In complex mode I can quickly press one after the other g MEM, f PREFIX, f (i) (or the other way round) - the display changes from memory state to full mantissa, then off(!) and only after the remainder from 2 seconds again on. Perhaps this is what happens during subsequent PSEs in a program - the next PSE hits before the previous finished and the display is gone.

Best regards
Dirk

- in Addition to
http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/articles.cgi?read=1089

First I had tried PSE in a loop before I realized that when I saw only the very first result and then a blank display it was exactly what's written in the first bug description.