Pantera EFI

Hi, you are correct, NO patent infringement as the Patent Expired in 2006.
The team members, I am one such person, where not the FIRST DIS car inventors, it was a Ford Model "A".

Do you use my coils ?
What is your Coil Dwell ?
What is your Anti Dwell ?

I use a 48-2 distributor target wheel, ONE sensor.

I too have another question, the VR sensor's wire polarity data, the (+) & (-), seems to be reversed in some of the shop manuals.
Would Blue be (+) and White be (-) in your eyes ?

Lance

Kanibaals

This could be used for distributor replacement - EPS made from Suzuki called SUZUKI 33100-77E20 Camshaft Synchronizer.
And a signal converter that convert the signals to fully comply the HONDA's distributor's timing, thus cause no CEL at all.
The EPS itself won't fire the engine it must work with converter.

zamiex

any write up on this project? just wondering how complicated the wiring will be

Roel Dus

Hi amadeuscr,

Props on the arduino COP project I'm in the process of trying it myself. I was wondering did you take into account the slow speed of the arduino's digitalWrite()?
It's basically about 10X slower than using the port config method :

http://www.instructables.com/id/Ardu...how-to-fix-it/

Now it could be you already know the speed of digitalWrite() and just added a delay on top of that to get the propper delay. In that case never mind.
Reading from the link I posted there is quite some speed difference.First I'm going to try the arduino. Question though what type of coil on plugs are you using? From a K24?

Best regards,

Roel

majdi

Any idea how to modify the code to make degree adjusment by using variable resistor?

Roel Dus

Yes the Arduino has analog inputs. You can use one to ref the delay time off. Have to be carefull with that though. Don't want to burn coils or have too weak of a spark. Basically the dynamic dwell methode is good. It's just that the dwell time is added to the DigitalWrite() execution time.

ES1_OBD1

My arduino will be here tomorrow and I've already got the obd1 ecu in the car and datalogging working. Thanks for the information! Just a question, how would this react to ignition cut type rev limiters?

Roel Dus

Well I'm guessing ignition type rev limiters cut power to the coil in the distributor. So now that has to be set up the cut the power to all the four coils-on-plugs.

ES1_OBD1

It cuts the ignitor some how, that much i know. I was just trying to think in my head if the ICM signal disappears how is the arduino going to react. I received mine today and I am about to wire it into my ES1.

EDIT: I have thought about it and i believe this is what is going to happen is when the arduino misses the ICM signal its going to throw the sync off. For example, it hits sync, see's ICM signal, decides to fire off cylinder 1, hit rev limit(miss ICM signal), ecu and dizzy move to cylinder 4, arduino recieves ICM signal to fire next cylinder, but it still thinks we are cylinder 3 due to the missed ICM signal and we get a misfire. Deep down i feel the engine is spinning so fast at 4k rpm it wont be felt.

Using 4 sync wires, one off each injector, and the ICM signal, it should work like stock. Logic says if cyl 2 injector was the last to fire, its coils 2 turn to sprark. It would use 3 more wires, but i feel it would be more accurate.

Roel Dus

What do you mean when you say "dizzy move to cylinder 4" The mechanical moving to the next cyl is removed (was done by the rotor). The ICM is the trigger and moves over to the next cyl to fire by programming. It can only be off sync until Inj1 fires again. True using all 4 inj wires would be safer but not more accurate. Timing wise the ecu will provide the exact timing, as long as the arduino can keep up it will always set the next cyl to fire. Keep in mind I had to 0.25 speed a movie clip i made of it at idle to verify it was working correctly (watching the led's). But why even bother? There is an RPM counter in the programming. You can easily program it so the fire_coil functions only work under a set rpm hereby creating your own rev limiter. The software will actually still be syncing but no signals will be sent out until rpm is dropped again. Once the firing resumes it will still be in sync. Btw i'm doing the micropython programming right now. I/O timing is no longer an issue. Will keep you posted.

Best regards

ES1_OBD1

I mean the ecu will be firing the next injector, but the arduino will be firing the coil we skipped. I would rather not hard code my RPM limit in the arduino, as this is for two step/full throttle shift with ignition cut. I don't think i will have a problem, i just don't want to blow apart my plastic manifold. although, i have been looking for a reason to do a y8 manifold besides just because. If I do have a problem I will do the 4 wire sync method and will report back. I am going to try to get it running tomorrow fingers crossed I have an early day with work.

Roel Dus

Alright cool! I may go the same route too. Right now I'm experimenting with interrupts for the sync. If it works I may as well sync all the cylinders that way.
I'll see if can can get some pics in.

Best regards,

Roel

ES1_OBD1

Thinking about it, the 4 wire sync seems the best way to do it, I think the code might be shorter also, but i haven't made it that far yet. Long day at work and I am not sure if I will get to it today. The arduino is cool stuff though, I killed a few hours yesterday just playing wit the on board LED LOL!

Roel Dus

Ok so made some great progress. Got the interrupts working. But because I had only 4 inputs I had to work with 3 injectors and one for the coiltrigger. But I'll get an additional opto board soon.
The program basically doesn't even need a while true that always loops, it works solely on interrupts. But I'm going to do some RPM calc and with that the dwell calcs so I'll have use for a loop
routine running calculations and then passing those to the interrupt methods. Soon I'll connect the signals to the car and see if it can keep up.

Roel Dus

Yay some testing done.

So connected all the ECU inj signals and made a simulation prg cuz I can't yet connect the coiltrig. But all the interrupts work. They are actually high signals that go low when injector fires but I just had to change the interrupt to trigger on the falling edge. All injectors are registered even at high rpm. It was scary at first to see all the injector leds dimm but it's normal when you let off the throttle it cuts the injectors until it idles or when you press the throttle. Really happy with the testing. I'm going to get the actual COP assy's and the additional opto board and wire everything up. Also the rpm and dwell calc will have to work and all should be fine. Maybe some tinkering with dwell times ect. Then real world testing by actually driving. Then I plan to make a PCB with all the signal processing all integrated on one board. I did make a movie clip but it won't upload. maybe it is too big? Will post it on my fb.



Best regards

majdi

Basically i test the arduino code, i got knock while 4000 -6000 rpm. Adjust with ostrich neptune still same knocking..

Roel Dus

Ouch watch out for that! May want to check all the wiring / solder connections. Are you using shielded wiring or straight in ECU? Did you use the diodes and resistors? Also what coils are you using? The dwell settings also have to be matched to the coil hardware. May want to compare sparks by running just one COP next to a conventional one. Adjust dwell from that to get the right base line. Are you using that variable resistor you spoke about? How is that setup and what kind of delay is it making? You will need a multi turn pot to get the needed resolution. Just a few things to consider. Good luck.

majdi

Im using K20 coil on plug.. Yes i used resistor for pin inj signal and icm. Wire i not used shielding one, but i twisted with ground wire. And no i not using diode.

Any how i add diode to each coil? can tech me?

I will try change the wiring and the diode add and see if there knocking or not. And NO in not using variable resistor to adjust the degree.

Roel Dus

coil output arduino ---|> ---- Input COP plug = diode 1N5817 or whatever type
did you remove the old ignition coil and replace with 680 or 1K Ohm resistor?
When using shields only connect shield on one side. (it can carry the interference)
Keep wiring out of the way of possibly interference sources.
What are your dwell settings? keep in mind the DigitalWrite() is already taking 3840 uSec. The total delay may be too much for the K20 Coil, in that case lower it.
I don't know what type a K20 cop is? You need the exact specs of it.

Pretty brave just throwing it on there BTW. You should always measure http://static.tiepie.com/gfx/Product...P750_InUse.png

first.

Compare to a conventional connected spark plug or distr coil. When signals look the same (same high time/low time), then do the complete conversion.

Best of luck

Roel Dus

Ok time for an update. I've been working on my B20z swap a lot so that's why I've not been posting but I did receive the Honda Fit coil-over plugs and got it firing.
I also ordered some scope tools for measuring. I'll be connecting cyl nr 1 to my mcu and will run the engine like normal to compare the timing. I'll post a vid of it later.

Roel Dus

So got around to do some COP tuning on cyl4. and here's the result. Had to tune in a delay of 525uSec before firing and signal is kept high 600uSec for now. Will see how the coils deal with it and how the engine runs.
Unfortunately the coils I bought for 5 Euro a piece were really shitty. They got really hot. one of em blew.Spark is also weak. Thinking about getting some Cm11-102's and will be driving it with a sr20det ignitior connected to the mcu via an opto board.

https://www.youtube.com/watch?v=q9NhyVOj724&feature=youtu.be

Link to the cm11-102's

So yeah, keep ya posted

Roel Dus

So it's been a while since my last post and here's some updates. I've been working on my B20z civic alot and the engine is in and running great. Just finishing up putting alot of dots on the i's.
I did just get my CM11-102's in so expect some testing soon.

amadeuscr

Interesting ... didn't know the delay using digitalwrite() , that explain the difference in timing , I'm going to try the fast one . very good insight , thanks .

amadeuscr

it should resync , but we should test how it works in real life, my concern its the delay for resync ,.

amadeuscr

if the arduino loses the ICM signal , as soon as it gets it back it will mark the RPM as too low and will desync , and it will be forced to resync again and will catch up at some point , the problem its the delay resyncing I would say