brainstorming area for developers
AfreshTiny, VemsFrontier/MiracleBoard, VemsFrontier/ArmEfi or some similar simple hardware can be used for interesting applications. Use your fantasy.
Boost target as page and other boost control related things
Certainly it´s really cheap having a 0-5v input and solenoid connected to the VEMS to run your boost control. Extremely cheap even,
However without connecting to the ecu in any way you are going to be limited in knowing what your boost target is.
I´d like to see a output channel in (??) where the page is changed in the Round when you change the boost target by more than X kpa.
i.e changing of the target forces the round to a specific page (where boost target is shown) for a number of seconds
The same would be in effect if the round is not connected to the vems and just a round by itself, making the round a fully fledged boost controller with updating display and could be marketed as such.
Flashing display for certain values beyond-limit - feature proposal
Also ideal would be that if a sensor input goes above a value the round swaps to that page and flashes repeatedly until the page button is pressed to acknowledge the alarm, the page would not change while the button is being pressed.
Since there are 8 pages, a single sent over byte could run a
force action to the round, i.e when bit 4 is enabled then change to page 4, if two bits are enabled then ignore and so on.
This gives the VEMS ecu some control over the view on the round.
Flap Substitute - same HW for MAP substitute
Many cars have a primitive and rather restrictive flap-type MassAirFlow sensor. It outputs a logarithmic airflow signal.
We could replace the flap-sensor with a small board with inputs:
- MAP
- RPM (HALL or coil-primary-input seems easy; VR requires LM1815)
- CLT (optional)
It also needs a serial (FBUS, see CommMatrix) port, so a cheap cable can be used for mapping from PC. No max232 of course.
output signal (only 1 output-type needed on a unit):
- analog voltage (filtered PWM). The Nissan AFM has a maximum +5v output (otherwise it's a logarithmic curve), when the ECU see's maximum it is injecting at what ever its peak mapped amount is.
- but some ECM's might eat unfiltered PWM or
- frequency
So the
- primitive, mechanical, restrictive, slow, unprecise (hysteresis!) failing unit
- fully mappable, electronic, good airflow, fast, precise, reliable unit.
Fully mappable
means that serious engine mods are possible (cam, turbo): besides the remap only the injectors need an upgrade:
- without replacing stock injectors with bigger ones the system would be maxed out soon (most systems sized to apply 85% injector duty at max power: marginal headroom for even more power)
- with bigger injectors, the remap (lean-out the map) is essential of course, since same airflow will need shorter pulsewidth.
For MAP substitute the most notable advantage is just the full mappability without serious harness modifications. (MAP based SD systems are reliable and usually fast anyway).
This could be bolt on to thousands of vehicles. Simpler implementation, manufacturing, doc and support compared to our high-tech toys. And they could finance the development of our next generation of high-tech units.
This is a great idea, depending on price it could be very intresting for diy turbo systems. Here is a link to the similar commercial ecu: http://www.mapecu.com/
Optional stuff:
- display (eg. LED or LCD) used for SimpleUserInterface
- 5 buttons for SimpleUserInterface this is the real question
- real measurement of airflow (hotwire requires similar code as our wideband; hotfilm I don't know). This is not really necessary.
- more advanced comm than the FBUS (not really necessary)
Note that this makes it possible to properly tune such a car without messing with the EEPROM (safe feeling of easy revert) or doing unclean surgery (such as CLT-hacks). Also, it's naturally a good supplementary unit for the wideband (or the other way around).
Note that we could apply lambda-feedback. (with a slightly improved board: like our 2-cake round, but preferrably on 1 board - need not be circular)
Note: The flap is the ECU on all the older guys; the box just take the NBO² + spark + temp + flap voltage and with analouge mixers make a PWM (whatever has the NBO2 feedback and spark output that should be called ECU; but anyway it doesn't matter what we call it, the point is we can plug in a simple circuit that boosts the possibilities with that engine). The flap box has an offset pot and a thermister and the like 10 taps to trimmed resistors. As the flap moves it selects one of the taps so that resitor + the offset pot + the temp thermister makes the output voltage going to the pwm box.
issues:
- EGR is a serious issue. Any engine that has EGR needs real air-flow metering, or a wideband sensor in the intake (or it could work with LSU4 sensor in the exhaust too). First we just say that EGR engine is not supported.
[PhatBob] Many people block off their EGR when they start modifiying their cars, so no EGR Support may not be much of an issue (most of the people who would use this system would be blocking their EGR's off anyway)
TODO: check Bosch 0280203023 unit output curve, pinout. I think connector is: [5 pin recepticle housing] and where is the PCB header? I gave up [ here]. It has some tyco 5-1 pin JPT (the 5th pin is missing for keying).
TODO: find similar
Sensor Curves
- - Nissan S14/RB25/VG30/VH41 [Sensor values and graphs]
- - '94-'95 Mustang GT [See bottom of page for values and graphs]
TODO: start to collect a library of default curves for several engines
[air-flow sensor prices] are normally 180..416 Euro.
See also