DIY ECU build thread

It probably needs more fuel, then more air, then even more fuel.

Couple of things- you REALLY need to find a better way than just injection pulse width- I know that can be done in the future, but it will make the module much more portable from one engine to another. Which means that a roughtly good running Neon would run a Miata, if you change the injector calibration and MAF/MAP calibration. So very much easier to calibrate.

Second- you need ~6x more fuel during crank compare to normal running. Really, that much- you have a cold and dry manifold that you need to establish a puddle that will move into the chamber, and then hope there's enough fuel to fire.

Are you sequential injection during crank?

Lastly- buy and a/f meter- when you do that, you will see that on a cold engine, you need to add fuel to maintain the same a/f in the exaust. The enrichment need will go away both with time and engine temp.

Really- the a/f meter will answer a lot of questions.

edit- that 6x thing- that's for an engine at 70F/20C. The amound needed goes up as the temp drops. The most an engine will ever use is a -20F start, not WOT.

Privet! Im so interested in this project and would love to test this on my 93 miata but unfortunately I have no experience with ecu tuning, or coding but I would much rather put my time and effort learning into this device instead of the much less powerful and more expensive megasquirt. Tell me on a scale of 1-10 what experience level would I need to get this running? I know what my engine wants and needs as ive played with more antiquated tuning methods but nothing this modern ahem fancy. Keep up the good work man were rooting for ya

In reply to alfadriver:

I hope that tonight I will finally wire the wideband to this thingy... It's all about the lack of time :( Too much little stuff to take care off - bugs, wires, bugs, performance, wires, boards etc :( I need to sleep less but this would make me a slower angrier person.

I think once I have the algorithm and figure out the right values as in milliseconds for my car, I would make the default auto-calculate based on displacement and injector flow pro-rate from my values. Step by step :)

In reply to chiodos:

At least with a Miata you would be far from alone - there are four Miatas playing with this from time to time, one of them we are actually troubleshooting as we speak: http://rusefi.com/forum/viewtopic.php?f=3&t=729

Coding is definitely not required, but car skills are needed. I have zero tuning experience and my neon is running :) I am still learning, still somewhere in the start of the curve.

Awesome! I can definitely handle that, I now know whats going to power my miata when turbo time gets a bit closer

That's timing while idling - just a test of how stable it would be:

http://www.youtube.com/watch?v=xcxzZYOdPJs

russian wrote: In reply to MadScientistMatt: Any idea on what kind of testing budgets we are talking here? 10k, 100k, 1M? That's if we are talking certification tests.

The numbers I've seen have been in the mid five figure range.

Great build! Would it work with a Vortec 5300 V8?

G_Body_Man wrote: Great build! Would it work with a Vortec 5300 V8?

So far we have only started ten inline-4 and one inline-6. From the ECU standpoint, there is no difference between engine configurations, so this should work on V8. Wanna give it a try? :)

Please note this is a work-in-progress ALPHA version.

Check this out: a guy from Thailand has made his own rusEfi-compatible board

before I've got my own board with the chip on the board:

Big day for me - first rusEfi dyno run.

Stock ECU run: 121hp/127 lb-ft

1st rusEfi run ever - wooping 44hp/69 lb-ft. Right, 12 degrees advance does not work so cool

8 runs later: 123hp/128 lb-ft. More then stock ECU! better timing.

We did not tune fuel at all (Speed Density with default table), the plan for the day was to get close to stock values and just test everything. Mission accomplished :)

full logs

Awesome. I love checking in on this thread.

When are you going to tune for power?

Read through this thread! Way above my knowledge base. That's pretty awesome! I wonder if there would be any advantages to aftermarket ecu on my Speed Triple. Hmmm

Keep up the great work!

ProDarwin wrote: When are you going to tune for power?

Not for a while probably. For now I am happy to change the state of this project from Alpha to Beta, kind of need to figure out what's my plan for it now. Dyno is not free :(

escort1991 wrote: I wonder if there would be any advantages to aftermarket ecu on my Speed Triple.

ECU is just a piece in the puzzle, it really depends what your goals are. If it runs, do not touch it, unless you really want to :)

russian wrote: Big day for me - first rusEfi dyno run. Stock ECU run: 121hp/127 lb-ft 1st rusEfi run ever - wooping 44hp/69 lb-ft. Right, 12 degrees advance does not work so cool 8 runs later: 123hp/128 lb-ft. More then stock ECU! better timing. We did not tune fuel at all (Speed Density with default table), the plan for the day was to get close to stock values and just test everything. Mission accomplished :) full logs

I can tell you from personal experience that I would be less worried about HP and torques and be much more attentive to the AFR's. From your graph it looks like it went lean in the middle of the power ban when you were producing a lot of HP and torques. This could result in bad things happening to that motor.

BUT all that said I am hugely impressed by what you have done. Keep up the great work!!!! I have always thought that there would be a market for a stand alone that was not trying to do everything but instead was designed to use a very specific set peripherals and only use MAP.

For many people they will be much more comfortable making a specific type of sensor and/or coil work with there motor than sifting through all the options in a computer screen hoping that they got the settings rite.

Great work!!!!!

From your graph it looks like it went lean in the middle of the power ban when you were producing a lot of HP and torques.

Is that the blue line we are talking about here? That's stock ECU.

Somehow along the way this project has shifted from building a unit for my car into building a product maybe. You have a great point that people are horrified by computers and options, they need a nice & user-friendly platform. We'll see :) Still a long way to go.

Ahh yes the blue line if that is the case and yours are the other lines it looks like yours has better AFR's!!!!!

1st dyno day with the Festiva: better low end torque, for some reason lower HP above 5500

https://www.youtube.com/watch?v=1UKX0zRmWlU

Time has come to implement acceleration enrichment. As you can see, currently WOT (green) drivers the rpm (white) DOWN before it goes up. Wonder if anyone know how much (in absolute terms) extra fuel would I need and for how many engine cycles?

https://svn.code.sf.net/p/rusefi/code/trunk/misc/logs/2003_dodge_neon/DataLogs/2015_Mar_25/

http://youtu.be/Fn18tE8MrLA

We've started our first v8 today!

http://www.youtube.com/watch?v=qghT6uhH6_I

Looks like rusEfi is becoming a ECU of choice for education purposes, so

http://www.youtube.com/watch?v=HsKoD4tMfOU

I love every time this thread updates because it's always so cool.

A live-teaching environment with the ECU is a crazy good idea.

Missed this thread the first time around so I'm glad I caught the updates. Software engineering (and EE as well) are above my pay grade but I might have some insight regarding info about EPA /CARB compliance.

As others have stated, meeting emissions parameters is not enough. For components to be EPA /CARB approved they must pass certain durability tests. EPA requires components last somewhere between 80-100k miles (this might be stricter now) in an automotive environment at certain temperature extremes with a certain amount of airborne contaminants (or, as in the case of an ECU, the ability to ward off those contaminants for the required mileage. That probably means designing a case for the components). There are also requirements regarding the end user's ability to alter the ECU. That would be one reason we don't see plug-in chips on OEM computers anymore. There would be requirements regarding ease of altering the software (really the lack of it). These requirements change constantly so it's a moving target.

As you know, many states don't give a E36 M3 what happens to the emission controls after the car gets bought. That number gets smaller while the number of states requiring CA-type emissions profiles increases.

It is a daunting task to design a system for EPA /CARB compliance. So keep thinking about your original idea, which is building the least expensive ECU possible for the purpose of running a turbo on your LeMons car and achieving Total World Domination in that arena before taking on the EPA. Or even anyone else's tune. Get your own car running well first.

Now with automatic warm-up fuel correction. Currently hard-coded to keep AFR at 13 while CLT is below 80C, will probably make it a curve to gradually get to 14.7 while engine gets warmer

http://www.youtube.com/watch?v=Mdyr1dbYeW8