forbiddenera

Oo /subscribed

Someone should have pointed me to this. I'm all about hacking these ECUs and Arduino programming.

You have a willing test mule here. Have an Arduino, chipped ECU with data logging and ability to burn chips and a turbo b18b1.

I don't have any COP stuff but a trip to the JY can probably fix that - what vehicle do I need to pilfer coIL packs from?

ALSO @adeel please post link to the user using his Arduino as an ECU.. interesting, I've thought of doing that but the amount of work involved vs burning a chip seems rediculous, curious what he's using for ADC.. most Arduino dac/ADC I've seen don't even support 5v.

amadeuscr

Well , I changed the code and added a macro to avoid using digitalWrite() function , this code works on Arduino NANO , I'm adding the picture using the NANO , which is smaller , I'm still figuring out where to mount it .

Digital Pin 3 -> DIODE -> COIL 1
Digital Pin 5 -> DIODE -> COIL 2
Digital Pin 6 -> DIODE -> COIL 3
Digital Pin 7 -> DIODE -> COIL 4

Digital Pin 8 -> 1k Resistor -> ICM
Digital Pin 12-> 1k Resistor -> INJ1

Also , I'm using 5v instead of VIN as power, 12v from the car is not that stable. So I connected the 5v power to D19 which is power source for MAP (5v power source) , so power goes like this :

5V PIN -> D19
GND -> B2 , ANY GND IS GOOD

Also , I'm not using the ECU Pins for coil output, I'm using direct cables with a connector , I guess its easier .

Attached the ZIP FILE with the code.


Cheers

amadeuscr

Get it going brother ! , in the last post you will find the updated code and info.

ES1_OBD1

Is the updated code still getting the 4-6k rpm knock?

Roel Dus

Good to see some new life in this thread. Props on the func.h amadeuscr the DigitalWrite had me worried before. For me I'm still putting in the final touches on the B20z swap in my civ. But it's itching to get back on the cop project. I do have the right coilpacks now so I think it should be up and running soon. Will update Great work guys.

Roel Dus

Alright made some progress today. I had to update the MCU firmware to fix an inverted signal. that worked and then I got the engine running with one coilpack and the other 3 conventional by the rotor. I tuned it and timed it just right. So then I went to wire up the rest to run on all four coilpacks and then it seems like it wants to start but the immobilizer kicks in! Alarm lights begin to flash and I have to press unlock to reset the alarm. It's now dark out so tomorrow I'll see If all the coilpacks actually fire and I dunno If I can't get it to work I'll have to move everything to the civic. Kinda weird, it's like it wants the feedback from the original coil or something. Any idea's?

forbiddenera

Wtf? Thought we were doing obd1 here..!

IIIRC, You can remove the immobilizer from obd2 ecus.

Though it shouldn't kick in with key in ignition (rf tag in ring)

I do automotive wiring / remote starters / immobilizer bypass for a living so lmk if you need help, I am definitely interested in this project.

Roel Dus

Yeah it is obd1. It's a 95 accord. Alarm/immobilizer whatever it is. maybe it's not OEM but it resets on the OEM keyfob?? After a couple of cranks I hear a relay kick in and it intterrupts the starter so bad it almost seems like a mechanical jam.
Had me really spooked. then the hazzard lights blink until I press the keyfob. I'm definitely close so I'm not giving up so easy. Thinking about just running the ignition power from the key switch directly to the plug at the dizzy.
But... it also interrupts the starter. So that needs bypassing as well. Or it could be I need to place a resistor in place where the coil is to have the system think it's running with working coil..
Otherwise, like I said might move it all to the civ and get it running on that first. Thing is the Accord is my daily and the civ still needs to pass a special MOT inspection because of the B20z engine swap.

ES1_OBD1

I finally got a set of coils, wired up the arduino, put it in the ecu. All good. I warm up the car a bit on the distributor ignition, shut the car down and install the coil packs and arduino. The arduino takes several cranking attempts before it will sync, but once it does the car starts then dies. Repeat this 2 or 3 times before i have to leave for a bit. I come back and the car is completely cold now, same thing as before, several cranking attempts before it will sync, and when it does sync, the biggest backfire in the world came out of this exhaust pipe. Tried again for a second time with the same result (this is the definition of insanity). At this point I unplug the fuel injectors and hook up a timing light but the smell of e85 at this point is so strong we decided to let the fuel evaporate for a while and that is where the car sits as of a few days ago. We used coils out of a 2003 accord, instead of d17 coils, i figured they would be about the same internally, wiring matched up. I suspect that the timing is massively retarded but i have not confirmed with timing light yet.

amadeuscr

are you using INJ 1 to synch ? , can you post some photos of the setup ?

Sergio Garcia

Hello here I leave the diagram on the connection to Arduino

Greetings from venezuela and many thanks to the AMADEUS friend for his help

ES1_OBD1

I am using an UNO, and the original design, and coilOnPlugRE.zip

my injector sync goes to A1 in the ECU, and pin 8 on the arduino through a 10k resistor.
my ICM sync goes to A21 in the ECU, and pin 12 on the arduino though a 10k resistor.
pin 4 on arduino goes to coil 1 through a diode.
pin 5 on arduino goes to coil 2 through a diode.
pin 6 on arduino goes to coil 3 through a diode.
pin 7 on arduino goes to coil 4 through a diode.

I am powering the arduino from D19 in the ECU, and it is grounded to A23 inside the ECU.

I tried the ecu in another car just to test sync and it synced very quickly.

My ECU throws a solid CEL and will not give me any codes, but also is not in limp mode. When I first soldered in the arduino i soldered the ground to B1 on accident and the eeprom got hot, but the chip was fine, I am still using it in another ECU with no problems, and the arduino seems to functioning just fine as well.

svobik81

My engine is spinning for about 5 seconds before synchronizing

ES1_OBD1

I messed with this again. I also experience long crank before sync. My car will crank for 5-10 seconds, sync, start and run for about a minute where the car warms up just the slightest and it will stall. All future attempts result in same 5-10 sync, then a backfire out of the exhaust lol! I think I am going to look into adding three more sync wires off the other injectors so the arduino isnt "guessing" what coil needs to fire, also wouldnt have to wait for sync.

Spudz76

I run a 1995 F22 distributor on a 1999 F23 and simply got a 1999 subharness (chopped the longest I could get off a 1999 CL), hooked up the two crank sensors (four wires) to the harness where the F22 originally had them by the distributor, and made sure the shielding was grounded decent. Runs great, no difference other than physical location and if you don't get the shielding working it won't work (I tried "just wiring" them first, too much interference so it acted like the sensors were different / threw all kinds of sync codes - OEM shielded it for good reason it's carrying hall effect analog AC signals not amplified or digital so interference and loss are not tolerable).

Never got the F23 distributor to work at all even with the two crank sensors set up correctly... maybe that sensor is the one that is different.

spAdam

iTrader: (1)

First off, the stock distributor sensors are variable reluctor, not hall effect. Life would be quite a bit easier if they were, because it's easy to find an edge but harder to interpret the analog vr signal/zero crossing. Simple computers don't like analog signals. When I built my simulator I used a MAX232 style chip (MAX235, specifically) to convert the Arduino's 0-5V edge to a 0 crossing signal that the ecu would accept as a distributor signal. Quite possibly you could use a similar setup as a rudimentary adc to go the other way.

That being said, have you guys tried tracing the distributor signals through the ecu? It's my understanding that the ecu takes those analog signals and processes them into edges for the cpu to deal with, and that the adcs that do it live on the two vertical boards right above your Arduino there. I've never had any need to go after those signals so I can't confirm or deny this, but it makes sense. Maybe you could do some tracing and find where those pulses are coming out the backside of an adc and probe there. Find the cam sensor and you have your sync, find the cyl sensor and you open up opportunities for cylinder by cylinder timing adjustment or even knock windowing.

ES1_OBD1

The ECU is still processing the sensors, the arduino is tapped into the ignitor signal wire that tells it to fire the coil, and inj 1.

spAdam

iTrader: (1)

Oh I wasn't suggesting that you take over engine positioning from the ecu, but thinking that you may be able to get your pulses for any of the three signals right there. If there are nice clean 5V edges there, you're life gets that much easier. Use the one tooth signal for your sync and the 4 tooth signal to decide which coil to fire. The 12 tooth wheel could still be useful for calcs. Since the tooth count is low as crank sensors go (you would get 12 edges per 360° of crank rotation), even way up at 10000rpm it would only be 2000 edges per second which a 16MHz arduino should be able to deal with... depending on how the code is written, obviously.

forbiddenera

What happened to this?

I was thinking of Trying this myself. I also found someone who made a mini proprietary cop module that goes where the knock board connects..which is probably a good place to get your edges..! (Most knock controllers have a window and thus require crank position)

forbiddenera

No obd1 honda should have an immobilizer!

Trace your wires from your ignition switch to the fuse box and find an alarm in the middle.

forbiddenera

I have coils and arduino. Not sure if I want to try this yet as I'm using the car daily but..i might?

but if this project is dead then..

bank code

Is it possible?

forbiddenera

Dev possible and I'm itching to try it but I also heard about other boards like COPmini which is a specific board for our ECUs.

bank code

I tried it and it still does not work. arduino nano
Is it available?

bank code

arduino nano no rpm siagnal