FreeEMS firmware feature wishlist! (out of date)

[Lundgren417]

What about a hand held or small realtime display such a a palm baised freeware, or other handheld display, since sometimes we are not the only ones in our car. The use of a toutch screen would be pretty snazzy to

[Fred]

All of that sort of thing will be external to the EMS and communicate using the exact same WELL documented serial interface. I'm sure such toys will come with time though, I love playing with nifty toys too. I have an obsession with undersized electronics :-)

[AbeFM]

I wonder if I'm slipping in my old age, or what, but I definitely agree with things more and more over the past couple months. Jean's P&H boards (or any form of P&H in hardware) is obvious, adjusting an output based on load (cams, for instance) is surely manageable and can come along pretty quickly once things are working.

The free floating points are awesome, though I have a feeling they will be hard to tune? Or harder to model, etc. At any rate, it shouldn't be too hard to have the routine that goes out and get the value from the "table" be blind to if it's a table, independent points, or god himself whispering polite suggestions. I'm a huge fan of trying it though!

Perhaps some dedicated I/O stream protocols will have to be dreamed up. Maybe CAN? Anyway, I'd see two versions, an interacting and non-interacting - it would be nice to have dumb displays that just show what they've been told.
-Abe.

[Fred]

Free floating points should be actually easier to tune than a table. however, being able to do that is totally dependent on performance! If we can't make it run fast enough when the time comes, that's it... we are out of luck and will have to use boring old tables. You are right that it should be independent of the implementation though :

VE = short getVE(rpm, map);

[AldoQ]

Hi
is possible to add a thing like a valet , for limit the power of the engine, when your car is driving by another driver
add a g force sensor for a very rough virtual dyno
and a ABS module
hello from Mexico
best regards
AldoQ

[Fred]

Hello to Mexico, hello Aldo,

I would like to see valet mode too, so that should be in there if it isn't.

I want to have a G meter attached to it somehow for logging also, virtual dynoing should be outside the EMS though in PC software.

ABS should definitely be a standalone unit. I believe Lance at msefi.com is working on such a thing after his trans controller comes up to speed for GPIO use.

Thanks for your input :-)

[AbeFM]

Valet mode and all sorts of anti-theft code would naturally fall into the same category. Certainly something that put a 5 mph speed limit after 200 meters would make a stolen car stand out.

[AldoQ]

Hi
my pleasure

the "ECU" have a input for a car speed sensor?
regards

[AldoQ]

Hi all
for security system, is posibble to read the currient trasponder of the car key?
regards

[thebigmacd]

Here's another idea for mapping: fuzzy logic.

Define several parameters and conditions such as "IDLE_RPM_HIGH", "IDLE_RPM_LOW", "DECEL_MAP", "IDLE_MAP_LOW" etc and use fuzzy logic to define advance, pulse widths, afr targets, etc

It would simplify setup as the basic tuning would transfer from one engine to the next just by redefining the parameters.

People love to hate fuzzy logic, and built-in fuzzy functions have been dropped from a lot of microprocessors, so I don't expect pure fuzzy to rear its head anywhere in this project

But in a way, if you want the "floating point" method to work, it might actually make sense to model the algorithm after the fuzzy method. Abstraction can be a good thing if well placed.

For example:
If [RPM Is WARM_IDLE_SPEED]
And [MAP Is WARM_IDLE_MAP]
And [TPS Is TPS_IDLE]
And [CLT Is WARMED_UP]
Then [IDLE_PW Is OFF]

The math for this is something like the following:
IDLE_PW[OFF] = Min( WARM_IDLE_SPEED(RPM), WARM_IDLE_MAP(MAP), TPS_IDLE(TPS), WARMED_UP(CLT))

Where the functions WARM_IDLE_SPEED(RPM) etc return a 0-1 proportion representing how close the sampled value (RPM) is to the specified parameter (WARM_IDLE_SPEED). The minimum proportion of all the included values is written to the location [OFF] in the IDLE_PW array. This value determined how closely to "OFF" the idle control value will be. If several other condition statements are written that determine when the idle pulsewidth will be MINIMUM, MEDIUM, and MAXIMUM, the weighted mean of all four pulsewidths will be written to the pulsewidth routine.

Using the fuzzy method, this will not just shut the ICV off when all the conditions are exactly met, but as the engine state approaches these conditions, the idle control pulse width approaches 0, and as it approaches other sets of predefined conitions, the idle pulsewidth will approach the ideal for those conditions.

I've lost you all, I know.