If a sensor fails, you have to flip switch 2?

If a sensor fails, MyECU doesn`t work if you don`t flip that switch?

Unless Cliff introduced any of the stuff we talked about earlier in this thread, the ECU will still do its job but depending on which sensor failed, and failure mode, the engine will run too lean or too rich, possibly to the point of not running at all.

Just look at the map: For example, if you loose your air temp sensor (going open circuit), the ECU will think it's like -30˚C and you will run in the order of 25% rich. You will probably do fine without flipping that switch (except for poor mileage and fouling spark plugs). But if the engine temp sensor failed the same way, you would end up 65% rich and I guess it wont run at all.

Flipping that switch will make the ECU ignore all baro sensors and use generic values that should take you home in most conditions. It may be harder to start and you will probably get poor mileage.

Raz is absolutely correct and thanks for the explanation.

It is possible to improve on it. If it never gets below 0, for example, you could adjust the air and oil temp correction tables in the map such that the values below 0 are all the same. This would limit the deviation.
Also if running closed loop the sensors will be over ridden to about 25% correction.

Ah, that never occured to me! I think I can even improve it further: Since a sensor is very unlikely to go half bad, the following should work even better (considering my normal riding temp is 10-30˚C):

# Prime gives the fuel injection time in ms while cranking
#Temp C   -30   -20   -10     0    10    20    25    30    40    50    60    70    80    90   100   110   125
Prime      12    30    25    20    17    12    11     9     8     7     6     5

# Crank is the %% boost of the map injection times right after start
# this boost decays to 0 over about 20s
#Temp C   -30   -20   -10     0    10    20    25    30    40    50    60    70    80    90   100   110   125
Crank      60    70    65    60    60    60    55    50    35    25    25    15    15    15    15    15

# OilT is the permanent choking boost
#Temp C   -30   -20   -10     0    10    20    25    30    40    50    60    70    80    90   100   110   125
OilT%    +0.0 +53.9 +44.5 +35.9 +29.7 +19.5 +10.2  +9.4  +5.5  +0.0

# AirT is the barometric adjustment for air temp(based on ideal gas law)
#Temp C   -30   -20   -10     0    10    20    25    30    40    50    60    70    80    90   100   110   125
AirT%    +3.4 +19.8 +15.2 +11.0  +7.1  +3.4  +1.7  +0.0  -3.2  -6.2

This way, if a sensor claims -30˚C, an air temp of 20˚C or engine temp of 50˚C is assumed instead. In most conditions you would hardly notice the problem at all. I think I will use the above.

You can't do the same to the air pressure correction though. Will the ECU accept a low-end reading and apply the formula? If so, it will be like 90% lean I think? That would necessate a DIP switch flip.

OK, may work but a tad McGyverish.
Sensor failure will most likely not occure but what if?

Think I`ll wire my SW2 to a waterproof switch
in my P8 case if possible.

raz wrote on 05/19/10 at 21:23:20:

You can't do the same to the air pressure correction though.

Or can we? If I'd change the line [for My16M]

MyECU Cfg Pressure=249mV,170mB,4751mV,1050mB

into

MyECU Cfg Pressure=2960mV,700mB,4751mV,1050mB

Would that change anything? Or does the ECU just find out the straight line formula and end up the same? I picked the standard pressure at 3000 m, the altitude where I'm not recommended to soar without oxygen mask 8-)

Wouldn't it be easier to carry a resistor equivalent to normal operating and connect that up instead of fiddling with the map?
Silly me, I know you would much rather tweak things LOL :)

That is hardware tweaking, I prefer soft tweaking whenever possible :)  But you are right of course. A common 3.3K resistor could replace the air temp sensor or a 1K one could replace the engine temp sensor.

I guess if my pressure sensor would fail in the middle of nowhere, I could just connect that input to the 5V feed. That should produce a somewhat rich mixture, instead of one too lean to run or even start the engine. With your sensor sitting inside the case, it would be a little harder to achieve. Anyway I can't remember ever hearing or reading about a failed Guzzi pressure sensor.