Home > Projects > Megasquirt

Coil Harness | First Prototype | Second Prototype | Electrical Interference | Second Coil Harness Attempt | Ignition Control Box

A coil on plug (COP) ignition setup has been on my list of things to do for a couple years now.  I first got the idea for this from Tim's Megasquirt Project.  He figured out that you could use the coils from a late model BMW and drive them in wasted spark configuration from the Ford EDIS ignition control module.  Details of this are on Tim's web site.

My intended use of a COP ignition was to directly control the coils from my Megasquirt controller instead of the EDIS controller.  I intended to wait until I could run fully sequential injection and directly fire the coils independently so this project was a little ways down on my list.

In early September I was contacted by Tom from www.02again.com and asked if I wanted to work with him on developing a mounting bracket to allow a similar COP setup (he had also seen Tim's setup).  I had not planned on starting this project but when presented with an opportunity like this, I could not refuse.  Over the next 4-6 weeks we developed the concept from an initial prototype that Tom had already fabricated into a final design that works very well on the M10 engine.

This page details the development of this COP setup from initial prototype through to final parts.  The coil mounting bracket is a future product from 02again and should be available in the near future.

Here is the original prototype coil mount bracket that Tom sent me to try out.

First prototype coil holder as I received it from Tom.	First prototype test fit on a spare head I have.

 

Coil Harness

Before I could convert my wasted spark coil pack ignition to COP I needed to fabricate a new ignition harness.  I had already planned on this option a couple winters ago when I did my Megasquirt ignition upgrade project.  During that project I had installed four high current coil drivers into my Megasquirt controller and had modified the engine harness to accept different ignition coil harness configurations.

The coils used for this project are used on almost all late model BMWs from about 2002 onward.  There are two BMW part numbers for two different coils that will work:

• 12 13 1 712 219
• 12 13 1 712 223

These coils are available from most auto parts stores for anywhere from $25 to $65 each so you will want to shop around for the best price.  I cross referenced the part numbers and used a 2004 BMW 330i sedan as the car for the parts store to reference.

Coil packs used for COP conversion are from a late model (post 2002) car. BMW part number 12131712219.	Close-up of the coil's mating connector installed.

I fabricated a wire holder for the ignition harness out of some brass tubing.  This kind of tubing is normally found at hobby shops or arts and crafts stores.  I soldered the holder together using regular plumber's solder and flux.  Here is a drawing of this holder.

I ordered a set of mating connectors and pigtails from my BMW dealer.  The BMW part numbers I used are:

• 12 52 1 437 985 - Connector housing (4 required)
• 12 52 0 007 170 - Socket pigtail 0.2 - 0.5mm (4 required)
• 12 52 0 007 172 - Socket pigtail 0.5 - 1.0mm (8 required)

The cost of these parts was about $50.  A full wiring diagram for my coil harness is here.  I used the thinner gauge wires for the ground connection on pin4a of the coils.  The heavier gauge wires were used for the coil (+) and coil (-) connections.  The coil ground wires were terminated to a ring lug that is attached to the coil mounting bracket.

Ignition harness wire holder frabricated out of brass tubing.	Wire holder cleaned up and ready for paint.	Painted wire holder.
Coil connectors and pigtails ordered from BMW.	Assembled coil connectors with a length of 3/8 inch convoluted tubing and heat-shrink.	Connectors and pigtails installed in wire holder.

 

First Prototype Installed

Once I had the harness modified, the COP setup with the first prototype bracket when together pretty easily.  A few issues were identified.  The most significant issue was that the runners from my header extended out from the block fat enough to be located directly under the tops of the coils and extended out beyond the heat shield of the exhaust gasket.  Within a few minutes of idling the engine, the coils were too hot to touch.

First prototype and coil harness installed on engine.	View of heat shield gasket and first prototype bracket.	Tops of coils exposed to direct heating from the header.

 

Issues Identified With First Prototype

I fabricated a simple extension to the heat shield out of some sheet aluminum and attached this extension to the prototype bracket.  This solved the problem with cooking the coils.  I was able to drive the car for over an hour and pop the hood and grab the coils without any difficulty.

Paper template for a heat shield extension to shield the tops of the coils from the header.	Heat shield extension made out of some sheet aluminum.
Heat shield extension added to first prototype coil bracket.	Bottom view of extended heat shield.

Another minor issue with the first prototype was that the bracket needed to extend out about another 1/8 inch to allow the coils to properly seat in the bracket.

Coils did not fully seat into first prototype bracket.	Close-up of coil fit on first prototype. The rubber top of the coil does not completely seat into the bracket.

Second Prototype

I provided all this feedback to Tom and after a couple weeks he had a second prototype out to me with all the improvements I identified.  He integrated the extended heat shield into the bracket design and adjusted the bracket dimensions to correct the coil fitment issues.

Second prototype bracket with heat shield extension integrated into bracket.	Bottom view of integrated heat shield on second prototype.	Coil fit is much better on second prototype.

EMI Problems

The most significant problem I encountered ended up being electrical.  I had mounted my high current coil drivers inside the Megasquirt controller's case and I think this was a mistake.  The electrical noise generated by the four high current drivers as well as the electrical noise transmitted back from the coils into the controller's case cause significant electrical noise problems.  When I initially started driving with the COP setup, I lost my serial communications and experienced several lost sync events and even a couple reboots of the controller.  The noise was so bad that my TechEdge wideband O2 controller crashed and needed to be re-flashed to recover.

I identified the electrical noise problem by switching my car radio over to AM and was immediately able to hear the noise which changed pitch with engine RPM (ignition noise).

I addressed this noise problem with several modifications to my ignition harness.  These changes were:

• Add a shield to the coil (-) wires entering the Megasquirt controller
• Add a noise suppression capacitor to the coil (+) connections
• Add ferrite beads to the coil (-) wires entering the Megasquirt controller

The combination of these changes have eliminated the electrical noise problems.  (My AM radio reception is also fixed...)

I cut into the harness and added the silver ground braid to the wires that enter the car. I also added a positive lead for the noise filter capacitor.	The black wire that goes to the ground braid and the noise filter capacitor connected to the positive leads is shown here.

Second Attempt At A Coil Harness

Ok, does anyone see the HUGE mistake I made in the design of my first coil harness?  Here's a hint, do not use thin walled METAL tubing to run your wires if you cannot fuller de-burr the inside of the tubes.  I had sharp edges at each tube junction where the individual coil tubes were soldered to the main tube.  I managed to chafe the insulation on several wires, one of the wires had the insulation cut through to the conductor.  Fortunately I found the problem while the car was in the garage and not driving.  I'm also fortunate that I did not burn up anything when the wire in question intermittently shorted out to the brass tube.

The way I fixed this was to create a new harness based on the BMW design.  I purchased an E36 ignition harness off of ebay and cut two of the 6 cylinder's coil connections off of it.  A little plastic fabrication work to close up the end and it looks almost like a factory piece.  Most importantly, it is made of non-conductive plastic this time.  The factory rubber grommets and wire covers also work much better than the convoluted tubing and heat shrink tubing used on the first harness.

Here is the complete process showing how I modified the E36 piece.  I still need to fabricate some simple brackets to tie this piece to my COP bracket.

The E36 ignition harness.	Carefully remove the cover.	Remove the wiring. All of the rubber grommets and wire covers will be used except for two of the coil connector wire covers.	The E36 connectors are not compatible with the coils used. The newer coil connectors are slightly smaller than the E36 connectors and the boots don't quite fit.
A small wire tie pulls the rubber boot into the connector for a tite fit.	This resistor was used on the E36 for the coil ground return. I will try using it to see if it reduces ignition noise.	Test fit on my engine to make sure the cylinder spacing on the ignition harness will work.	I cut the wire holder between the main harness junction and the two end coil wires.
I then cut the end wall from the removed piece to glue it back into place.	End wall epoxied into place on the modified wire holder.	Same thing was done for the cover portion of the wire holder.	New connectors installed and connected to the harness.
Cover installed. This almost looks like a factory piece.	Close-up of the end showing the modification.	New coil harness test fit on engine.

 

Ignition Control Module

I was having a lot of problems with ignition noise causing problems with my Megasquirt controller.  I suspected that these problems were caused by the fact that I had located the high current ignition drivers inside the MS controller case.  This choice resulted in a direct connection between my ignition coils and the inside of the MS controller.  Not a good choice on my part.

I resolved this problem by fabricating a stand-alone ignition driver box to house the ignition drivers that could be located in the engine compartment.  I used a combination of a water resistant aluminum box and military-style connectors.

I decided to build two copies of this box.  If an ignition driver ever died on my, I would be in serious trouble.  The spare box was tested on the engine and then added to my spare parts box in the trunk.

I used a Hammond model 1590WSFL water resistent enclosure.	Here is one of the Hammond enclosures drilled for the military style connectors.	A 4-coil driver board from Glenn's Garage.
One of the coil driver boxes fully assembled. The coil harness goes to one connector, the other goes the the Megasquirt controller.	Why build just one when you can have two at twice the price?	Two drivers built up. The spare goes in the trunk.
An aluminum mounting bracket for the coil driver box.	Coil driver box mounted behind passenger side headlight.	Final connections made to new ignition driver box.

 

What's Next

The second prototype coil mounting bracket appears to be the final design.  I'm waiting for a sample of a production bracket to verify.

I have been driving with various versions of the ignition for the past month.  I think the engine is running a little smoother than it was with my wasted spark setup.  This might be explained by a plug wire that might have been going out on me.