What limits the use of LDO at higher frequencies?

[iVenky]

I am designing an LDO. The frequency of the input wave is about 26 Mhz. I read a classic paper by Mulliken where they have designed the LDO for a UGB of about 100 khz. I tried increasing the UGB but it was very difficult for me to increase both UGB and as well maintain the required phase margin.They have used an auxillary current path for sudden transient current. Now I thought the circuit wouldn't work for 1 Mhz or 10 Mhz but I was surprised when it worked for 10 Mhz.(simulated in Cadence Virtuoso) How could this circuit work? i.e. there was very less ripple in the output even at those input frequencies. This means that UGB doesn't matter but the auxillary path matters the most for it to work at higher frequencies. Anyway it didn't work for 26 Mhz or 15 Mhz. So I need your help in making it work at those frequencies. What should I do?

Thanks a lot.

 

[erikl]

Can't you afford a LPF after the rectifier? Costs just one pin and a cent for an external cap.

 

[leo_o2]

I also read that paper. It is difficult to get high psrr at high frequency. You should to have experiement for the result.