What would it require to manufacture your own crankfire system?
 

I was just thinking that where does one need distributor anyway?

Could one just put an extra disc (wheel whatever) with holes to indicate firing sequence to crank "wheel" (the one with tdc and such markings) like in this cstreits engine?

http://forums.pelicanparts.com/uploads/P1010126.jpg

I was mainly thinking putting the optical pickup system from Allison Crane XR700 thingie.

There would be only couple parts, the disc/plate/spindle that attaches to the crank next to alternator/blower pulley. This would have holes that would indicate when to fire a cylinder. Then there would be this optical pickup which is already inside the distributor, this would be at the end of an adjustable arm to set the timing.

HMM, during writing this stuff I remembered (heh.. =) that one would need to know which cylinder would fire. What kinda equipment does one need to do that? Multiple coils...yes...what else?

This goes under the heading of: There's nothing really simple!

The crank fire systems need the timing pick-up, some method of shaping the pulse, some method of advancing the pulse in relation to the crankshaft rotation and some method of amplyfing the pluse to get it across the electrodes of the spark plugs.

There it is. I have probably left something out.

Good luck,
David Duffield

Not only that, you'd have to be awfully darn accurate drilling those holes... Not something a guy like me could accurately do on the home drill press...

Basics of crank-fire systems.....
1...Disc on end of crank with holes or sawtooth edge to indicate a count...part of count for position of crank...part for firing position.
2...optical or magnetic pickup to sense the holes or magnetic markers in the disc.
3...electronic circuit to fire the coil at the correct time.(at this juncture...why not 6 coils? like coil on plug?) and get rid of the dizziy.
4...why not also use the timing pulses for a timed sequenial injection system?
Coils for coil-on-plug are about $150 a piece...maybe less.
No dizzy would clean up the engine bay nicely.
All sounds too easy....but remember the rule...cost of construction will always exceed the cost extimate by a factor of 10 !!! LOL
Bob

I'd choose a hall sensor over optical, they're easy to use and would eliminate any risk of oil or dirt interfering with it.

MSD makes crank fires that are real simple but it uses the distributor to send spark to plug and adjust advance I suppose. I have heard there are issues with it picking up interference? Check a summit catalog (online?) they have all of the parts available individually. But you still end up with a distributor which would IMHO be nice to get rid of for aesthetics if nothing else.

What would it take? I thik it would take an IEEE degree....

hmm, disc on crank end, 2 sensors. One for synchronizing the spark, lets say cylinder #1. After it has sensed cylinder #1 is at right position it would just use the other sensor to count which cylinder (hole in the disc) is supposed to fire.

Electronics: sensors give a signal to this six-counter that gives a signal to high current transistors which have capasitors in-line. These capasitors have enough juice (enough joules) stored for coils.

So basically (in theory):

2 optical/hall effect sensors, other only for syncing, the sensors are positioned in different radius, sync being the inner.

Sync puts six-counter on pre-positioned cylinder number. -> the other sensor gives positions of other firing positions.

6 transistors which load up capasitors and fire at given time (signal/base-current from six-counter) giving hmm, lets stay still with 12V -> 30000V (from coils).

hmm, I really gotta think about this :)

I've mused about this before. I have a "chopper plate" (optical, I think) inside one side of a 964 dizzy. I didn't do it, so I don't know enough about it...

I assume it could be inside a regular dizzy too. This is possibly cheaper than getting a crank sensor. In fact, someone should clear up for me why a Pertronix or Crane could't do this job.

Assuming you use a sensor inside the dizzy, you can use the dizzy to distribute spark - no need for crank position etc to be measured. Again, this is in an effort to reduce cost. You still use your same distributor, coil, CDI, etc.

So all the computer (that you add on) is doing is adjust timing. With a throttle position sensor or airflow sensor and measuring RPM, you could fully map the ignition. I bet it would improve driveability heaps.

I am too lazy to read the instructions to see how it would work, but a suitable timing only computer is:

http://www.link-electro.co.nz/lem_ignition_link.html

http://www.link-electro-usa.com/v1/home.html

I bet the NZ$ price is cheaper ;) I think I asked once, and the unit is about US$500, plus sensors and the tuning program etc. I could be way off though.

Note if you ran a crank sensor it could run individual coil packs. These are probably cheap in New Zealand, as I am driving a Holden (GM) Commodore with either 3 or 6 of them in the engine bay, and there will be 100s of these in the wreckers here.

Yea :) Distributor would be the traditional method, but it would be nice indeed to get rid of it (less parts that wear). It would clean up the engine bay if one could mount the "electronics" elsewhere.

I completely agree. My main objection is the ridiculous price the vendors attach to the optical trigger and crank pulley wheel assembly. Using a donor dizzy should be cheaper.

(edit) Specifically this:

http://www.clewett.com/products/adapter.htm

Bring Money!

http://www.sdsefi.com/specific.html

This assumes you already have an EFI system on the engine and have an O2 Sensor.

If you're statring from scratch, add manifolds, fuel rails, throttle valve, throttle position sensor, MAP sensor, CHT sensor, not to mention injetors, etc, etc.

James

re: computer stuff for mapping

Feel free to go at it, the computations are
pretty simple, enough that there's been
completed projects shown in some magazines
I read. Of course, the difficult parts of it
are knowing what your curves should be:

You REALLY REALLY don't want to screw
up the advance and blow your $$$$ motor.

Mark S.
'70 914-6

I left out one important thing in my previous post....you also need a second sensor to tell when the #1 cyl is in the firing position.
Some companies put a sensor on one of the rocker arms...some put one on the cam drive.
It all works the same....the #1 cyl is only at TDC on the compression stoke once per 720 deg....thats when the ignition is fired.
Of course...advance is easy...just pre fire the plug before TDC.
One of the timing rings I have been playing with has 65 little humps on it....with 63 at regular intervals...and 2 very close together...between those and the second sensor, the computer can figure out where it is in the rotation.
Now...just have to find 6 coils to install on the beast.
Bob

nevermind!

Well it can be done easily, but advance-thing means that there are microcontrollers involved and then it get's nasty...

Really, there are DIY-projects like this...wait for Megajolt, sucessor and add-on to Megasquirt!

Has anyone tried this? And, if so, what was your experience:

"PORSCHE 914 CRANK FIRE DISTRIBUTOR"

http://c767170.r70.cf2.rackcdn.com/R022905205CF.jpg