We have a Tektronix SD-24 sampling head (used for TDR and eye diagrams), which used to work fine, but now fails (plugged into a CSA mainframe) with a power on self-test failure. Seems to be a hard failure.
The messages say "error index m1661", and say the problem lies in the "EEPROM area" of this sampling head. The error definitely follows the sampling head, if one swaps heads with another similar unit.
Inside, one finds a common NatSemi 93C46N eeprom device (8 pin DIP). No visible soldering (or similar) problems.
I'm tempted to stick the EEPROM into our eprom programmer and see what's in it (and whether it's dead or just lost its memory). Without knowledge of the format of the contents, it may be hard to work out what bit dropped off (one might be lucky and identify a flakey bit by voltage or temperature stress, I suppose, but somehow I doubt it).
Does anyone know what's stored in it? According to the manuals, when one runs head calibration, the values are saved there so they follow the sampling head if one moves it to another mainframe. I think there's a factory default set of calibration values there too for when one can't be bothered to run the calibration sequence. There's also a user-programmable label for the head. But is that all, or are there other important numbers hidden in it?
Is there some secret way to re-initialise it (hidden menu maybe?)
We have a second similar (working) sampling head. I presume if I read that one's EEPROM and cloned the data (into a new 93C46N) for the failing head, then it would pass the checksum (or whatever) test on the EEPROM. But would this (after running the "Enhanced Accuracy" calibration sequence) lead to full function? Or is it more likely that performance (spec: 18ps risetime and 20GHz bandwidth) would be dire without it going back to the factory for re-certification?
Thanks for your views...
Regards, Mike.