Project: LS2 Coil Interface

[derbybrit1]

The LS2 coils are real flamethrowers

[xlh1460]

I realized that in my last update, I had mentioned that the interface had covered the following requirments:

2. Convert the 12V (active low) signal from the ECU to a 5V (active high) signal required by the coils.

3. Allow for an adjustable dwell to be able to maximize the coil output while also maximizing the life of the coil.

Those were both covered in the discussion related to the oscilloscope waveform screenshot.

I neglected to mention:

1. Be compatible with the ECU’s spark event detection to avoid having “EFI” flashing on the dash continuously.

Since the ECU will no longer be directly driving an inductive load (the stock coil) it won't be getting back the signal it expects to allow it to determine that an ignition event has occurred and will cause it to display the "EFI" fault on the dash.

Obviously that would be undesirable to I had to come up with a solution.

The solution was to simply include an inductive load into the interface circuit to mimic the stock ignition coils.

The following photos show my test of that in the garage.

This is the bike running with one of the coils disconnected.


And this is the expected result on the dash.


This is the inductive load test circuit attached.


Fault, what fault?


Of course requirement:

4. DON’T DESTROY THE ENGINE

is yet to be seen.

[xlh1460]

As promised, here are the video clips.

The first one is of the stock coil.

They were taken with my iPod, apologies for the quality.

The gap of the tester was set to close to the maximum that the stock coil could handle.

I was able to open the gap while the spark was maintained since it apparently requires less energy to maintain the arc than it does to start it.

I then continue to open the gap until the stock coil can no longer handle it.

In the second video, the LS2 coil is able to start where the stock coil petered out.

I was able to continue to increase the gap.

I finally got impatient (and nervous about the energy) and opened the gap to the maximum setting.

The LS2 coil had no problem starting and maintaining the arc at that setting.

I love the smell of ozone in the morning.

http://s673.photobucket.com/albums/v...=StockCoil.mp4

http://s673.photobucket.com/albums/v...nt=LS2Coil.mp4

"Those coils are real flamethrowers."

[xlh1460]

I got time to "cobble up" some PCBs today.



Next I need to drill the holes and mount the components.

[derbybrit1]

Put me down for couple of boards, if you decide to make and sell them after perfected. My spare coils have been waiting for a new home.

BTW, I have had good success with Magnecor wires. The can custom make any length and terminations. I'll look up my ordering specs if you need them.

[xlh1460]

Put me down for couple of boards, if you decide to make and sell them after perfected. My spare coils have been waiting for a new home.

BTW, I have had good success with Magnecor wires. The can custom make any length and terminations. I'll look up my ordering specs if you need them.

You have a PM.

[Polux RSV]

Hi,

Be carrefull if you run the bike with signals like those on the pics. Your pulse is shorter which is ok. But it makes the coil fire BEFORE the original signal of the ecu. You are actually ADDING ADVANCE, which could cause damage to your motor. You should have end (firing) edge at the same position to keep the advance.

Angelo

[xlh1460]

Hi,

Be carrefull if you run the bike with signals like those on the pics. Your pulse is shorter which is ok. But it makes the coil fire BEFORE the original signal of the ecu. You are actually ADDING ADVANCE, which could cause damage to your motor. You should have end (firing) edge at the same position to keep the advance.

Angelo

Wow, you are absolutely right, nice catch!

If I turn my adjustment all the way to the other extreme, I will be retarding the timing slightly.

Hmm, so I wonder how one would make the dwell adjustable without altering the timing?

From what I have observed on my scope, the dwell on the Aprilia ECU seems to remain at a constant 7.6 ms regardless of the RPM.

I have only observed this in the garage, not on the road or on a dyno.

I am thinking it may be better and safer to just have the inteface follow the ECU signal exactly.

As long as the stock dwell is sufficient for the LS2 coils, it should work fine.

[xlh1460]

I have the PCBs assembled and bench test and ready to try on the bike.

I am going to have to wait until the weather and roads improve a bit before I road test them however.

[RossGuzzi]

I am gobsmacked with amazment !

Good luck.

I didnt understand a thing your doing, but it`s a good read

[derbybrit1]

I have the PCBs assembled and bench test and ready to try on the bike.

I am going to have to wait until the weather and roads improve a bit before I road test them however.

My suggestion to finish these up would be to put it in a little box with 2 male spade connectors for input, and a standard LS2 pigtail for an output. Plug and Play.

[xlh1460]

My suggestion to finish these up would be to put it in a little box with 2 male spade connectors for input, and a standard LS2 pigtail for an output. Plug and Play.

Good idea. I had a case already chosen when I made the boards.



This current design is actually intended to drive both coils on the one cylinder.

I am going to scrap this design though, now that I have decided against the idea of making the dwell adjustable.

The good news is that the circuit will be a bit simpler/smaller and I am going to go with a seperate interface for each coil to retain the original redundancy of the OEM setup.

I am hoping to make the circuit small enough that it will more or less just fit in a harness that goes between the OEM connections and the LS2 coils.

Lately I have been busy on figuring out how to mount the coils.

To me that is the tricky part of this project and almost caused me to throw in the towel. How sad is that?

[xlh1460]

This is what the next PCB should look like.

It should be close to actual size if I created the image properly.

This one should actually fit in the pigtail/harness between the standard coil connectors and the LS2 coil plugs.

[xlh1460]

This is a shot of one of the interface boards next to the coil connector.

It has bench tested ok so now I just need to assemble 3 more and make up the harnesses.

[derbybrit1]

I bought my pigtails pre-assembled from an eBay store (and they were in San Antonio close to AF1). Good quality and good service.
http://www.ebay.com/itm/GM-IGNITION-...item43a8834845