I doubt your issue is related to the CE# pin. If your device is the only device on the bus (in addition to the processor/host) then connecting CE# to ground should not cause any problems. The CE# pin just tells the 9914 that the current transaction is intended for the 9914 instead of some other device on the bus. If there are no other devices on the bus there shouldn't be any transactions that aren't directed at the 9914. When I say bus I mean the bus between the 9914 and processor, not the GPIB.
Can you confirm that the *IDN? command results in a valid response from the 9914 at the host? If so a lot of your firmware/hardware can be assumed to be correct.
When you run "scan for instruments" does it say there is a device that did not respond to *idn? or does it not detect any device at all? The GPIB signaling that allows MAX to detect the presence of an instrument is handled in hardware. The response of the instrument to the *idn? given by MAX requires firmware intervention.
If MAX does not detect any device there is some fundamental issue like pon being asserted. I think *idn? first asserts IFC, is there a chance your firmware would set pon based on an IFC assertion? Is there any case in which your firmware would set pon and not clear it immediately afterwards? It is also possible that the GPIB address of the 9914 has been reset to 0 (same as GPIB controller).
If MAX does detect a device but says the device doesn't respond to *idn? there is likely a firmware issue. My guess is that the firmware leaves the 9914 in a different state after the first *idn?. Possible an IMR is not set correctly.
