place as it is greater than the incoming voltage of 12v.

 

[Vbase]

Using 16V cap is too close to the limit. You can use 4 of 22uF 25V in parallel, they are common and easy to get.

 

[jimmykk]

thanks for reply....or a single 100uf 25V will do the job...right

 

[Dan Mills]

Depends, tants running at more then about 50% or their ratings are bad news from a reliability perspective when used across a stiff supply.
That 12V, is it 12V, or is it automotive (So 14.4V most of the time in reality with plenty of nasty transients and near zero source impedance)?

Personally unless there was some pressing reason it had to be a tant (Potted in a high pressure environment or something) I would be reaching for an electrolytic from a good manufacturer with some suitable MLCCs in parallel.

Regards, Dan.

 

[dick_freebird]

Input filter ESR plays into wallplug efficiency and you
may prefer multiple parallel lower value caps for this.
In a simple buck your output current runs into the
cap for the output-low period and out of it for the
high phase, always scraping across the resistance
one way or the other for a loss term. You also want
to look at ripple current ratings which can also drive
you to multiple parallel caps in higher current buck
converters (your ripple current is your output current,
in these input caps).

 

[schmitt trigger]

From personal, very painful experience, I can tell you that I fully agree with Dan_M.

For reliability, Ta capacitors must be used at 50% or lower of their rated voltage, so in an automotive environment you MUST USE a 35V device.

 

[jimmykk]

what if the 12 V i am providing is from a power supply for lab experiments.

 

[schmitt trigger]

Then go for 25 volts