Does anyone know the output voltage for the coil packs?

I'm a former DSM owner (95 Talon Tsi AWD, god I miss that car), and now a current 93 Miata owner. Our ignition systems are very similar. There's some talk on our forums of swapping our coilpacks with DSM packs (in fact, I've already done it, they work fine). What we can't seem to find any conclusive information on is the output voltage of the DSM packs. From my research on the Miata forums, it seems like the 1.6L Miata packs are 25-28KV. Would be nice to know if the DSM packs are a good upgrade for the forced induction guys.

Thanks!

Here's some pics of my DSM from the early 2000s:

the FSM says 40K volts for the 420a coils, it doesn't say anything about the 4g coils. i would think the 4g63 coils would be in the 40-50k range but probably closer to 50k

That seems to be the consensus, but I can't seem to find an actual source to confirm. Maybe someone here has the equipment to test them or something. I'm not comfortable screwing around with 50kv, .

That number is generally a meaningless number. If you look at the voltage rating from the manufacturer, it might be as simple as that is the max voltage the epoxy can handle, or that is the max voltage the coil will produce before shorting out(which doesn't mean you will see that voltage on the car).

The coil output is more related to the coil input. Get the dwell etc correct. Look at the inductance and dwell times to see if they are similar and hold energy well.

I'm starting to realize I didn't know that much about coil packs, . Does anyone have those specs? I'm having a hard time finding them for either car.

So ultimately, how does one determine which is a "better" coil pack? What makes COP better than oem coils (both cars have a lot of people who convert)?

BTW, it'll only output whatever voltage it needs to jump the gap.

JoeG said:

So ultimately, how does one determine which is a "better" coil pack? What makes COP better than oem coils (both cars have a lot of people who convert)?

Click to expand...

It's not simply just which is a better coil pack, but more importantly which is a better coil pack for your application.

You have to weigh the time available to charge the coils(which is car specific), the time it takes to charge the coils, and the output specs.

And people wanting to convert to COP coils from OEM is because many people don't understand what they are doing. The most common misconceptions are that you lose a significant amount of energy in the plug wires, or in firing the second spark plug with a wasted spark coil. For as many people that swap to COP on a DSM, you have ones that have found it didn't work as well as they hoped and go back to an OEM style coil pack.

Looking at the Miata forum though, it looks like there are some people that know how to properly compare and set stuff up. This thread has some measurements of the stock coils.

https://forum.miata.net/vb/showthread.php?t=316886

JoeG said:

What makes COP better than oem coils (both cars have a lot of people who convert)?

Click to expand...

I'd guess 80-90% of DSM's with COP ignitions are worse off performance wise than the OEM ignition. Without some aftermarket CDI or other device to drive the sticks it's just Rice. The factory wasted spark system works extremely well for a OEM solution.

brads said:

. The most common misconceptions are that you lose a significant amount of energy in firing the second spark plug with a wasted spark coil.

Click to expand...

That is simply not true. Firing 2 plugs probably half's the available spark energy. The current is constant in a loop, so each plug has to have the same current flow through it. Each plug has it's own voltage drop, if they are equal voltage drop the coils spark energy is halved. I'd suspect that the voltage drop across the non firing cylinder is less than the firing cylinder, but it is certainly not nonexistent.

The coil is just one part of the whole system. It is just an energy storage device. The power transistor has a bigger job than the coil, it has to sharply open the primary circuit, or else some of the stored energy will leak out in the form of back emf through the primary. (that's why arcing points in old cars have weak spark)

it all deals with Faraday's law, essentially, a big inductor opposes changes in the magnetic field it experiences, and since current in an inductor sets up a magnetic field, it opposes changes in current as well. An inductor will generate emf to drive current to stabilize the magnetic field. An inductor is an energy storage device, and the energy stored is proportional to the current squared. So if you drop suddenly drop the current in an inductor, you essentially drop the magnetic field it produces. It will use it's stored energy to create the emf necessary (up to farradys law) to keep the current/magnetic field up. Since resistance and magnetic flux leakage exist, it uses up that energy by turning it into heat, and the field and emf then collapse. If we had perfect superconductors, and 0 leakage inductors, the field would never dissapate!!

Now if we apply this to automotive transistor ignition systems like the dsm:

At the end of the dwell time, there is a big field. When you open the primary circuit, the field starts to collapse, and the primary and secondary start generating emf. Since both circuits are "open circuits", no current happens until insulation breakdown happens. Breakdown can be through air in between the spark plug gaps, or the primary side power transistor, or even through the insulation of the plug wires. Since we have good plug wires, and the secondary has 100X more turns than the primary, the plug gap gets to breakdown voltage first. The energy stored in the coil is dissapated in the form of heat at the plugs, heat from resistance in the wires (very low since low current), and then magnetic leakage from the coils.

From Faraday's law, we see that the quicker the magnetic field is dumped the higher the EMF that can be generated, thus the quicker you open the primary circuit the higher the EMF that can be generated. Keep in mind that EMF is to be thought of as open circuit voltage, and not "real current flowing voltage" this is the voltage that "could be generated", and not what "is generated". Continuing on, while you are opening the primary circuit and the current is falling, the magnetic field is leaking off too, as well as some current leakage through the primary and secondary this is all stored energy in the coil leaking off. If you don't open the primary quickly enough, the leakage can beat you, and you can get no spark, since you don't generate high enough EMF for breakdown to occur.

https://en.wikipedia.org/wiki/Faraday's_law_of_induction
https://en.wikipedia.org/wiki/Electromotive_force
http://web.iitd.ac.in/~pmvs/courses/mel713/mel713-25.ppt


So anyway the big important things on the coils are inductance, turns ratio, resistance, and efficiency.
We talked about turns ratio and inductance. Now lets talb about resistance, since we see that energy is proportional to current squared, it's essential to have a very low resistance or else the 12v will not allow enough current flow to saturate the coil - ohms law V = IR.

Lastly on coils we will talk about efficiency. Leakage is a big portion of efficiency, and for low leakage there needs to be a closed magnetic circuit. Coils like dsm coils with the laminated core wrapping around the outside are very good. Old school V8 coils are poor. A torroid would be best, but hard to do due to the high turns ratio for ignition. There are also losses due to eddy currents, as well as saturation, but that's getting hard to explain, and my hands are tired.

Saturation is basically when the iron can not make any more magnetic field regardless of current - pretty much the more iron the higher the field it can have before saturation.

To sum it up, coil output voltage is meaningless. Inductance, turns ratio, resistance, and weight are the characteristics you should care about. Oh and a sharp cut off on the power transistor.

Waste spark splits your spark energy between the 2 plugs.

bastarddsm said:

That is simply not true. Firing 2 plugs probably half's the available spark energy. The current is constant in a loop, so each plug has to have the same current flow through it. Each plug has it's own voltage drop, if they are equal voltage drop the coils spark energy is halved. I'd suspect that the voltage drop across the non firing cylinder is less than the firing cylinder, but it is certainly not nonexistent.

Click to expand...

Find the equations for the sustaining voltage. Find the equations for the breakdown voltage. You will see that the voltages are higher on the spark under pressure. And Power is Voltage x Current. So while you are correctly understanding the basics of current flow in a loop, you are missing the rest of the equation.

bastarddsm said:

Waste spark splits your spark energy between the 2 plugs.

Click to expand...

Yes, but it doesn't split it equally.

Hi, I'm a phd candidate studying engines, and I'm well aware of all of this. I have not had the need to figure out how much energy is wasted with waste spark, but just for sport here we go:

An engine is running at 30psi boost. The pressure in the non firing cylinder is probably 60psia, with a gap of 0.020" thats's 150 torr-cm. At time of spark cylinder pressure is some where around 460psia, and that's 1250torr-cm. If we look at the Pashen curve, we see breakdown voltage for n2 and 150 torr-cm is about 7500V, and at 1250torr-cm the breakdown voltage is about 45,000V. So total breakdown voltage is 52500. Non firing has 14.2% of the spark energy, and the firing one has 85.8%. So yeah it's a bit from half, but 14.2% is far from nothing.

before everyone get's their panties all up in a tizzy of 14%, remember that spark energy gain does not equal power. Spark energy will have a minor effect on the ignition delay, or what we call the "flame development period" Basically more spark energy/larger gaps slightly speeds up the very early stages of combustion. We are talking slight. Like 1* of advance kind of difference slight.

bastarddsm said:

That is simply not true. Firing 2 plugs probably half's the available spark energy.

Click to expand...

bastarddsm said:

Hi, I'm a phd candidate studying engines, and I'm well aware of all of this. So total breakdown voltage is 52500. Non firing has 14.2% of the spark energy, and the firing one has 85.8%. So yeah it's a bit from half, but 14.2% is far from nothing.

Click to expand...

Did I say nothing? NO. I said "not significant". Is 14.2% ? I don't think a PhD is necessary to figure that out. But in your comment below you basically make the point for me. No, it is not signficant. Now dropping from "probably half's(sic) the available spark energy" to 14.2% is a significant error. But what do I know? I don't have a PhD, I have to resort to my 20 years of experience designing ignitions for a living.

bastarddsm said:

before everyone get's their panties all up in a tizzy of 14%, remember that spark energy gain does not equal power. Spark energy will have a minor effect on the ignition delay, or what we call the "flame development period" Basically more spark energy/larger gaps slightly speeds up the very early stages of combustion. We are talking slight. Like 1* of advance kind of difference slight.

Click to expand...

Back to the original question. As I mentioned, if you can get a table that shows dwell time vs. current for the coils, and dwell times that your ECU provides(not an issue if your ECU can adjust it), that is important in determining coil suitability. Some of the other specs may be tougher to evaluate without more equipment, but seeing what your system can provide for dwell and energy on the new coil vs existing will help you weed out a lot of options.

You stated that we don't loose a significant amount of spark energy to the wasted spark, and clearly we do loose a significant amount - it's the same order of magnitude. And I'm not convinced it's only 14%, there is a lot more going on in the cylinder, its not N2, it's got turbulence, particulates, and all sorts of shit flying around. The pashen curves really are not relevant, but it's the best I can do without a lab setup to measure this kind of stuff.

Weather or not the amount lost matters to power production doesn't effect the fact that it is a significant amount. So shoot me I shot from the hip and guessed that it was PROBABLY 50% lost, and a quick estimate shows really it's probably more like 14% Still better than saying zero or close to it, which is what insignificant implies. I mean if I sell something for $100, and the guy tries to pay me $85, I'm sure as hell going to notice..

If you really meant that it was "some but not very much", shouldn't you have mentioned that? Perhaps explained it a bit? Are you not a wiseman? An ignition designer? Shouldn't you have an enlightened understanding that would allow you to provide a better explanation?


You keep talking about designing ignitions for 20yrs, and I've asked for examples of what you have done, test setups, technical details, anything really - Nothing back. As the ignition expert, why didn't you actually do the calculations? As the expert I would think you should be showing guys how to measure what dwell time is needed. Oscilloscopes are dirt cheap now days, just need a function generator or some ingenuity. I guess we should just refer to the nonexistent data on the junkyard coils people are looking for data on.

bastarddsm said:

You stated that we don't loose a significant amount of spark energy to the wasted spark, and clearly we do loose a significant amount - it's the same order of magnitude. And I'm not convinced it's only 14%, there is a lot more going on in the cylinder, its not N2, it's got turbulence, particulates, and all sorts of sh** flying around. The pashen curves really are not relevant, but it's the best I can do without a lab setup to measure this kind of stuff.

Click to expand...

I find it hard to believe that a PhD candidate can't calculate all of that, or model it? What do you need real world stuff for?

bastarddsm said:

Weather or not the amount lost matters to power production doesn't effect the fact that it is a significant amount. So shoot me I shot from the hip and guessed that it was PROBABLY 50% lost, and a quick estimate shows really it's probably more like 14% Still better than saying zero or close to it, which is what insignificant implies.

Click to expand...

"Probably 50%". Pointing out the OBVIOUS BASIC error in what is freshman level electrical theory, had you immediately back down that number by a very significant value.

bastarddsm said:

If you really meant that it was "some but not very much", shouldn't you have mentioned that? Perhaps explained it a bit? Are you not a wiseman? An ignition designer? Shouldn't you have an enlightened understanding that would allow you to provide a better explanation?

Click to expand...

It was an adequate explanation. It's going to vary, but it is always relatively small, especially compared to the losses from mismatching the coil and application.

bastarddsm said:

You keep talking about designing ignitions for 20yrs, and I've asked for examples of what you have done, test setups, technical details, anything really - Nothing back. As the ignition expert, why didn't you actually do the calculations? As the expert I would think you should be showing guys how to measure what dwell time is needed. Oscilloscopes are dirt cheap now days, just need a function generator or some ingenuity. I guess we should just refer to the nonexistent data on the junkyard coils people are looking for data on.

Click to expand...

Why do I need to do the calculations? I have known for a long time that a wasted spark puts most energy into the cylinder under compression. It seems like it is commonly known to people outside of the PhD in combustion world.

If someone had some of the equipment and asked, I would give them some pointers. But the best thing I could do for most people is design an ignition that simplifies the dwell, makes it so most any coil will work, and plugs in. Oddly enough, there is a CDI on the market like that, designed about 15 years ago. Coincidence?

And to answer your next question, no I never worked for Dynotech.

Wow, there's a lot of good information here. Also, not that it really matters since voltage isn't the deciding factor, but I got confirmation back from Delphi this morning, the DSM coilpacks produce 40KV.

Is Delphi producing those? I'm pretty sure there was another OEM supplier originally for the turbo one. Diamond Electric.

You really have to be careful with quoted voltage specs from the oem. If they have you a spec sheet showing testing method, it may be a legit number. But more likely it is a non technical person answering email with numbers they are getting off sales brochures.

The OE should be capable of that 40k voltage, but I'm pretty sure the last time I looked at one of the datasheets, they were closer to 30KV measured from their test. It has been a number of years since I looked at it though. The datasheets don't tell you a lot from what I recall.testing in system to measure charge current vs dwell time is the way to go.

The delphi ones were the OE replacement I bought, not the originals.

brads said:

That number is generally a meaningless number. If you look at the voltage rating from the manufacturer, it might be as simple as that is the max voltage the epoxy can handle, or that is the max voltage the coil will produce before shorting out(which doesn't mean you will see that voltage on the car).

The coil output is more related to the coil input. Get the dwell etc correct. Look at the inductance and dwell times to see if they are similar and hold energy well.

Click to expand...

Dear Mr. Brads

We can measure the transformation ratio of a coil by applying for example 5 volts AC 50 or 60 Herts in the primary of the coil. If it is 40 Kv, Turns ratio 83:1.
So 5 x 83 = 415 Vac in secondary
Why 83:1?
480V x 83 = 40KV
Why 480V
Because when the power transistor turns off, it generates 480 V in the coil's primary,
480 x 83 = 40 Kv.
Unfortunately, I don't have an original DSM coil to measure.

That test assumes the coil primary stores 480V, which is a very big assumption. That would probably only be a safe assumption if you were running all the different coils on a testbench where you controlled the voltage and dwell, or were simply feeding in an AC signal. The kV rating that you get from the test just doesn't mean too much in terms of ignition performance in vehicle.

When I tested coil input/output voltage like that years ago, to check the turns ratio, I would usually follow it up with an inductance test. That is another way to calculate the turns ratio and perform a sanity check on the numbers.