In a question about an ignition system for a 4 rotor, I offered that it would be quite simple. A crank trigger gets the job done with the very best timing accuracy, and the system is very cheap to build.



This was in a Mazda powered GTU (IMSA) under two point five litre car. It started out as the prototype for all Lola formula atlantic cars. At the time it wasn't worth much. Had I not cut it up for parts it would now be worth more than my house.



Later I added it to a first generation GT-2 with a Pport engine. I had the leads from the pickups running through a piece of braided cover teflon hose in order to protect them from the high energy of the secondary voltages. I twisted the pickup leads in different directions so they wouldn't cross fire by inducing voltage in each other.



It didn't work. It crossfired and broke the front iron through the dowel hole.



But it did work in the Lola when I just kept the pickup wires well away from each other.



So, if you have not gone to sleep, here is one way to build a crank trigger system.



I used an early RX-2 crank pulley. It was smaller than the rest of the pulleys bolted to the front of it. It hase two sheaves. The 4 bolt holes are handy to bolt on a Mr. Gasket degree wheel.



I made an aluminum disc a bit larger than the degree wheel. With a spigot to locate it in the center of the stock cast iron pulley. And bolt it on using the original bolts. The disc is installed on the pulley and the pully is marked for a single reluctor.



For mine I used a smooth shank 1/4 bolt. With a hydraulic press, I inserted the bolt into the edge of the disc and pressed it into place.



I cut off the excess bolt and trimmed the disc in a lathe untill about 1/3 of the bolt was visible.



Two magnetic biased distributor pickups from a 70s Chrysler product are installed on made up plates on the sides of the engine, so that the pickups are within .030" of the surface of the trigger wheel. The pickups are moved up and down on the bracket to alter timing. Wires from the pickups can go into a Chrysler product amplifier. Any parts store has them . The real expensive one is an MSD. There are two other models. One with 5 pins and one with 4 pins. You need a biasing resister to run the Chrysler amplifiers. About 5 bucks. Better yet, trigger an MSD with each pickup.



The output of the amplifier goes to a double ended coil from an Oldsmobile. One wire from each end goes to a leading and a trailing plug for one rotor housing.



Here is a picture of the poor old RX-2 front pulley. The picture is captioned.



Lynn E. Hanover

Here is the front of the same pulley/disc. The timing marks are on both sides of the disc, (TDC and BDC) because you have to clip your timing light onto the number two rotor housing ignition leads and check that timing also. TDC for number two is at BDC.



Sorry about the dirt and rust. This piece has not been run in years.





Lynn E. Hanover

Your accent is a little different than what I'm used to, but I was still able to follow your posts at least 90%. That's cool that you made it work. I had an idea like that a couple years ago, but settled on modding the distributor instead, for my 0/180º 20B ignition.



So how do you think a spark 90º after TDC would work?

You cut up a LOLA?

The crank trigger bit is good. The wasted spark coil is not really a good way to get fire into a rotor. The reason is that, in normal operation you have one end of the coil firing into exhaust gasses. These are highly ionised and so look like a near short to the plug. So you only get a little 'pop' at that end, and a big bang at the end you want to fire. If you are firing both sides of the coil into 1 chamber, then you will get 2 small bangs, which will not guarantee good ignition.



I'm building a 4 rotor ignition system at the moment. I must admit to having wimped out on one area and bought an autronic 4-channel CDI. The reason was mainly for risk reduction. I could have grafted a load of igntors together to get sufficient current drive, but the systems issues (40 odd amps needed to supply them for starts) made me decide to put that on one side for further experimentation. So the total cost is running around $500 but for that I get a fully 3D mappable igntion that will ignite almost anymixture, of which 300 is the CDI box.

Yes, I cut up a Lola. But first we converted it to a single seat CanAm car.



Then we cut it up and made the second IMSA GT under car out of the running gear.



It was black with gull wing doors. I narrowed a Chevron nose. Made up all of the body molds etc. FT200 Hewland, two seats, right hand drive, tube frame and of course a rotary engine. I ran it in A sports Racing in SCCA. It got some wins even.



A collector has that car now. It is still in central Ohio.



You get old, you get smart.



No way around it.



Lynn E. Hanover

Well,



If you have one MSD 6T driving each double ended coil, rest assured the either end of that coil can weld you watch to your arm, and make you sit in the grass for a while trying to remember your name. An arc will jump from the back end of the sparkplug boot even though the engine is running and the plug is firing.



Here is a tip that can save you some pain. An MSD fires once when you turn on the ignition switch. If you have been screwing with the fuel supply, and there is a bunch of fuel in the exhaust system, the explosion can just about knock it off the jack stands, and open a closed garage door.



I ran it with just the Chrysler amplifiers and no MSD, and it works fine up to 9,000

RPM. That was all the RPM we ran back then.



The MSD runs the coil at 350 volts, so some coils will not stand up to the job without overheating. MSD has a list of coils to use. And makes an style of coil you might want.



Lynn E. Hanover

I could not remember where the apex seal is at 90 degrees, so I ran out to the shop and stuck a degree wheel on a front plate, rotor and housing.



At 90 degrees the apex seal is less than 8 degrees from uncovering the trailing plug hole. That hole has a recessed plug and a cavity in front of the plug.



I would worry about any residual burning fuel in this cavity could set off the next charge. Just one of those can break the cast iron. I wouldn't do it.



Lynn E. Hanover

This is good stuff. Someone sticky it https://www.nopistons.com/forums/pub...IR#>/bigok.gif



J

i ran a dual coil setup on my 3 rotor street car (triggered by the stock cas and an ecu) i always had problems with the spark being weak, i now have individual coils and its much better. it might be that the gm coils are much better than what i used, i had fd coils

Lynn,



Sorry, missed the bit about the MSD reading your original text. I'll read more carefully next time.



Curious that it worked with a stock ignitor though. I'll have to test that on the bench. Ignition is one of those areas where theory and practice still often have major disconnects.



I'll post some pictures as I go along with the system I am building. The advantage of using the EDIS system is that you get a fixed advance even with no ECU connected. In fact you could build a controller using no more than a 555 as a monostable and a pot on the dash. The EDIS modules are about $40 from a U-pull.

Bill, so you're saying that a DIS coil on both LEADING plugs is ok (on a 13B), but a DIS coil on leading AND trailing on ONE rotor is bad?



Mike, the setup you had was one FD leading coil per rotor housing, right? You're running six individual coils now with the E11?



Lynn, how about 90º for leading only? So each leading plug sees a spark at 0, 90, and 180 degrees. Do you think 'chamber pressure' would be rather high at that point, thus preventing the spark from jumping very easily?



The MSD basically treats the coil like a step-up transformer. It's best to get a coil that was built with this in mind. If you try to use a tradiational kettering type coil, it will probably burn up.

I'm saying that theory says that each coil will get 1/2 power compared to normal wasted spark operation. Accepted wisdom is that rotaries need nice fat sparks, so this wouldn't work well. However Lynn has proved this wrong. I can only assume that it speaks volumes for his tuning prowess in that he had a good (read easy to light) fuel mixture throughout the range.

There sure are plenty of theories out there. One of mine is to have a dedicated coil and ignitor for each leading plug because it overcomes the shortcomings of a DIS coil. Using cheap readily available parts was just a fringe benefit. I call it DLIDFIS. My 20B ignition is based on it. We'll see how well it works right after I'm done with my current project.

Its a fine theory, and if you want fixed timing then you can do it with about 3 ICS and a standard CAS. But I want to be able to adjust advance to make the beast easy to start and drive and give retard when you hit the NOS.



I do have drawings in my lab book for driving standard mazda ignitors from the EDIS, I just wasnt happy about the risks when doing a 4 rotor that wasn't mine.

yep.



concieveably with lynns setup you could go super old school and put an advance/retard lever on the dash, like a 20's car. or run that cas signal to an ecu



for a roadrace car you bascially dont need an advance curve, although it can broaden the power peak a bit

Having a mapped ignition has some advantages. Easier starting, stable idle at lower RPM etc.



But the KISS way of doing things has a certain attraction. I'll go away and try and document up all my notes for posterity, as I came up with about 6 ways of doing it.



I need to escape from the house and get a proper ignition test bed up and running as well.

Three ICs? As in integrated circuits? That would be kinda cool. I also like the 555 timer idea.



I decided to keep mine simple (because I'm stupid) and just use an FB distributor with three pickups spaced 120º and a stock reluctor in the stock location (bottom of the shaft). I did build a second pickup holder with three more pickups for trailing and a two-point reluctor at the top of the shaft, but the amount of power trailing adds vs the amount of parts needed to accomplish trailing direct-fire could not be justified at the time. Besides, leading ignition with 0º and 180º sparks takes care of the squish very well. I didn't notice a difference while driving my 13B with trailing switched on and off through all sorts of driving conditions. Plus there are no other 20Bs with 0º and 180º sparks, so it seemed like a cool bonus, if unexpected at first (you see, I was under the impression the JC Cosmo did have 0º and 180º leading sparks from the factory, heh).



Trailing does have one nice feature. If the engine has leaky oil seals, when you fire it up after it's been sitting for a while, the leading plugs can sometimes get fouled with oil. This is where trailing shines. The trailing hole is small like one of those so called spark plug no-foulers you can get at the auto parts store. This allows the engine to run while the leading plugs get cleaned by the incoming AF mixture. I've actually experienced this feature once. If it turns out to be something that I truely need, I will setup trailing as soon as possible. Otherwise, I'll just run leading only untill everything else around here is running (I may need to extra coils and ignitors for some other engine around here).

You take advantage of the fact that the 2nd gen CAS has a pulse every 30 degrees and a home pulse every 360. divide by 3 and you get a pulse every 90 for a 4 rotor, divide by 4 for every 120 for a 3 rotor and divide by 6 for a 2 rotor. Put the output of a the divider into a simple counter, and reset the counter with the home pulse.



Its so easy its just a pain I haven't had time to knock I up. there are even some $1 8-pin microcontrollers that would do it easily



Bill

Wow, that does sound easy.

Before I was "saved" I raced a Fiat Spyder. Pretty cool twincam 1800CC.



I ran a crank trigger with the Chrysler pieces. The pickup has two leads and a nice connector plug on the end, so I used the whole thing. We had 250 PSI cranking, so it could kill a starter trying to start it with 35 degrees in place.



So I added a chrysler reluctor wheel to the stock points style distributor. I ran the wires from that pickup and the wires from the crank trigger to a double pole, double throw AC switch on the dash. I ground out the advance slots, so I could start the engine with 5 degrees of advance. That way I had two sources for triggering the amplifier. Both the distributor and the crank trigger would give 35 degrees at full tilt.



You could do the same thing on a rotary with the stock distributor or the crank angle sensor and the stock crank reluctor wheel. I am lost from this point because I am not at all famillier with the computor stuff. But it can be done without too much trouble. Then you can drive around on the recommended curve and advance. Then stand on it and once above the max advance of the stock curve, you have max advance you want. Or you could change it around to retard when the nitrous comes on, or after X amount of boost comes in. Even do it with as manifold pressure sensor.



In a 4 rotor, you can run the two rear housings from the crank trigger, and the front two from a stock distributor or crank angle sensor, or the stock reluctor wheel with a bit of electronic tomfoolery.



I run an electronic distributor with just the leading pickup connected to both the leading and trailing MSDs. No stock igniter. Just a shielded pair of leads (in a dash 4 braided hose) to the MSDs. It is just too easy. When I get some time I will return to the crank trigger system timed off of the flywheel.



When we got the first Mazda electronic distributor, I hooked it up using the stock igniters that came bolted to the sides. I was trying to run split timing with it, and the trailing timing would bounce back and forth between 25 degrees of the leading and the 20 degrees I wanted.



I asked another Mazda racer about this and he didn't even look up. He just said "your using those stock igniters!" "Throw them away". Not only would they pick up signals to fire from each other, they could be triggered out of phase when the rear housing fired, and kill the engine. Mazda later moved them to the fender to get them away from the secondary ignition wires, with some success. Then gave up all together. With a powerful CD or MSD added, it was a disaster.





Lynn E. Hanover

What kind of signal is the pulse coming out of the CAS, connectivity, voltage, etc? And what wires is it on? I ask because I am trying to use a CAS off of my parts car to measure the speed of something. I have a cheap datac unit that has 0-5V analog inputs and can be programmed to count pulses if I knew how to set up the circuit. I have used a small multimeter to check for connectivity between the red and white wire pair and the green and white/black wire pair but I cant find any change in the resistance when the CAS is spinning or stopped in any position.



Sorry about the thread jack https://www.nopistons.com/forums/pub...IR#>/blush.png



Thnx

Andy

Its a variable relutance sensor, so voltage out is proportional to speed. At a couple hundred RPM you will get about 2V. At a couple of thousand you will have about 100V.



For you application it would be worth using a J109 ignitor as an interface. In this case the output that normally connects to the coil would be used as an open collector line.

i have been comtemplating a german style coil on plug system, using a crank "buzzsaw" from a cosworth DFV for quite some time now.

the rx8 actually has a system like this. its got one pickup and a trigger wheel on the front pulley. the pickup is a really funky shape, but it bascially tells the computer where the engine is in its cycle

Ok, I just checked that out. The 90 degrees, 180 degrees are OK. The spark is still in the exhaust cycle. The next 90 degree point is at 270 degrees and that is also 90 BTDC on the next rotor face. So another spark at 270 is out of the question.



So there is that answer.





The stock coil running a "lost spark" system (so they don't have to buy another coil) Fires one housing at (whatever BTDC) and the other end of the coil fires housing 2 at a point way late in the exhaust cycle so it has no effect on performance. So late in the exhaust cycle, there is little pressure, so there is a good arc (conductivity) on that end, and the plug in the firing housing is more likely to fire.



A basic rule is that the higher the cylinder pressure, the higher the voltage it takes to get an arc across the plug gap. So higher voltage and shorter gap, or, both.



As a NA engine revs up, the actual cylinder filling goes down. In other words, each cycle has a smaller volume of fuel air mixture to compress, so (A) the compression ratio is going down. And (B) Even without the compression ratio changing, the smaller volume of fuel/air would make less power anyway. So we could get past 9,000 RPM with a breaker points distributor. in a NA engine.



With no turbo, this only means that I run a leading plug gap of .010" to .015" on Ice cold NGKs ($25.00 each). The trailing are protected from the charge speed so the gap dosen't seem to bother that too much. The rotor drags the fuel air mixture past the plugs at the speed of a rifle bullet. The chance of this working at all is remote. So the rotary used at high boost, with nitrous, at high RPM etc. May have a problem with the ignition system.



If you have a high speed problem, like it feels like a rev limiter or kind of goes flat above a certain RPM, (because the leading plug can misfire and unless the trailing also misfires, you won't feel it) try a set of cold plugs gapped at 0.010". If the problem goes away or moves up in the RPM range, you need more spark energy.



An over rich mixture is easier to light than a over lean mixture. The MSD was first used on Chrysler V-8s called the (Lean Burn engines) They were jetted so lean that without the MSD they would not run at all.



So if you can control mixture, make it richer and if it runs better you have three choices. It needs more spark energy, and, or, it was too lean.



MSD has a manual sold at hot rod shops that is full of useful information. Pick it up when you go in to buy the giant degree wheel.



If you run one coil to drive both plugs on one housing, it is likely that you will have to shorten the leading plug gap, or both gaps. And for turbo installations, even drive the coil with an MSD.



Lynn E. Hanover

Yep, there is a 270º problem, but I figured out a way around it with a crank triggered ignition. Here is a quick little diagram. Each leading pickup will be mounted 180º opposite of each other. There will be a magnet at each area of the pulley where I marked it with a degree (0, 90 and 180).

And if you're gutsy enough to attempt this in a distributor, here is what it might look like.



Note, this is what it MIGHT look like. I think the leading pickups would have to be spaced 90º apart.

I have never done anything with the stock igniters once I had cross fire problems.

I took them off and have never thought about them again.

I wouldn't recommend them to anyone.



I did the first version with Chrysler pickups and amplifiers.



Then again with MSDs.





Lynn E. Hanover

How about this.



Four pickups at the four corners of the compass.



one reluctor.



Two pickups (say at 9 o'clock and 3 o'clock) run the front two housings, the same as I ran mine.



The second set mounted at 12 o'clock and 6 o'clock run the rear two housings.



If the timing retards as you rev the engine, reverse the leads from the pickup.



You have to check the timing on all four housings. I moved the pickups in slotted holes to adjust timing.





Lynn E. Hanover

Oh yeah, I forgot about the 3rd and 4th rotors. Yep, I'd mount the pickups in the same positions like you said with adjustable sliders. Infact, my 20B setup has manual adustment of each pickup because it was necessary.



I marked an old pulley every 120º and tested it on a 13B. L1 was used to run both leading plugs. L2 and L3 fired into the air. All I had to do was shine the timing light on the pulley and check the alignment of each pickup. I adjusted leading and trailing this way, but will start off with leading only due to space and parts limitations.