MSD Coil Ballast Resistor Needed?
 

I was about to replace the ignition coil with an MSD high vibration coil on my 86 3.2 Carrera with Bosch DME.

As you know the instructions say you may need to install a ballast resistor.

Does anyone know if a ballast resistor is needed for my car? (or has anyone installed an MSD in a Carrera 3.2?

Thanks!

Ballast resistors, high wattage low ohm resistors in series between the source voltage and the coil, are only required with an actual Kettering ignition system, points or solid state.

They are used to prevent overcharging, saturating, (think HEAT the coil at low RPM.

Your '86 has a modern day implementation of the Kettering system that uses PWM, Pulse Width Modulation, to prevent coil overcharging. Superior to most CDI systems, coil is charged prior to ignition time, not at the point of ignition, also results in wider, hotter spark vs CDI

You might be better off with your original coil unless there is something wrong with it.

Thanks!
I put the MSD in without a ballast and so far no problems. I did have to put a little bit of rubber strip in the holder to secure it because the Bosch has a bigger diameter than the MSD.

Anyone got any comments about this?

According to this guy A lucky eBay find: NOS SWB Coil the DME Bosch coil is 4 mHenries and the MSD is 8.
For say a 4 millisecond dwell wouldn't the energy in the coil be half for the coil with 2 X the inductance (not even considering the primary resistance)?

Does it matter?

I'm sorry I can't resist

What?

What?

Hi Rick. I am "That Guy." Does the inductance matter? Hell yes it matters.

To the original poster: Find yourself a good used original Bosch coil for a 3,2 Carrera and don't go anywhere the MSD, the specs are wrong. Last time I checked a control unit for a Motronic car was about 1800 bucks, it's just not worth it to stress the coil drivers.

And West before you comment please read this thread:

http://forums.pelicanparts.com/porsche-911-technical-forum/640218-crane-xr700-capacitive-discharge.html

Thanks for posting the link to the Crane thread, as post number 3 there links to a
website for coil values like your post on the other forum but with more info;
Diagnostics

Just to summarize, I would think the consensus is:

Using the MSD Coil

• Best Case is it works but with degraded ignition performance over the stock coil tailored for the DME (weaker spark)
• Worst Case It saturates because the core will not support the current (hinted at by the ballast resistor comment) and stresses the driver (current limiter) in the DME.

There is no published data, only what individuals have measured with their LCR meters and I haven't found any data on what current the coils will support.

If anyone has a 10 Amp current probe I would really like to see the results. ;)

What?

With a standard Kettering system the dwell TIME varies with RPM, idle dwell TIME is long enough that the coil would be over-charged, saturated, absent the current limiting (BALLAST) resistor.

Our DME Porsches use a FIXED dwell time ("OFF" time varies) that at idle is only long enough to fully charge the coil, PROPER, OEM coil.

What?

With the inductive system the coil can be FULLY PRE-CHARGED, to a point just short of core saturation. With CDI the plug typically fires long before the coil becomes fully charged, resulting in lower spark energy.

Didn't see any conflict....

I would not replace the stock coil. The DME has a built in dwell map that is matched to the stock coil. You will likely dwell that MSD coil for to long a period of time at which point the internal circuitry for the coil driver within the DME will go into over current protection mode but it's not designed to work in that mode. Over current protection is only a safety feature. I don't advice you use that MSD coil it could easily result in failed ignition drive circuit as well as a toasty MSD coil. If the dwell map is properly matched to the coil you don't need a ballast resistor.

It is possible to properly calibrate a chip for use with the MSD coil, I've done such calibrations to the dwell map but it's not easy to do and requires a special chip.

The stock coil has a optimal dwell time to reach saturation of about 3.8ms and it hits that number right up to about 5000RPMs where each crank turn takes 12milliseconds and you have 3 spark events per crank rev so this gives you about 4ms max dwell time available at 5000RPMs.

The way the dwell map works is that up to 5000RPMs it hits the dwell right on the head at about 3.8ms after 5000RPMs dwell becomes a compromise as the dwell time gets shorter and shorter as RPMs head north of 5000RPMs.

Basically above 5000RPMs it dwells the coil based on crank degrees and does so at about 110 degrees of crank rotation. I'm not speculating here, I know the code within the DME and how it works along with the dwell map.

Bottom line is that below 5K RPM it dwells to optimal saturation above 5K it dwells for as much time permissible to fit 3 spark events into the crank rev.

In a system such as the DME that tightly controls dwell time it's not a good idea to just guess at a coil's optimal dwell time. If you find a coil with a saturation time of 3.8ms then go ahead and try it but if you have no idea what the dwell time for the coil is your taking some risks or at best just compromising the ignition system.

It was also mentioned that in the DME system it's the coil off time that varies, that's true till you run out of off time and once that happens then on time starts to vary. This occurs at about 5000 RPMs.

The 3.8ms time is assuming that system voltage is at about 14.0vdc if the voltage drops significantly below this then the dwell time goes up by a bit.

I was wondering about the Pulse Width Modulation (PWM) you were talking about. Your definition of PWM must be anything that turns on and off regardless of period or duration. BY that loose definition points and a condenser would be PWM. The DME turns the driver on to charge the coil and then turns it off to fire the plug. If the current in the coil exceeds a set point there is an analog circuit that rolls off the charge voltage to the coil. There is no chopping or averaging occurring.

With a CDI ignition the energy is stored in the capacitor and the coil acts as a transformer and the energy is immediately transferred through it. The magnetization inductance is an undesired parasitic which is why a CDI coil is designed differently from an inductive discharge coil as pointed out in the last thread. Charging the mutual inductance draws energy away from the spark. Weaker spark? Compare apples to apples. I'm sure a CDI could be designed that would equal the inductive system but at higher cost.

"..immediately transferred..." No, the magnetic field builds "slowly" in accordance with inductive reactance... The secondary voltage then rises in synchronization(?) with the rising magnetic field.

On the other hand with an solid state (no condenser) inductive system the secondary VOLTAGE will rise virtually instantly to the level required to fire the plug.

The reason an inductive ignition coil can be so readily used with a CDI system is because 300 volts will overcome the coil's inductive reactance quite quickly in comparison.

I think of PWM vs analog/linear control functionality.

I think of PWM as an ON/OFF control function with either the "on" time or the "off" time varying in accordance circuit load demands. Our DME ignition varies the "off" time in accordance with RPM.

So no, the Kettering system is not PWM.

As indicated in the link of post #8, the MSD Blaster primary inductance is about 4-4.5 mh,
whereas the Bosch coil used in the 911 3.2 has an inductance of about 3.6 mh. So using
the MSD Blaster will not damage the 3.2's DME ECU, but at higher RPM the MSD coil will
produce a weaker spark. So there's really no benefit to using the MSD unless one no longer
has the original black Bosch coil and is forced to use the unreliable silver Bosch coil.

I know very little about the construction of these automotive coils but will the MSD coil, designed to be used as a transformer, support the current levels here? There is a difference between transferring a pulse of energy through the core and storing all the energy in the core. It has to be kind of close because in one (inductor) application they want you to externally limit the current with a resistor.

For a power supply example a transformer would have a core made of continuous magnetic material whereas a coupled inductor (like a fly back) would have a core with an air gap to store energy.

If you read the specs of the MSD Blaster coil on MSD's website, you'll find that it's not
designed exclusively for a CDI application. Furthermore, again using the data from the
link in post #8, you'll find that the typical CDI only type of coil has primary inductance
values less than 1 mh, which is what one would expect since no energy is being stored
in a CDI type of coil.

I think I see a problem with your transformer

I don't have any specs for the magnetic materials in any of the coils especially that showing the part of the B-H curve where the inductance (resistance to change) falls off.

It still might be a good idea to stick with the original coil.

Even with coil specs in hand I always bench test a coil for dwell times and across various system voltages. Here's a video of a crude bench test setup I use from time to time, it gives you a good idea of how a coils optimal dwell time can be found on a bench:
https://www.youtube.com/watch?v=a7dFdIdsZeU
Often I've found that coil specs are off and the actual saturation time differs on the bench than the time calculated mathematically. The very best test is the bench test.

What you need to do with the MSD coil is test it the same way I tested the 964 coil dwell and compare the results. Worth noting, the 964 coils are the same as the 3.2L coil.

The 3.2L (and 964) coil dwell times are as follows:

12vdc - 5ms
14vdc - 4ms
16vdc - 3ms
20vdc - 2ms

I suspect the MSD coil results will have longer dwell times.
The best way to know is to bench test a coil.

As seen system voltage impacts dwell time considerably, at 14v you can no longer fully charge the coil above 5000RPMs but at 12v this drops to 4000RPMs, just a 2volt system voltage drop effects the coil optimal dwell considerably! And at 16v your good till 6600RPMs

"..I suspect the MSD coil results will have longer dwell times..."

Higher inductance value = REQUIRES longer dwell times to reach a specific charge level.

The worse conditions for firing the spark plug is at WOT, lower RPM range, and a HIGH engine load, RAPID acceleration. So not fully charging a coil at maximum engine RPM is not necessarily a design flaw.

And that is..???

But peak cyl pressures in our 3.2L cars happen at WOT 5400-6000RPMs (peak torque) and even the stock design is a compromise here if the system voltage is below 15v. Putting in a coil that needs more time to dwell is a bad idea as it makes this problem worse.

I won't argue if a coil should be fully charged or not but I will say that the stock setup fully charges the coil so long as time permits. That's what the engineers designed in the 84-89 DME. I do agree that some other EFI system play with dwell times to reduce charge for low cyl pressure conditions but the 3.2L does not do this.

This is an excellent source for dwell and testing for optimal dwell time:
Ignition Coil Dwell Calibration
You really need to bench test, don't trust a published mH number!

Even after bench testing you need to double check with coil in car because the DME (or EFI system) could limit current somewhat from what you tested on the bench. The in car conditions and the in car electronics can cause some differences. In the perfect setup you'd do in car testing but it's difficult (not impossible) to vary system voltage in car with engine running.

Consider how little % the coil charge level changes beyond 3 time constants.SmileWavy

You strove for coil saturation for determining full charge, how much does that differ from the 5 time constants it takes to reach a full charge, or not?

Will require longer dwell times!

The DME ECU doesn't know what ignition coil is being used and as such it uses its predetermined
dwell time based on the battery voltage and an assumed coil inductance. If the ECU were
actually monitoring the coil current (driver current), then the ECU could adjust the dwell
time based on a desired current.

Remember:

I (dwell current) = T (ECU dwell time) X V (battery voltage) / L (inductance - unknown for a non-stock coil)

So since the ECU doesn't know the inductance, the actual dwell current is unknown.

My comment "I suspect the MSD coil results will have longer dwell times." was referring to bench test results to find optimal dwell time, I was not referring to in car dwell times.

But 100% agree that the actual in car dwell time will not change since it's baked into the DME via a dwell map. I've created new maps for different coils over the years, my advice is stick with the stock coil, it's fine for the naturally aspirated 3.2L the engineers did know what they where doing :)

this
voltage appears on the secondary immediately.

In the context of a transformer being used to convey energy from the primary winding to the secondary winding the rising magnetic field must "cut", pass through, the secondary winding in order for the secondary voltage to rise in accordance with primary voltage applied.

The more input inductance the transformer has the more the secondary voltage will lag the primary's applied voltage.

ELI the ICE man.....

Speaking here of TRANSFORMER action, not inductive ignition coil action.

Right on!

Sal

Way off topic, but I watched your video on RhodeMap RI......scary stuff.

As a former Rhode Islander I am very interested......good analysis on your part.

Dick

So the transformer equation does not apply to your transformer.
Look at Faraday's law for both windings (load on the secondary) and note the shared total flux appears in both equations and can be eliminated (the dt part). I think that is how you derive the transformer equation.

Input inductance is another of those standard terms in a transformer model. Is this leakage inductance, magnetization inductance, maybe all of the above?

If you apply a Sine wave to the primary with an unloaded secondary the input current will lag the voltage (90 deg) but with a load the real part dominates and the phase shift of the current goes to zero.

Dick,

Yes, these HUD initiatives are scary stuff and most community leaders don't grasp the threats they pose to our way of life in America. Single family homes are looking to be wiped off the map if HUD has it's way. Rest assured, if not stopped it will creep into every community in America. It's eye opening to read what HUD writes, they are very clear on what they have in mind.

Thanks for watching.

If the sine wave frequency is low enough relative to the transformer's overall inductance the "lag", phase shift, will be minuscule, loaded or no.

But "here" what is of interest is the transformer reaction to the application of voltage with a SHARP leading edge, FAST rise time.

Which is exactly what you get.... FLUX IN FLUX OUT.
The sine wave is to give an idea of what the magnetization inductance does. Perhaps you should read up on this a little.

I'm going to be honest. I'm not electrical enough to understand the last 25 or so posts (I still test a 9-volt battery by touching it to my tongue).

In short, I've been cruising around town with the MSD coil for the past few days and it never feels hot or anything. Do you think I need to replace it with a Bosch coil to prevent permanent damage?

YES...!!

At high RPM the MSD coil will not be as fully charged as the OEM coil.

Agree, find a good used old black coil from a 3.2L or a 964. The new ones are now silver and made in Brazil, quality of the new coils is not as good as the older black coils.

.....You are right about the silver Brazilian made coils in that two failed in my 1973.5T. I had a hard time finding any suitable replacement until a few folks mentioned the MSD Blaster coil, which I bought. Now I am experiencing intermittent "sudden shut down" that may be attributable to that coil that may be contributing heat to the CDI unit. Please correct me if I wrong here...................the CDI unit was recently rebuilt, I use a 6500 limiting rotor, standard points and this MSD Blaster coil. After a long ride, the car shuts down!!!! After a long cool down it runs again. Wrong coil? rotor?. I just ordered the original black replacement (not the silver one) from a shop in Arizona, so I hope that helps. Other have stressed to use a "non-resistor rotor".

Bob