melted an IGN-1A
 

Finally got some sunny days and decided to go out and work on the FD again. It's been sitting for the last year. Got the battery charged and tried starting and it would try and catch but no luck. When I got out and looked back under the hood I saw smoke from one of the coils. I have upgraded all my grounds and am running the Sakebomb IGN-1A kit with battery in the stock location and the ground from the coils straight to the battery. I am also running a Adaptronic PNP with the coils wired for waste spark. For some reason one of the rear coils melted down. When I inspected the coils both front coils were cool to the touch, but both L2 and T2 were too hot to touch for any significant time. Any ideas why it would only be the rear coils?

There are two grounds. One for the battery / chassis, and one for the ECU ground. Without the ECU ground connected, you will melt coils.

As it comes, one ground is labeled battery ground and the other is engine ground. Advanced install does mention using ECU Reference, specifically tapping into the MAP sensor ground. Should this give me a more secure setup? I can do this fairly easily.

My experience with SBG products is that they are overpriced and half-assed. With that said, the problem here is your EMS or their harness.

seen it happen for others, and then the replacement was fine

but you can buy the coils direct from the source; Lance Nist @ Pantera EFI, and be 100% guaranteed they’re legit, and usually the best price too

I’ll look for his info …

edit: sent you a pm
.

Yes, do that. I melted 5 coils using the basic grounding method. Then I looked up the coils specs, and they specifically tell you to ground to the ECU ground for one of the coil pins.

I'm still not sure why SB suggests any other way

^Why would you run your IGN1A's wired for waste spark? Presuming you have 4x of them, they should be run in direct spark mode. Besides, the IGN1A only has 1 spark plug wire terminal, and the purpose of waste spark is to fire 2 rotors from a single coil (L1 & L2)

It was just trying not to change too much on the initial connection. I am changing it to direct fire at the same time that I update the wiring for the ground

What are you guys talking about? There are THREE grounds and they NEED to be ran properly. The Sakebomb kit is known to be incorrect. (Instead of cylinder head, read: rotor housing.)
https://cimg8.ibsrv.net/gimg/www.rx7...f5bd178266.jpg

I am pulling the harness and going to update it with a ground to the battery, one to each rotor housing, and one to a sensor ground from the ecu. Are all sensor grounds from the ECU viable?

Most ECUs have one or maybe two sensor grounds which all sensors share. Just use whatever ground is ran to the MAP or TPS.

https://cimg3.ibsrv.net/gimg/www.rx7...274dd257e7.png


Some more food for conversation, this was a post on the Haltech tuning group i saved a while back doing some research on how to wire in the IGN1a coils. This surely cannot be accurate, despite claudio have a TON of wire harness building experience. I understand star grounding to the block, and sending both the "cylinder head" and battery ground to the block (effectively zero offset grounding), but the signal ground should be going through the ecu, correct? Anyone else weigh in on this before I wire mine incorrectly? lol

You do not want zero offset grounding that is the whole issue here. Battery, sensor, housing. Just follow AEMs install instructions... the ones I posted. Before grounding my coils correctly I had major sync loss issues and it cost me a front iron before figuring out the correct way to wire these. Also a fried coil.

I want to also add that grounding any sensor to the block is a terrible idea. There is a reason aftermarket computers have a dedicated ground for the sensors. An engine block ground is extremely noisy. Grounding the sensor ground of the coils to the ECU sensor ground assure that the sensors and coils are sharing the same noise free ground. Assuring that the ECU is getting noise free reliable data. The old saying.. garbage in equals garbage out.

What is really concerning is that Claudio is the owner of TuningTechnology, and he does a TON of harnesses. I'm not sure if he was mistaken in what he was trying to say or if he's been doing this incorrectly for some time?

EDIT:

Found another thread on the haltech tuning page. This gentleman emailed haltech direct for support and this was their response:

https://cimg7.ibsrv.net/gimg/www.rx7...ea86debd58.jpg


What is the definitive answer on how these are supposed to be hooked up? Because i keep seeing conflicting messages...

Second post to follow up, direct from the Haltech page:
https://support.haltech.com/portal/e...ignition-coils

12V Power Supply

The IGN-1A coil can draw up to 15A. If high rpm and high power is being used, together with conventional 30A Relays to provide power, it may be necessary to have a separate fuse and relay for every pair of coils. This has been seen on some rotary installations where over 10,000 rpm is used together with high boost.


If the wiring is also not cable of supplying the current required it can also limit the spark output.

Grounding

The number of grounds (3x) on this coil often cause confusion. If in doubt, all grounds can be joined together and run the to cylinder head. This can eliminate issues if there is a poor engine ground strap that then pushes high current through the coil. It can also prevent a coil from firing if the engine grounding is not sufficient, and this happens as the ground lifts on the ECU signal preventing triggering. The fix is of course to improve the engine grounding.

Haltech did not design these coils, these are rebranded Mercury Marine coils. AEM were the first redistributors of these on the market and they would not have come up with those installation instructions without guidance. What Haltech thinks about it is incorrect and not relevant as they did not have any hand in creating them. I believe Lance Nist at Pantera EFI was the engineer behind these as a previous poster in this threat alluded to. If you dig you can find a thread and many others on this exact topic.

Ground them per Mercury Marine/Lance Nist's specifications and you'll never have an issue. Ground them per Haltech and Claudio's suggestions and you'll be fine 95% of the time, however when things go wrong, they go wrong badly and are an annoying issue to deal with that is difficult to diagnose unless you actually know what you're doing.

Also, no IGN1As do not pull 15A in common and proper use. This is more poor information that while technically correct in a vacuum just plain isn't the case when deployed on a vehicle and doubly so on a rotary. Since the IGN1As draw ~7.5 amps at peak saturation and are not a constant draw/charge average current will be roughly half of peak, then you'll need to factor in actual duty cycle. After that you'll know what kind of current capacity you need to accommodate for which isn't 15A a coil.

Also here is how you'd find duty cycle on a rotary, plug in whatever dwell you're interested in using, here is a basic (over the top) 5ms example:

For an 8000 RPM target and 5ms dwell:

Rev/sec = RPM / 60 = 8000 / 60 = 133.3
sec/cycle = 1 / cycle/sec = 1 /133.3 = .0075 = 7.5MS
Coil dwell/charge time = 5ms
Duty cycle @ 8000 RPM = On time /Total Time = 4 / 7.5 = .66

Lots of professional tuners have done things wrong for years, sometimes it works, sometimes it doesn't. Best and really the only CORRECT advice here is to follow the coil manufacturer's instructions for wiring - they specify 3 different grounds for a reason. Haltech does not make these coils, so they wouldn't be my go-to source for authoritative information. Here's my reply to another post elsewhere on the forum regarding IGN 1A wiring that tries to explain why they have 3 separate grounds:

just have @C. Ludwig make you a coil harness and you won't have any more issues

I disconnected the harness and this connector had quite a bit of moisture in it. Enough that I had drops splash out when I disconnected it. I would have thought this would be a weather tight connector. Any suggestions or ideas?


https://cimg5.ibsrv.net/gimg/www.rx7...42f9a8566c.jpg
https://cimg6.ibsrv.net/gimg/www.rx7...2207dc8e01.jpg

I’ve spent a lot of time on the phone with Lance over the years. Give him a call. He loves to talk. Plan your day accordingly.

The coils are designed for an OE application that utilizes misfire detection. The multiple grounds are part of that feature. Since none of us in the aftermarket are making use of it, the two low level grounds, pins B and C, should not be connected to ECU reference ground. They should be connected to the head or rotor housing. Connection of pin B to ECU reference ground could, in the event of coil failure, allow the coil to pull high current ground through the ECU causing ECU damage. This is the primary reason to not connect pin B to ECU reference ground. Combine that with the fact that the function that the coil reference ground serves to non existent in the aftermarket and there is not a good reason to connect pin B to ECU reference ground. The OP’s coil melt down wasn’t caused by pin B not being connected to ECU reference ground and won’t be solved by doing so. Claudio, myself, and many highly experienced builders don’t use ECU reference ground for these coils. The only place many of us differ is whether we take pin D to the head/rotor housing or the battery.

I’ve said it for years, decades now, AEM doesn’t know anything about the coil they rebranded. The diagram they use is from Lance and depicts the OE configuration.

The coil harness connector should be weather tight but it’s also 30+ years old and has lived a hard life in one of the harshest engine bays ever created. Water ingress isn’t surprising. New connectors are available. Dielectric grease also goes a long way to keeping water out.

Without being hands on, we’re all guessing at the cause of the failure. I’d start with checking to make sure the ECU is properly configured for the coil configuration and toning out the harnesses to make sure there are no shorts and the configuration is correct to suit the coil configuration. If you need help, feel free to contact me directly.

Thank you @C. Ludwig for the great explanation. Can you expand on this a little, at least from your preference? I assume if you have good star grounding that you could use either the block or the battery direct, yes?

I prefer the battery but I’m probably in the minority. Lance has told me to use the battery. There are guys that have great success taking all three to the head/rotor housing.

When I build harnesses, I take the ECU and ignition power ground directly to the battery with separate circuits. That way I’m not relying on someone else’s work to know I have 0v and proper amp capacity for those two critical circuits. I know the argument is that this creates a ground loop with the ignition but I haven’t seen an issue arise. Taking all three to the head/rotor housing would eliminate that ground loop and again, I’ve seen a lot of people have success with that.

Thank you for such a great response. I won't mess with grounding to the sensor then. Should I have the engine ground split to rotor 1 and rotor 2 instead of just a single engine ground? Also, can someone provide a reference or name of the connector I am looking for (Or a better connector if I am depinning and replacing anyway). I am comfortable depinning and putting the pins into a new connector (as well as using dialectric grease) to try and make sure I don't have water ingress being a potential issue.

good luck telling them that.

I ended up buying an SB RX8 kit from someone for 20 cents on the dollar that the original purchaser couldn’t get to run correctly from day 1 (misfires) and SB refused to address it. Pinning out the harness revealed the problems and correcting them solved the issue, but SB still swears it couldn’t have come from them that way.

I can only verify that the harness had issues and ran fine after being fixed.

my position has always been to wire it the way the diagram shows, glad to see a few people who disagreed before have come around to seeing it different now.
.