Any advantages/pitfalls with CS-CAP capacitors for hobby electronics?

[gildasd]

CS-CAP or Conductive Polymer Aluminum Solid Capacitor,

For normal hobbyist stuff (not radio or high end sound), is there any advantage?
Do we need to modify any calculation over normal electrolytic capacitors?

[mrpackethead]

Polymer type capacitors have several advantages over a normal Aluminum Electrolytic Capacitor. The polymer electrolyte has higher conductivity than the liquid electrolyte used in standard capacitors, resulting in lower ESR. This permits a single polymer type capacitor to replace several aluminum electrolytic type capacitors, resulting in reduced board space and total number of components on the board. Since the polymer is a solid, it also gives longer life, not following the classic Arrhenius formula. Instead of doubling life with every 10°C drop in temperature, the lifetime will be 10 times longer for every 20°C decrease in temperature.

Yes, they are a little more expensive, but for circuits where having low ESR is importnat ( Power Supplies ), they can make a big difference!

[gildasd]

Polymer type capacitors have several advantages over a normal Aluminum Electrolytic Capacitor. The polymer electrolyte has higher conductivity than the liquid electrolyte used in standard capacitors, resulting in lower ESR. This permits a single polymer type capacitor to replace several aluminum electrolytic type capacitors, resulting in reduced board space and total number of components on the board. Since the polymer is a solid, it also gives longer life, not following the classic Arrhenius formula. Instead of doubling life with every 10°C drop in temperature, the lifetime will be 10 times longer for every 20°C decrease in temperature.

Yes, they are a little more expensive, but for circuits where having low ESR is important ( Power Supplies ), they can make a big difference!

So you read Digi Key too 
http://www.digikey.com/us/en/ph/nichicon/AlumCondPolymer.html
I just happened to get some instead of electrolytics and wondered if they were any major pitfalls.
Will save them for power supply filtering and as pull caps on the ICL7660 negative voltage generator.
Would they be good for using as pull capacitors on a LM7660?

[mrpackethead]

Lower ESR == worst control loop stability. Keep that in mind.

Yes, but longer lifetime, and less ripple. ( potentially )..   Everything in engineering is a compromise.

[gildasd]

Lower ESR == worst control loop stability. Keep that in mind.

Yes, but longer lifetime, and less ripple. ( potentially )..   Everything in engineering is a compromise.

For those without solid math and control theory background, I always suggest to go with generic parts, not those leading edge parts.
To make a converter stable at zero load and zero ESR requires tricks in control loop and heavy calculation and simulation.
Even doing so, some other parts of the design may be compromised, such as load transient response.

On a more amateur level, these are some example uses:

1) On a generic 12V 1Amp wall wart supply,
a standard 1000µF electrolytic cap before a 7809.
Then a 470µF electrolytic after after it.
Taking the edge (or spike) off the ripple so that it is usable to feed op amps and the such without leakage.

2) As 10 µF pull caps on ICL7660.

3) As simple RC filter (think 100 Ohms, 47µF) before supplying DC parts in the 1V to 9V range.

And other boring basic stuff.

[gildasd]

On a more amateur level, these are some example uses:

1) On a generic 12V 1Amp wall wart supply,
a standard 1000µF electrolytic cap before a 7809.
Then a 470µF electrolytic after after it.
Taking the edge (or spike) off the ripple so that it is usable to feed op amps and the such without leakage.

2) As 10 µF pull caps on ICL7660.

3) As simple RC filter (think 100 Ohms, 47µF) before supplying DC parts in the 1V to 9V range.

And other boring basic stuff.

2 and 3 are okay, 1 should be thought twice. Inside a wall wart it is fine, but outside a wall wart, inside a final product, this may not be true. Low ESR will not effective damp inductance of power cable induced transient, hence can break down linear regulator. LT has an app note describing this phenomenon.

Thanks for the answer.
For 1, my methodology is to get a rough estimate with T=R*C
Then to prod around that value with a varying caps while watching the effect on an oscilloscope.
I start low (lets say 470µF in this case), and go up: 1000µF, 1500µF etc.
A certain point the gain in "cleanness" is not noticeable compared to the value before, I stop at this value.

What you are saying is that I should add a load test AFTER to see if my values are stable?

I feel like I'm explaining how to sharpen sticks to MIT staff... This is very crude...

[mrpackethead]

For those without solid math and control theory background, I always suggest to go with generic parts, not those leading edge parts.
To make a converter stable at zero load and zero ESR requires tricks in control loop and heavy calculation and simulation.
Even doing so, some other parts of the design may be compromised, such as load transient response.

in deed. IN fact i was working on a design just the other day, using a Micrcel Buck converter, where additional ripple is actually designed into the compensation network.