The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support Rix Racing
Please Support Morrison Fabrication

OEM Coil Pack Vs. COP Ignition Dwell

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

95REGF150

10+ Year Contributor
447
242
Mar 7, 2012
Elk Grove, California
Well it's time to bring back up the stock coil Vs. COP debate. I know that if you read through most threads on here about the subject you will find the general consensus to be that not only are COP setups not an upgrade but that they actually provide worse performance than the OEM coil pack due to the dwell time being fixed in the ECU.

However, I had never seen any actual data on the subject mostly just theory and opinions. Now I personally want COP simply for the more modern look of it in the engine bay over the wires. So I do not care if it performs better than stock. But I would not install something simply for looks that does not perform at least as well as stock.

So while some threads got derailed looking at total spark energy or using CDI boxes I want to focus on the one argument most made as to why a COP setup would be worse than our factory coil pack, the required dwell time. Everyone seems to think that the COP coils will require a longer dwell time because they are designed for use on a sequentially controlled engine and only fire on a cylinders compression stroke rather than both compression and exhaust strokes as in our wasted spark setups. The longer time between firing should allow for a longer dwell time between plug fires and so the coils should be designed to take advantage of that. So when they are used in wasted spark configuration they are now being short charged and not reaching the same total amperage as what we were receiving from the factory coil pack when ran at the same dwell time

At least that was the theory but again no one ever offered any evidence to back it up. So I set out to see if that was really the case. I went to the parts store and bought 4 of the 300M coils and have 2 seperate factory DSM coils in my parts bin. It's pretty easy to measure resistance with your average multimeter but that will not give you enough information to calculate dwell, you also need the inductance which is much harder to measure. Luckily I have a fancy LCR meter at work that can do this easily. Long story short here are the numbers from the two coil setups

OEM DSM Coil Packs
Coil 1 - R=1.03 Ohm L=7.254 mH
Coil 2 - R=1.04 Ohm L=7.471 mH
Coil 3 - R=0.98 Ohm L=6.997 mH
Coil 4 - R=1.03 Ohm L=7.395 mH
Average R=1.02 Ohm L=7.279 mH

300M Coils
Coil 1 - R=0.49 Ohm L=2.441 mH
Coil 2 - R=0.51 Ohm L=2.375 mH
Coil 3 - R=0.52 Ohm L=2.425 mH
Coil 4 - R=0.50 Ohm L=2.467 mH
Average R=0.51 Ohm L=2.427 mH

For the 300M coils we will wire 2 of them in series for each circuit so it will be
Averagex2 R=1.02 Ohm L=4.854 mH

Just by numbers comparison you can see that the 300M coils exactly matched the resistance of the OEM coil pack but the inductance is much lower. So now with this information we can use a formula to calculate the amperage of the coil vs the dwell time. The below charts were all done at 14V for the sake of comparison. Obviously this changes with voltage but all we are interested in here is the comparison between the two.

OEM Coil Pack:
You must be logged in to view this image or video.


300M Coils x2:
You must be logged in to view this image or video.


LRC circuits are exponential/asymptotic so since both setups have the same resistance they both approach the same amperage of about 13 Amps. Inductance though plays into the time it takes to get there and that is what we are looking at here. And you can see that actually contrary to everyone's assumptions for the same dwell time they actually end up with MORE current on the 300M coils.

Now this can be a problem if the coil is being ran at an amperage beyond what it can withstand but as many people have been using these coils for a long time I guess its safe to say they can take it. Interestingly enough I found some threads where people were increasing the dwell time by DOUBLE what the factory settings were because they assumed the coils needed it. They didn't report any blown coils so they must be pretty tough. It's just funny because it was completely unecessary.

For reference here is the factory ECU dwell table on a blackbox ECU:
You must be logged in to view this image or video.


So you can see at 14V the OEM dwell time is around 5.5 mS. So for the OEM coil pack that is just over 7 Amps. On the 300M coils we are seeing over 9 Amps so there is more energy being input to the system. And now I will make the assumption NOT a weaker spark than with the OEM coil pack. Whether that actually equals a stronger spark I can't say.

That's for smarter people than me such as @bastarddsm to explain. I tagged you Kurt because I know you were pretty vocal in some of the COP pro/con threads I read and majorly against them. What's your take on this now that we have some coil values? Am I still missing something?
 
Last edited:
Its not so much the coils themselves but what your using to drive them!

Ya I didn't really look at the power transistor with this so that's a good point. Unfortunately I don't think there is any information available on what the PT is actually rated for. But from 7-9A is only a 2A increase in current per channel. That seems within the realm of being ok.

And there has been quite a bit of field testing from those running the 300M coils for a long time. I haven't seen anyone complaining about killing the PT. Although I'm sure the guy who doubled his entire dwell table using these coils was the most likely to fry his PT.

In light of this though I'm actually thinking about dropping my dwell when I run these coils to around 3.5-3.7 mS to try and keep the current at the same 7A as factory.
 
Mike, very interesting, as I am an old timer and understands dwell. What are you going to do to drop the dwell? Is it adjustable somewhere?
 
Mike, very interesting, as I am an old timer and understands dwell. What are you going to do to drop the dwell? Is it adjustable somewhere?

I use the ECUFlash program on a blackbox ecu for my 2G. So that table I posted showing the factory settings I can edit to whatever I want and reflash. That was what I found some other people doing but I wasn't prepared to just start throwing numbers at it like they did.

I have never used Link but I would have thought that it would have some provision for setting the dwell as well. Is that not the case?
 
Not that I have found or heard of Mike. I wish it were an option, it may well be and I haven't run across it yet. Nice to know about the black box tuning though, since I have a 98 TSI with the blackbox, Tatrix Cable and the phono plug dealy. Obvious that I don't know much about those ECU's yet.
 
You can't adjust dwell via Link. I recall reading Tom, or maybe it was Dave, explain the dwell timing is too complicated to screw with, and not just a simple value, in the OEM ECU's assembly code. I could completely be misremembering.

The specs on the 300M coils that I recall reading years ago put their dwell close to our ECU's stock dwell timing which is why they seem to work out of the box, somewhere in the 6+ms range at operating voltages. However in a few years of running them with the OEM PTU and ECMLink racing I found that setup just too unreliable as the coils would just fail too often. I wondered if the dwell wasn't quite right and perhaps overcharging leading to premature failure. I gave up on them and switched back to an OEM setup for a while eliminating my failure issues. Later I switched to a setup still using ECMLink but without a PTU with some Denso coils that need dwell times in the 2-3ms range. I use a custom circuit I developed to scale the dwell adequately (and eliminate the PTU entirely, tach signal provided by the coils themselves). You wouldn't need that circuit with the black box of course. If you're asking why I did this, I simply wanted a clean engine bay, but a bit of weight loss as well, and don't think I really gained anything otherwise.

I am just throwing this data at you in case it's relevant, not really sure it is.
 
Last edited:
You can't adjust dwell via Link. I recall reading Tom, or mabe it was Dave, explain the dwell timing is too complicated to screw with, and not just a simple value, in the OEM ECU's assembly code. I could completely be misremembering.

The specs on the 300M coils that I recall reading years ago put their dwell close to our ECU's stock dwell timing which is why they seem to work out of the box, somewhere in the 6+ms range at operating voltages. However in a few years of running them with the OEM PTU and ECMLink racing I found that setup just too unreliable as the coils would just fail too often. I wondered if the dwell wasn't quite right and perhaps overcharging leading to premature failure. I gave up on them and switched back to an OEM setup for a while eliminating my failure issues. Later I switched to a setup still using ECMLink but without a PTU with some Denso coils that need dwell times in the 2-3ms range. I use a custom circuit I developed to scale the dwell adequately (and eliminate the PTU entirely, tach signal provided by the coils themselves). You wouldn't need that circuit with the black box of course. If you're asking why I did this, I simply wanted a clean engine bay, but a bit of weight loss as well, and don't think I really gained anything otherwise.

I am just throwing this data at you in case it's relevant, not really sure it is.

It's definitely relevant John thanks! Good to hear from someone who had issues running the 300M coils. And good to know overcharging them actually does cause failures overtime. Did you have PTU failures as well or just coils? That's pretty cool you came up with an external circuit to change the dwell and use Denso coils. I'm in the same boat I just want to go COP to clean up the engine bay not really looking for other gains.

Interested to learn that Link was not able to provide a way to change dwell. I know most link users run the factory coil pack so maybe it wasn't a big enough issue to spend time developing. However, I have used many other aftermarket EMS and they always have some type of dwell setting available whether it's just nominal dwell or a dwell vs. rpm table etc.
 
Just to be clear I only speculated overcharging was causing the failures and didn't confirm it with any certainty, I just couldn't think of another reason for the failures. I didn't have any PTU failures over six years of racing on the OEM PTU, just the 300M coils. It's just one data point so not even close to statistically meaningful, but with racing it's not like we have much data for the platform.
 
Suscribed!

Looking to switch to sequential/direct fire, and I have a 300M COP plate laying around.
 
Correct me if I’m wrong but running LinkECU, AEM, Haltech, etc. One could wire up coils in a sequential firing configuration without the use of a CDI box? R35, Yaris, or 300M coils for example.
 
Last edited:
@TSiBOLO all standalone ECUs have some kind of adjustment over dwell. The lack of dwell adjustment in the DSM platform is limited to ECMLink. That was a big reason why I didn't go with ECMLink for my current Eclipse.
 
I understand the dwell adjustment. But I’m referring to each coil only sparking on the top of the compression stroke.

Correct you can get full sequential with any standalone as we have a cam position sensor from the factory which gives you phase/stoke of the cycle. The factory ECU actually uses sequential for the fuel injectors but opted for wasted spark for ignition I would assume as a cost saving measure to use a traditional 2 channel coil pack as these cars were produced before COP setups were the norm.

So to do sequential spark with the factory ECU someone would have to get really fancy with Assembly Coding and get into the timing code and repurpose the EGR and purge solenoid outputs as coil ouptuts. While possible it would never be worth it for the amount of work that would take. Mainly because sequential spark is not going to gain you anything performance wise for most all the setups and builds that use the factory ECU.

That's because wasting a spark on the off cylinder that is on it's exhaust/intake stroke does not effect power at all. As long as your coils, transistor, charging system, etc. can handle firing all 4 plugs every rotation of the crankshaft which the factory system can then there is nothing to be gained. The NEED for sequential spark only comes in with very high revving motors where with wasted spark you have less time between plug fires than your Dwell time. And this is calculatable:

Coil Dwell Time: 6ms
1rev / (6ms / (1000 * 60)) = 1rev / 0.0001min = 10000 RPM

So above 10000 RPM you will not have more than 6ms between coil fires so the coil will be undercharging. Most likely this will start to happen slightly before this as if your battery voltage is lower the dwell time will be longer as you can see in my first post above.

Now if you go with a standalone and sequential spark you only fire the plug once every 2 rotations. So it doubles the above equation:

2rev / (6ms / (1000 * 60)) = 2rev / 0.0001min = 20000 RPM

Which obviously no one is going to get above. Cliffnotes version - the only time going to sequential spark will gain you anything is if you plan on running over 10k RPM. So if you are doing that it's definitely time to go with a standalone, sequential ignition, and Denso type coils which will be no problem to setup as sequential spark using the factory cam/crank sensors.
 
Last edited:
I'm just now seeing my tag in this LOL. I was mostly against COP stuff because it's a lot of money to spend for no gains. Stock works why f*** with it.... CDI on the otherhand, A dynatek always made my car run like crap, and I had to tighten gaps from what I was running. Have also read that CDI ignition on blown alcohol can cause detonation. Might just be an old wive's tail, but that what I've seen from blown alcohol guys, and a few others trying to run like an MSD on big boost methanol stuff. So, I've been on stock ignition with Taylor wires for almost a decade now. I have my coils relocated to the shock tower, I was told that a ground wire from the coils to the engine helps things. It seemed to fix some of the light throttle missfires it had with tight plug gaps.

Ok lastly. One thing that is missed in your analysis is that the stock ptu has a like 5a hard current cap. No matter the dwell it's cut off at 5a. I'm not sure how they do that, maybe a resistor? If that's the case that could be f***ing up your analysis. You could put together a simple arduino sketch to make a output waveform to drive the ptu and use the scope to watch the current rise through a small precision resistor to verify your results.

Next, isn't the spark energy go as the inductance squared? If that's the case a dsm coil has 9x more spark energy. It might not be square, might just be linear, but it's still 3x then.

Also could be insulation breakdown on the 300m coils.
 
So to do sequential spark with the factory ECU someone would have to get really fancy with Assembly Coding and get into the timing code and repurpose the EGR and purge solenoid outputs as coil ouptuts. While possible it would never be worth it for the amount of work that would take. Mainly because sequential spark is not going to gain you anything performance wise for most all the setups and builds that use the factory ECU.
It's probably possible to do it with an and type gate and another cam signal. Add another cam sensor that has a high output from like 40btdc to like 220atdc on #1. and the coil signals with this signal, and that's for 1 and 3. Then run the extra cam signal through a not gate and then and that signal with the coil signals for 4 and 2. probably have to do some funny buisiness to make it work and generate a tach signal, but totally doable. It might be possible to use some sort of logic and just pick off 1/3 and 2/4 right from the cas signals. I'm sure it's doable with an arduino, but maybe even with some discreet logic gates and no mcu...
 
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Latest posts

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top