Capacitors with a MIC2940A-5.0 voltage regulator?

[SteveT]

A quick question with reference to the attached schematic.

The circuit uses a MIC2940A-5.0 fixed 5v LDO voltage regulator. The datasheet recomends an output capacitor of at least 10uF and it must be low ESR. OK, no problem with that I just bought some nice 15uF Sanyo Aluminum solid capacitors, however, whenever I use a 'normal' linear regulator such as LM7805 I always add an electrolytic (and) a 100nF ceramic to catch the faster nasties. Would I be ok to do the same with this reg?

N.B. This question was prompted by Dippy's excellent thesis on decoupling capacitors. See, some of us read it
View attachment 5v LDO with Crowbar.pdf View attachment mic2940.pdf

 

[Paix]

Hi SteveT,
It seems that the answer to your question isn't coming very fast. Everyone must be busy elsewhere at the moment.

I believe that the reason why you normally use an electolytic and a 100nF ceramic, is that the Electrolytic isn't low ESR. it's performance at high frequency is impaired by the method of construction; more often than not it is rolled and so has a small amount of inductance and the nature of the electrolytic process introduces a small resistive component too. With a low ESR capacitor the high frequency performance is much better, obviating the belt and braces approach which was standard for so long due to the lack of affordable high quality high value capacitors.

I'm far from being one of the regular experts on the subject, so i could be wrong, but until the cavalry arrives I think my best shot is fairly to the point, but not particularly erudite.

 

[fernando_g]

An LDO internal circuit is significantly different to a plain vainilla 7805.

They require certain amount of ESR which forms a zero to maintain stability.

Having said that....usually solid aluminum caps have very decent high frequency performance and an additional ceramic may not be required and could even be detrimetal to stability.

However, always follow the datasheet recommendations.

 

[premelec]

Just to remind you that LDOs are great and have a strange characteristic as they go into very low drop voltage - their idle current goes way up [see data sheet]

 

[SteveT]

Thanks to Paix fernando_g and premelec for their responses.

premelec, I've noticed that, I think I might design the circuit board to have the polyswitch adjacent to the heatsink. I know it's not a thermal switch as such but could possibly be made to work that way.

I forgot to mention that the 6v input will be from a wallwart so the chances of it going into dropout will be minimal.

Thanks again everyone.