[time-nuts] possible issue with LEA 6T eval board

[nuts]

 Hoping that the
> receiver had not perished as well I decided to replace the regulator.
> I could not identify the part on the board but replaced it with a
> MICREL MIC5205-3.3YM5 . Now all is well, at least from monitoring
> over a usb cable.  
> 
> Regards
>  
>

Just a FYI, those bipolar LDOs can get a bit unstable near dropout. One
thing you usually don't see on a Micrel datasheet is the efficiency as
you approach dropout. That is because the BJT is near saturation, the
beta has dropped, and thus the chip itself needs more current just to
regulate. Generally for that type of design, you add a bit of hidden
circuitry to "sat catch", i.e. keep the BJT out out saturation for
stability reasons, though it is itself a second feedback loop. It is
usually never shown on a functional diagram. However the pass device
beta is very low near saturation, which can stress the part. The part
does have a thermal shutdown, but with bipolar, the part can fail
pretty quickly.

Unless you need very low noise, a LDO with a Pfet pass device is more
reliable.

This patent explains one approach at sat catching:

http://www.google.ca/patents/US5410241

So those bipolar LODs are a lot easier to fry than you would think.