hey,

I am thinking about just messing around with a CnP ignition system, to just get my feet wet in some electronics stuff. My idea is to use my distributor's timing outputs (that would normally directly fire the coil) to now drive a circuit that fires coil packs.

I am going to use:

http://rocky.digikey.com/WebLib/Texa...017B,4022B.pdf

This simple counter should sequence the coils in order, while a cam sync sensor Resets the counter after all of the coils have fired. Some form of power transister (likely one designed to limit current) will drive each of the coils.

My main concern at the moment concerns coil dwell and saturation. Using this chip, each coil would remain off for 90* of crankshaft rotation, and then turn on for 630* (two engine revoultions minus the 90* that it was off for). Would this be too long for each coil to remain on? I would think that the added dwell would be compensated for by the increased 'off time' of the coil. Would it overheat the coils or not?

If so, would the fix be to use a 4pin hei module to drive each coil, since these have active dwell adjustment?

Any ideas or comments would be appreciated

 

Before I can suggest anything, what coils are you planning to use? IOW: just regular coils, or a CnP setup like used on the GM LS1 vehicles?

RBob.

 

most likely I will be using something very similar to the ls1 coils. Either that or a wasted spark set of coil packs, either ford edis or the gm coils ( GN and the like).

 

The LS1 coils have a driver built in. Just need a 5 volt signal to start dwell and then fire the coil.

As for the actual dwell it shouldn't be more then 5 msec. This will prevent both the driver and coil from overheating. At high RPM (IE: 6,000) there isn't enough time for that amount. Can use the ECM signal for dwell as the GM ECMs have logic to control dwell time.

RBob.

 

What happens if I don't have an efi setup

Maybe I will just have to do the math and see how long (time wise) that the coils are off/on when driven by the chip that I plan to use.

For the LS1 coils the 5V line goes high when you want to start charging the coil, and low when you want to initiate spark? Just curious cause sometimes the OEM's do weird things, and I haven't quite been able to find much info on the LS1 coils.

Anybody have ideas on how to control the dwell with this setup?

 

LS1's are high to start dwell, falling edge firs them. They will also self-fire after 100msec of dwell. I'd recommend a PIC or such to control the dwell. Could also use a state machine based on a PROM.

How are you handling the advance portion?

RBob.

 

Well initially I was thinking of just using the hall effect trigger in my distributor (Pertronix ignitor) along with the mechanical and vacuum advance in the dist, to control the timing advance.

Eventually, when I upgrade to efi, I will just use the EST from the ecm to control the advance.

However, this dwell issue is becoming a problem. At idle, the coils would be getting far too much conduction time if driven directly off of the IC I have chosen. I am still toying with some different ideas to get around having to acually use a microcontroller to control the dwell:

1.) Have the voltage driving the coil, a function of rpm. Either have a continuously increasing voltage with rpm, or just have say 6V going to the coil at idle and low speed, and then 12V after a certain rpm is reached. Thats not too difficult of a circuit to design.

2.)Use the charging times of capacitors, along with a comparator, to delay the onset of dwell from what the IC would dictate. This isn't terribly accurate, but should be able to get it somewhere in the right area.

3.)Use a CDI to virtually eliminate this whole dwell thing. Problem is, would I need to use a MSD DIS box to achieve this? I can't think of a way to get multiple coils to fire from a, say, MSD 6al. That kinda eliminates the 'cheap' aspect of this project

After that, I would just use something like a PIC or one of the Zylog microcontrollers to control the dwell. But when you get there, you might as well almost just design your own ignition system.

As always.. would love to hear what anyone has to say

 

There are some single-chip ignition controllers around, like the SGS-Thompson L482 DIP.
All the ones I have seen were for a single-coil setup however. Look up the SGS application
notes and you can see how they handled the logic for dwell calcs (including coil feed-back).
Try this link for ST:

http://eu.st.com/stonline/books/pdf/docs/1339.pdf


It doesn't really bear on DIS, but FYI there is a little HEI write-up here that mentions the ST circuit:

http://www.corvetteforum.net/c4/doctorj/heicoilinfo.htm


Also, Electromotive advertises a stand-alone waste-spark DIS that uses no ECM control -
have you looked at their package?

 

I have yet to design something around the sgs chip, but it indeed does look somewhat doable.

I have been thinking of using a 555 timer to delay the onset of the charging of the coils, so the dwell is reduced (although still not ideal, or exact). This would basically only be used at idle and <1500 rpm, to keep dwell under say 50ms or so. Yes this is still too long, but unless some other ideas come up....

Eventually I might just design a microcontroller based interface that can take the stock EST signal from a '730 or the like, and directly drive a CnP setup. It could certainly be one cheaper than the eDist.

One other question is, can a msd (or similar) crank trigger be used to feed the base timing signal to a stock gm ecm? Basically removing the magnetic pickup from inside the distributor and putting it on the front of the engine. Both seem to be similar styles of pickups.. but I could be wrong here.

 

Just as a point of reference, in developing a new system note that the GM and ST ignition designs
contemplate current limiting of the output (as you mentioned) as well as schemes to avoid a non-
synchronized spark on system power-up, 'soft discharge' of the coil when the engine is not running,
noise supression, and a few other little odds and ends useful for a workable daily-driver...


As far as crank triggers go, the MSD pieces seem pretty expensive. If you were working on a small
block Chevy, it is possible to use stock parts from the later Vortec truck engines to make a timing
signal that runs off the crank snout. The GM truck pieces are cheap(er), readily available, and
very robust. They also reside inside the timing cover, rather than hang out in the air... The ECM
will neither know nor care where it gets its reference pulse, and I find the crank signal to be more
stable than one from the distrib shaft.

The only caveat is that the stock Vortec trigger is a Hall switch; as such the signal needs to be
adapted to feed a 727 ECM.