Here it is the ECU Finished with the module inside the ECU ... it uses 4 pins for the coils . A19 , B3 , B4 , and D5

 

Open source for Arduino? We really appreciate your contribution to the community.

 

Can you elaborate at all about this?

 

Well we've been testing the microcontroller and it looks it has been very reliable yes !

 

Well , it all started here, but not a lot of response from people . but the whole idea was to create a cheaper option for coil on plug for OBD1 .


https://honda-tech.com/engine-manage...-pack-3274406/

 

So what have you done with the distributor sensors ??? Using this ECU combined with the D17A architecture makes the job easier than using your board and ECU combination on a B, D, or H series engine. How do you intend to integrate your system with these engines ??? If you can remove the conventional distributor, you will have a winner.

 

With the D17 it's been a little hard to use the original sensors (cam and crank) but with no luck , because remember OBD1 has 3 sensors TDC , CKP y CYP , and 3 must be in the ECU to do synch . even tho I tried to use them , didn't work , they are not the same count and it only has 2 , Cam and crank , thats why in the project for d17 I took the road to install the dizzy , which is fine for now ... and even if it worked in the d17 my intentions were not focused on the d17 only , it was mostly focused for all OBD1 , that includes B , H , D , the idea is to remove the distributor functionality of high voltage distribution . but remember the CKP , TDC and CYP are inside the Distributor , so we can't get rid of it for good , but at least we can get rid of the high voltage distribution and convert it to Coil On plug, Im on my way to the Dyno to try this on a hatch with LS vtec , I will post results later ...

 

Very cool in for results

 

well , I just came back from dyno, I understand there is some tuning necessary on this but blue line is without COP and red with COP , Im having an issue with the dwell at high RPM . I have to lower it down to 2.3ms so it doesn't desynch but I lose a little bit of power , plus there is a difference of timing with COP , probably because the signal is processed faster in the microcontroller than the mechanical parts inside the dizzy , an offset of around 5-7 degrees adv , an offset conf will fix that , but not a huge issue that cannot be resolved . if there is someone that is willing to help in the project who knows about arduino and microcontroller programming is more than welcome , Im going to keep working on this to make it better and make a dynamic dwell programming also to avoid the desynch problem and finish it.

 

Those gains would be better realized on a car making more power obviously. But so far so good!

 

Yes that's true , also , I didn't retune the car , fuel and timing can be retuned and get a little more

 

hondata is probably looking for ways to sue you as we speak lol


post it on pgmfi.org forum we need a project to keep tat forum alive

pgmfi.org is somewhat part of Honda history by now...

 

Look into the uc32 boards from ChipKit. They're 32 bit Pic-based and run at 80MHz (Arduino Uno is 16bit/16MHz), so basically double the precision at 5x the speed. Same form factor as the Uno and their IDE takes Arduino code directly so you can drop your current sketch right in and tweak it from there.

chipKIT uC32: Basic Microcontroller Board with Uno R3 Headers - Digilent

 

very nice work, would you make public the code?

 

hope not lol , I think I have to re enable my account there , its been a while since I posted in that forum . its been dead for a long time . what happened with that lawsuit by the way ?

 

that uc32 looks very nice , but to be honest Im looking to make it even smaller, plus if you do the math , I only need a proc that can handle 6Mhz that will be good for 13K rpm , unless we are installing this on bikes lol . but thanks for the tip , Im actually getting one for another project I have .

 

why would I keep it to myself ? lol

 

You couldn't be sued for your device, there's nothing there to infringe on their IP. They didn't invent COP or hold patents on it.

I figured, just throwing it out there as an option to get into 32 bit without re-prototyping the whole thing or learning another code base. I would asssume that your end goal is a proprietary board. I think it would be worth the $30 to experiment with.

 

I was working on this at one point but never got to the point of testing it on an engine, just had it running with the engine simulator on the bench.

Feel like posting your code? I assume its all running in the Arduino IDE since youre using the board.

 

Im starting to consider a little more the board cause Im adding the Offset Retard to adjust the timing to the correct number and also thinking about the Dynamic Dwell .

 

I know how you feel , I was doing this even before hondata published their coilpack retrofit . And yes , everything is running on arduino IDE.

 

This is the first revision of the micro controller programming , as of now , I take no responsibility of the use of this information . this is the beta version and its just for development purposes.

 

Why is that a zip file? it's only 147 lines of code.

 

Good job!
In later engines (such as F23's as I know), the sensors are located at the crank pully side of the engine. I think in those engine the dizzy is just working as a distributor and coil, not like the older ones which contain sensors too. If I'm correct, you could completely remove the dizzy in those engines; Which makes the bay more shaved and clear. What you think?

 

Well the one on the crank in the later engines is a crankshaft fluctuation sensor, part of the obd2 system's misfire detection strategy. It doesn't actually have to do with timing the engine, but can be used for such with an alternate setup. You can't run an engine with only crank signal though. There has to be a sensor on the cam or the engine cannot sync.