Disclaimer: The technical information and opinions below are for education and discussion purposes only. They are not an endorsement or encouragement to alter factory parts on a vehicle of any kind.
High voltage and high horsepower engines are inherently dangerous, and do not tolerate fools gently. Any use of this information is at the user's sole risk and liability.
Oct 2003 - DrJ
Ignitions are not my forte, but I did a literature search on GM HEI systems, in preparation for a winter project I did last year. Below are extracts of the information I found. Note that I have not verified all of this on the bench first-hand, so a grain of salt is suggested and corrections/additions are welcome.
The original topic came up on CorvetteForum, from someone wanting to upgrade a non computer-controlled HEI distributor on a small block. Here is a portion of the information that was offered in response to the original Q:
1. David Vizard did some testing on the old-style V8 HEI mechanical advance distributors, and in Chevy Small Blocks on a Budget reported no problems with the magnetic pickups out past 6000 rpm. He did complain about the GM HEI switching modules though. I think there were reasons for the complaint:
2. As far as I can tell the old 4-pin GM HEI modules were internally limited to about 5 amps total current flow, and had a fixed minimum dwell of some 3+ms. The dwell is probably the source of the mythical "4500 rpm limit" attributed to the stock V8 HEI distributors, at least with the coils of that era.
(Note that newer GM 7-pin modules have a higher current limit & use adaptive dwell from the ECM.)
The specification sheet for the (now obsolete) Motorola controller
used in the original GM 4-pin modules is shown here:
(That link may not always work - if not, try a web search for " mc3334 " - you're looking for a pdf of a Motorola spec sheet.)
Here are some partial highlights of the Motorola specification:
The current limit & coil saturation (variable dwell) along with the 'zero-crossing' circuit are on pp 1&2. You can see quite plainly where that circuit begins to drop coil current above 3000 rpm.
3. Replacement 4-pin modules from MSD, Accel, etc have an advertised 7.5 amp current limit. I believe this is the 'extra energy' they advertise over stock GM parts. (They probably also use a smaller min dwell time, but I don't know exactly what.)
4. The current in the switching module affects the energy in the coil primary to the second power: If I remember my EE courses correctly, energy in an inductor = 1/2 L*I*I.
That means that if you double the coil current, you get four times the stored energy, in simple terms.
The extra current from the aftermarket modules by itself won't do anything special for the engine, but it allows faster coil saturation & greater stored energy in the coil. Directionally, this is better for engine performance at higher rpms.
However, if you consider using MSD products to boost ignition energy, note that their designs have other problems in my opinion:
5. What I have seen of MSD designs ( ftp://ftp.diy-efi.org/pub/diy_efi/documents/msd6a_02.pdf - that site has been up & down a lot lately) shows that they do NOT have an adaptive mag pick-up 'zero-crossing' feature - they invert the pick-up coil polarity, and fire on a positive going signal when used stand-alone. I think they have to do this for compatability with other systems, but to me it's not the best way to run an ignition.
An interesting alternate to the 'aftermarket' 4-pin HEI modules exists, for those concerned about spark at high rpm:
6. There is an MSD inductive ignition box (MSD-5900) advertised to also draw 7.5 amps. It has a circuit similar to the SGS-Thompson L482 example circuit here: http://eu.st.com/stonline/books/pdf/docs/1339.pdf. (If that link doesn't go through, go to the ST site and search for part # L482 & its application notes.)
Here is one diagram from the ST specification:
This box moves the coil-switching function outboard of the distributor to a big ST- BU941ZPFI ignition Darlington with its own heatsink, and has some sort of variable-dwell built in from the L482 controller - though I haven't verified its time constant.
Getting the ignition box away from engine heat is generally a good thing. And, if you trigger a 5900 box from the GM module (to use its zero-crossing feature), it looks like you gain the benefit of higher current switching & variable dwell without relying on the MSD trigger circuit. I don't think it would make a big difference, but (with the right coil) it should make more spark energy at higher rpm without using a CD box.
Others manufacture external inductive ignition boxes too - but I ran across a schematic of the MSD one, so I know generally what they used for a design and the major parts involved.
There are several after-market coils that seem to be oriented toward the higher current capabilitiy of newer ignitions. I note the Jacobs C4 coil has advertised primary characteristics that give a saturation time of ~2.4ms (@ 7.5 amps - storing ~112 mJ). There are other coils that might be able to throw off similar energy within the single-coil V8 time constraint - a brief comparison is given below.
Coil saturation time is taken from:
[ t = - LN(1 - (IR/V)) * (L/R) ]
where t = time
Coil Voltage (advertised as claims of 'high performance' or 'hotter' coils):
The HEI information above doesn't speak to secondary voltage
in the coil, but if anyone is interested there is a depiction
of the general case here:
A short discussion of voltage & field density is here:
A short discussion of arc time and current is here:
Secondary voltage (referencing the MSD-5900 example circuit) is induced by the collapse of the magnetic field in the coil primary. That voltage follows something like V= L* dI/dT; and is multiplied by the turns ratio of the secondary/primary windings. As far as I can tell, the stock GM HEI coils used about 85:1 for this ratio. The energy transferred from the primary winding to the secondary depends on how the coil was made and the properties of the core.
The Megasquirt developers posted their excellent simulation
program for evaluating ignition coils here:
Voltage claims for aftermarket coils may be an indication of their insulation breakdown voltage; or possibly a reference to their initial voltage when used with a CD box. (From the MSD6 paper, note that the coil primary in a CD application is fired with -400 to -500 v from the cap. With a 100:1 coil that puts the secondary in the 40-50 kv neighborhood.) Only the insulation breakdown V seems to have much importance to the ignition system.
Editorial Comment: If you read through the references above it is apparent that all the claimed advantages for 'hotter spark' coils, when used with a stock ignition module, are delusions of technically incompetent folks selling useless products to technically illiterate people. That is to say, such claims are horse-pucky, and such coils a waste of money, at least in terms of measurable engineering benefit.
Your mileage may vary. Spark energy is what lights the fire, not spark voltage. Energy comes from the ignition module current, the coil just stores it.
Spark energy stored in an inductive coil was given in the earlier equation; for a CD box the spark energy is from the capacitor in the form energy = 1/2 C*V*V. I haven't found much in the way of measured data, but apparently the Kettering type ignition only gets 1% or so of the primary energy out to the plug gap where it is used.
Coil Saturation Time (dwell):
At 600 rpm a V8 only needs to fire a plug every 25ms; at 3000 it's every 5ms; at 6000 it needs to fire every 2.5ms. Hence the problem of single-coil saturation (coil energy) vs. time at high rpm.
As near as I can tell, the stock HEI coils were about 0.5 ohm & 8 mH on the primary with 85:1 windings. (Measuring actual coil inductance, resistance, & capacitance outside of a lab is not doable - you have to disassemble the thing to get meaningful readings.)
You can plug those values into the B&G coil simulation (use 90* dwell) and see what happens to spark energy as rpms go up - remembering that the simulator only has one fixed dwell angle, like an old-fashioned points distrib. Also remember that an early HEI is current-limited to 5 amps, so you can ignore readings higher than that. The result doesn't look too good - that's why GM improves the systems in later years.
A newer design single-coil system can use a higher current limit and a coil with lower primary resistance/inductance (with variable dwell) to even out the secondary voltage/energy over the rpm range. (Run a coil with 0.5 ohm, 4 mH, and a 7.5 amp limit through the same simulation and note the difference.) The improved design just means you don't burn up the coil at low rpm, nor run out of juice at high rpm, and still show high energy values. That seems to be the direction of newer GM single coil ignition, among them the LT4 design and the Vortec truck engines.
Aftermarket ignition boxes & coils try to adapt the newer technology to old engines. However I note there is precious little design info available from box & coil vendors (and not much reason to believe what is published), so how it actually works is a crap-shoot.
There is a lot of interesting background reading on ignitions and other topics in this archive: http://yarchive.net/car/ignition.html
Also note that by the time you get to the LS1, GM scraps the whole design problem above, installs eight coils, and puts the distributor in a museum. As I understand it, the LS1 coils have current limiting built in, and do their own dwell calc as each plug fires.
The next-generation ignition is supposed to have per-cylinder
ion-sensing on board, detect mis-fires, & refire the plug
in real time - Saab has published some papers on this if I remember
correctly. For those interested, there are some Delco-SAE papers
for download here (recent stuff, not just ignition related) that
speak to future electronics:
HEI Coils - evaluation of published data:
To clean up my desk I collected my notes on 8 coils that might be useful in an HEI inductive ignition (single-coil), and stuck them in a spreadsheet. A jpg of the result is here.
Short saturation time & high energy are GOOD, slow times and low energy are NOT GOOD:
Note that I used the best info I could find for the original GM coil-in-cap; I have no specs for the LT4 coil (and module) but believe it is an improvement over the old-style. Everytihing else is from manufacturer's specs.
Mallory & Accel publish NO specs that I could find, so they are not represented.
Jacobs' advertising makes unusual use of electrical engineering language so I take their specs with a grain of salt.
Only MSD is technically competent in their writing - but I have NOT measured any of the values claimed by anyone, so use your own judgement.
The coils were compared via the simplified equations previously given for stored energy & saturation time, just to put them on a common basis. I used 13v and 7.5 amp feed (except for the stock design).
Conclusions of spec evaluation (which claim no connection to reality whatsoever):
The stock coil does a pretty good job as designed,
but dumps energy above ~4000 rpm.
These were considered ONLY from the view of inductive ignition. CD ignition is a whole different animal - coils there do not store energy, so a good CD coil will want properties different from the inductive kind.
1. Some time ago I ran across a web page for DUI ignitions & their 'Dyna' Ignition modules. Whoever wrote that ad copy has NO electrical training whatsoever. Their 'performance' claims are such erroneous crap I would put them dead last on a list of people to ever buy from.
2. For anyone who has read this far, I believe the GM 7-pin HEI modules draw around 6 amps. I have one to measure this winter, just to see if that info is correct.
3. I have seen specs that say the GM LT4 module draws over 7 amps, but have not verified it. If the info is correct, the best HEI conversion for higher RPM operation would be something like a Jacobs C4 coil (or an LT4 coil, which I believe to have similar specs), driven by an LT4 module. I'd probably use the existing HEI module to trigger the new one, and have a high-power, high-reliability conversion that takes up minimal space. If I build another HEI system that's what I'll use - just have to work out the connections.
The LT4 Module/coil are probably good up to ~6500 RPM or so. Beyond that point you are likely going to need DIS or a CD box.
Thus endeth the lesson for today. See Disclaimer Above