[sdiy] LDO regulators and ceramic caps

[Steve Lenham]

On 07/10/2015 13:07, Richie Burnett wrote:
> Hi all,
>
> Out of interest how are you guys and girls dealing with the issue of LDO
> regulator stability with ceramic capacitors?  The datasheets for most
> modern micro's and op-amps preach about the importance of low-ESR 0.1uF
> ceramic decoupling caps across the power pins of the device.  But the
> LDO regulator datasheets say that low-ESR ceramics across the output
> will make the regulator go unstable and are a no no.
>
> Just wondering how people are resolving this issue.  My board is quite
> small 2"x2" so it's unlikely that track lengths would add enough
> resistance to the ceramic capacitors to move them into the stable region
> for the LDO regulator.
>
> I've seen at least one commercial solution that had 100 ohm resistors in
> series with everything fed from the LDO output to isolate the decoupling
> caps from the regulator, but then that series resistance wrecks the
> regulation!
>
> -Richie,

Hi,

I'm sure you have seen it already, but others may be interested in this 
good article on the subject, which includes some of the theory:

http://www.eetimes.com/document.asp?doc_id=1225555

It implies a couple of solutions:

1. LDOs exist that are specced for output ESRs down to zero, for exactly 
this reason - could always simply use one of those!

2. It mentions that maximising the distance between the reg and the 
ceramic caps allows trace inductance to work in your favour. If your PCB 
is too small for this, perhaps deliberately introducing a _small_ 
inductance (either discrete or using "wiggly tracks" on the board) might 
do the trick without impacting on the regulation. Got to be better than 
resistors - yuk.

Hope this helps - please share your results.

Cheers,

Steve L.