Mike was having trouble with his computer. He was getting very
    discouraged. It would dump the B computer fuel map.

    Is a voltage drop causing computer problem?
    This is more likely to occur if your batteries are in
    the back of the airplane. I would run heavy
    wires for the starting and alternator  functions
    and a separate smaller wire directly from the battery for the main bus.
    The reason is the battery acts as a giant capacitor so starting
    and alternator glitches are less noticed on the main buss.

    Here is a sketch of his wiring design that Mike drew and that I
    have edited a bit.

    The other possibility is the computer is not sensing a power down
    situation and not cleaning up what it is doing in software. A way of
    doing
    that is add a small 5 volt rechargeable battery that keeps
    power on the computer long enough to clean itself up.

    This can mostly be done in software if one analog to digital converter
    is dedicated to monitoring the raw 12 volt supply. If the voltage
    drops below say 9 volts the computer starts cleaning up the software
    for shut down. Kind of like old Windows where you had to tell it to
    shut down and Windows would clean it self up before actually shutting
    down.

    The 5 volt regulator feeding the computer usually has a very large
    capacitor that can keep 5 volt power on the computer long
    enough to clean it self up. This capacitor can serve as
    the small five volt battery.

    Paul Lamar



Voltage spikes (possibly through grounding scheme) is a more likely
candidate.

I had one case where ground bounce caused this kind of problem and I
couldn't talk the builder through corrective measures (did not speak
english well).   Ended up having to add a number of filters & such to
the EC2.  The ones I can remember are:

Added a spike clamping diode  (1n4004) across the coil of the injector
relay.
Added a .01uf  capacitor to PCB ground to the Master Clear pin on both
controller chips.
Added .1 uf capacitor between ground & the air temp sensor connector pin.

Didn't hear of any more map table corruption problems after that so I
assume it helped.

Other than bugs in the code, the ONLY way corruption problems can occur
is through  large voltage spikes (usually through ground).   There is an
elaborate software procedure needed to do any alteration to EEPROM where
the table is stored.  The other way to erase it is through program
voltage (+12 v) being applied between ground and master clear pin.  That
is the reason for the filter added on that pin.  It would be better to
eliminate the root cause because there are other possible side effects
of ground noise.

Tracy

Can you do this on Mike's early computer if I send it back?
I added vibration isolators with braided case grounds to the mounting arrangement.

It appears to be well grounded to the air frame both before and after.
Also the 6-32 ground bolt on the computer has at least 4 wires going to it.
Maybe more.

Paul Lamar