Elio (in 2012) has done some work on the FEI rubidium <http://comments.gmane.org/gmane.comp.time.nuts/24826>.
On Tue, Mar 11, 2014 at 10:44 PM, Skip Withrow <[email protected]> wrote: > Hello Nuts, > > Looking for some help with an FEI FE-5650A. This is one of the units that > is surplus from the telecom industry with the Option 02 serial interface. > Unit was working fine - locked, and was able to control the frequency by > issuing commands to the serial port. Then something funny happened, > oscillator went unlocked and cycling the power does not help. I have > dumped what data I can (see attachments) and the software inside the 5650 > appears to be corrupted. > > Attached are two files, one from a good 5650A and one from the bad 5650A. > It appears that several of the data areas have been over-written with the > 9-byte sequence 2D 09 00 24 00 00 00 00 24. This would be the response to > the 2D query command, except the data in the frequency register was > probably not 00 00 00 00 at the time of failure and 24 looks like the > checksum for the header and is not the checksum for 00 00 00 00 (which > would be 00). > > Is there an ex-FEI engineer out there in time-nuts land? > Has anyone else seen such behavour? > Is there any known way to reset the software for the 5650? > Has anyone dumped the 5650 microcode (I vaguely recall one of the rubidiums > having been dumped)? > > I took the processor board from a good 5650 and put it on the bad unit and > it locks fine. My next step is to write a control word to the bad unit > then read it back and see if it shows up in the memory dump, to see if the > memory over-write is dynamic (the program has been premanently changed) or > if it looks like it happened only once. > > Any information would be helpful, or suggestions on debug. > > Thanks in advance. > > Skip Withrow > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
