3800 ignition coil question

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
I'm using a 3800 coil on my 2 cylinder waste spark garden tractor.

It uses points to trigger a transistorized ignition circuit that operates at 12v. Before I changed to the transistorized ignition I was using the coil with a ballast resistor and appropriate condenser, the spark was too weak, was hard starting, and still burned up the points over the first year.

With the transistorized ignition I don't have to worry about burning up points, so I can run the coil at the full 12v without the ballast resistor/condensor. Problem is, the coil gets really hot (though it runs really well). The coil basically shorted on the primary windings and took out the transistor that came with the kit after 3 hours of mowing the lawn. So now that I've upgraded the transistor and changed it to active cooling, so it can't also be killed, I need to figure out why the coils are getting so hot.

I guess I thought these coils were designed to work with 12v. So I'm guessing that the ICM is basically doing what a resistor wire on an older car does, supplying full voltage during cranking and supplying a lower operating voltage after the engine starts?

And, no, using the stock ignition coil is not an option. It's discontinued, and doesn't fix the hard starting issue.
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
Haha, didn't even think of that.

I had initially run a pair of standard ignition coils,
coil2.jpg


But had issues with hard starting, and they didn't last very long for probably the same reason the 3800 coils don't. And it was a mess of wiring and shit. So I figured the OE coil was just one of those coils with two posts, why shouldn't any two post coil work? Probably has a lot to do with the primary resistance.

The problem is I have no way of triggering two separate coils at two separate times off a single set of points. But the motor only runs at, what, 3k rpm to begin with? That wouldn't be much different than running a 3800 coil at 6k rpm. Guess i need to find a way to disperse the heat better, or build a relay circuit that adds in a ballast resistor when not cranking..
 

jdredd

Addict
Jan 29, 2014
734
298
Can you feed the coil less voltage? Does the 2 banger need as much voltage output from the coil to ignite ?

Is lack of dwell a problem?

"The dwell on this system is held low at low engine speeds and increases as engine speed increases. The ICM will always supply a little more dwell than is actually needed so the ICM relies on current limiting to prevent the coils from overheating."
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
I would have to measure the dwell (and I probably don't have a dwell meter).

The problem with this thing is definitely starting voltage drop no matter what ignition setup you use, but the compounding issue was coil availability and cost even for a used one.

I would think though that a transistorized ignition conversion wouldn't have an issue with dwell, I'll have to email the place that sold it as it doesn't mention anything about dwell in the documentation (ie: is dwell measured at the points a 1:1 relationship with the circuit output?)

But that statement 'relies on current limiting' sounds a lot like what a ballast resistor does. But this being a cheap garden tractor, it doesn't have a separate supply circuit for starting (to easily use a ballast resistor).

To answer the question, no I don't really need a hotter spark all of the time. The motor has hours on it, leakdown is around 20%, so not having a strong spark during cranking means it's just not going to start. Once it's running it runs like a champ until the ignition coil or ballast resistor crapped out (about every 3 months before changing to the transistor setup).

Actually, fun fact about this tractor, the ignition coil doesn't get power while cranking. So it has to fire at just the right time after it starts to spin down to start. I changed that problem when i put the transistorized ignition in, so with the 3800 coil and no ballast resistor it fires up on the first rotation... I guess I traded off one problem for another.

So I guess I'm going to hunt down a low resistance Ballast resistor and just wire it up to a switch and manually switch it over. I might get bored and try active cooling on the coil using a CPU heatsink and fan. If I can keep it cool, it shouldn't fail. I could sand the plastic off the top to get a metal on metal contact for better heat dispersion.

If that doesn't work, well, screw it. I'll just build a 3800 4x4 chassis for the snowblower and mega-mower. That could be fun.
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
OK, so I'm getting ready to fuck around with this again.

To recap, the important points that I've tried to figure out...

Standard 2-post GM DIS 3800 coil being triggered through a transistorized points circuit.

No, it's not running at any higher duty cycle than it would in a 3800. Per revolution it fires one cylinder and then the other, waste spark because they are 180 degrees out, just like the 3800 would.

Dwell: I now have a older analogue dwell meter. I need to learn how to use it. But I honestly expect that the way the transistor circuit works means it has a 1:1 dwell with the points. (No clue how this compares too a 3800 ICM)

Ballast resistor: I got something from like a 50's Cadillac or some shit. Can't recall, but that's what I've had all along. When it's plugged into the circuit, the motor is not happy at all, bitches hardcore anytime you try to change revs, (causes hard if not impossible starting if it's wired in 100%), have to creep the throttle up, and even then tends to kill the motor when loaded down like when you turn on the mower deck or hit a thick patch of grass.

Coil trigger circuit: The coil circuit is now on a relay, such that it gets full voltage during cranking and reduced voltage through the ballast resistor otherwise, also I can easily bypass the ballast resistor. With the resistor bypassed, god damn does it run great, for a while, like and hour, until something overheats and just shuts off and can't do anything for about 6 hours while shit cools back down. The transistor ignition circuit is the Velleman K2543, but I upgraded the transistor to a BU323Z, which has a far superior forward voltage bias operating range at 10+ amps compared to the transistor in the kit at 4 amps (that original crap transistor shorts out and dies, I've gone through 2 so far). The new transistor is on a massive heat sink and 80mm fan which I had to 3D print a box to mount everything. It literally gives no fucks for that first hour, haven't been able to get a temp reading on it after it shuts off yet.

The ignition circuit uses a +12v power that is supplied separate from the coil supply circuit. IE: the transistor grounds the coil circuit when triggered by the points. So I could put any voltage I want to the coil.

So I guess the big question here is what voltage do 3800 coils operate at? I know this coil gets hot, but unless it literally melts it shouldn't be shutting down after an hour when operated at 12v.
 
So I guess the big question here is what voltage do 3800 coils operate at? I know this coil gets hot, but unless it literally melts it shouldn't be shutting down after an hour when operated at 12v.

12ish V to mid 13s would be correct. But they run a max of about 25% duty cycle as far as dwell to total time period available on the 3800.
 

Turbocharged400sbc

3800 & 4T80E > ALL
TCG Premium
Jun 16, 2007
32,540
15,847
hangover park IL
You're probably not going to have good success with a resistor which limits voltage and current where the 3800 coil dis ignition are meant to function at high voltage up to 18 volts. while the circuitry limits the current to 800-900 mjoules on the ground control side via the power transistors.... the coils are fed full Battery / alternator voltage.
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
I'd be that as simple as this transistor kit is, it doesn't have any way of adjusting dwell, and I don't think it's really worth my time to try having that cam profile for the actuator welded and changed..

So that leaves me with finding a (cheap, cause it's a $100 garden tractor) universal ignition system.

The flywheel on this motor is nice, big, fairly accessible. I could probably put a trigger of some sort on it. Guess I'll have to see what I can find.
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
It's an '84 Ford LGT17. I've stumbled across a few NOS on ebay, usually for about $150, and I think there's another aftermarket that is close enough for most people but it's still $120, and there are other options too but I'm not really interested in spending moneys on this. I've got to make it work using parts from my scrap bins.

I do have on hand an old GM HEI distributor, which I think contains a 4 pin HEI module. I'd need to fabricate a trigger for it, and of course figure out how it's timed. Dwell of that one is about 15 degrees at idle and 35 degrees at 3k rpm. Still reading up on it.
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
So it looks like using the HEI module is fairly simple, I just have to put a reference point at about 20-23 degrees before TDC, then find any crank/cam sensor that generates 5v reference to trigger the HEI module. Another interesting thing is there is a 5-pin HEI module that will retard timing during cranking for easier starting. I might have one of those around as well.

Problem is finding a 5v magnetic pickup sensor, most only generate 1-2 volts, so I don't think the bizarre array of sensors I have will work. I found something online that might work, but it's $10 and I have to wait two days. I'm trying to figure out if there is an off-the-shelf sensor I can get.
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
Apparently there is a way to trigger the HEI module using basically a resistor and the points.
hei-jpg.2926295


So, phase 1, change everything over to the HEI module and trigger it with the points. If it works with the DIS coil, the ya' all were right about it being a dwell issue.

Then all that's left is to eliminate the points with a $7 proximity switch and a trigger mounted to the flywheel so I never have to worry about the points lobe wearing out and changing my timing.
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
So the 4-pin HEI circuit above seems to work, but is hard starting and getting kick-back. Surprise surprise. the 4 pin doesn't have any timing adjustment, probably too much advance based on the current points setting. So I'll have to try the 5 pin module that uses the 5th pin to retard timing to make starting easier.

Edit: I figured it out, that stupid circuit triggers the spark when the points close, needs to trigger then while they are open, so I'm going to have to use a transistor to flip the signal from the points. Should be fun!
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
I'm at that point where I want to burn this shit to the ground.

What does it take to bench fire a coil? I would assume +12v on one side of the primary, and a momentary connection (as in, a quick strike) to ground on the other side should make the secondaries light up like a dry Christmas tree on fire.

I'm getting zero spark out of multiple coils, coils I thought were good.

Edit: to answer the previous question, you have to have it hooked up to a condenser as well.
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
After much trial and error, I've given up trying to invert the signal from the points.

8G5PpqD.jpg


Going to start with what I know works, and that's using a pull-up resistor and trigger a ground using an NPN NO proximity switch (Depending on how the HEI module reacts, I might switch to a PNP switch and eliminate the resistor). I removed the useless mechanical fuel pump (thereby fixing that particular oil leak) to mount the proximity sensor, then drilled and tapped a hole for a screw (which is on so tight I snapped the head of, that bitch is never coming back out) in the flywheel to act as the trigger for the switch.

I added about 3 degrees of timing, factory had it set at 23 degrees advance with no adjustment. This should give me a decent gain in power across the board, and by using a 5-pin HEI module that is capable of retarding spark by 10 degrees I can actually power the ignition while cranking and not have to worry about kickback.

While I had the motor out, I figured I'd clean up a few other odds and ends, completely re-vamp the non-existant power distribution block so I can power the HEI module, coil, electric fuel pump, volt meter, cooling fan for the module, etc.... Then I figured, hey, why not make the hokey catch can setup somewhat more functional - and relocate the fuel pump and catch can somewhere that doesn't impede functionality. And, oh, heck, let's just re-do all the wiring, add some better lighting....
 

Mattstrike

Random Crazy Custom Car guy
Feb 13, 2014
542
1,139
Got to the point where everything is back together, and guess what? it doesn't fucking work, still kicking back. lol.

So the issue here is that I think the HEI module is waiting for a set number of reference pulses from the 'trigger' to establish RPM before igniting the coil. The tractor in question pulls power from the ignition circuit during starting because the timing is set to a static 23 degrees (and we don't want it to kick back on the starter and break shit). I thought I was being smart using the 5 pin module and some relays to automate the 10 degrees timing retard and maintaining +12v on the ignition circuit during cranking.

The issue here is that I cannot static test the circuit. IE: if I manually trigger the proxy sensor, it indicates the flip, but doesn't cause a spark. However, I have noticed that when I first turn the key on, I get a really nice strong spark off the coil. It looks like the flywheel has to complete more than one rotation after I let off the starter for the module to trigger a spark, at which point RPM has dropped off enough to kick back.
 
Old Thread: Hello . There have been no replies in this thread for 90 days.
Content in this thread may no longer be relevant. Consider starting a new thread to get fresh replies.

Thread Info