PDA

View Full Version : Haired Brained late Nite idea: Homemade dizzy


Captain Bondo
09-13-2005, 04:22 AM
Ok so I've had this idea for a while, it's 1AM and I'm totally wired and am still unemployed, so prepare to bask in the glory of a kid with too much time on his hands.

Concept: Home-Made Electronic Distributor

Now, the distributor in a typical MS car does two things:

1)It gives you 4 pulses per cam revolution. So you get the 4 sparks you need.

2)it has a rotor which basically acts as a traffic controller. It tell's each of the 4 sparks what cylinder to fire.


Now then, T6 motor's are a bastard because, well, they don't have distributors. And the rest of us might as well face it, distributors are weak sauce these days.

Now then, on a t6 life is a b*tch because it has six friggin coils. How you gonna drive six coils? Well, ,you can buy edist for 350 friggin bucks, or you can buy an expensive-ass engine management system for the safe of getting 2 more ignition outputs. So that's about 800 bucks an output. :roll:

But really- shouldn't it be a simple task to make a box that can take a single ignition output, and turn it into as many outputs as we want? We know the firing order, we know how far apart the pulses are, what's the problem?

So here's the deal. You need two signals.

Tach signal for timing reference. Use the 60-2 flywheel. ok check.
cam position so we can phase properly. Whiteblocks already have em. You could easily epoxy a magnet to the cam pulley and mount a pickup on a red though. It doesn't need to be accurate at all. You could easily set this up when relatively drunk/tired/confused.

Ok so we have some signals. sweet. ok.

so what we need is a standard edge triggered "counter" chip. only needs three outputs (up to 8 cylinders). I made one from nand gates with schmidt triggers or something once. but anyways they're cheap pre-made.

basically it has 4 outputs (we only need 3 for a six cyl). as it receives pulses, it counts in binary sending the outputs high to represent "1".

So at 0, all o/p's lo- it switches a solidstate relay or transistor that allows a signal to travel to number 1 coil.

an ignition pulse comes, fires that coil, triggers the counter, the counter increments to "1". Our series of logic gates result in a logic low input at one of the nand gates. The ign output of the MS is connected to all of them. When it pulse low, you have two lows. The nand goes high, an inverter after that (with an open collector output and a pull-up resistor goes low, and collapses the coil

The counter keeps incrementing with the ignition output, and when the motor completes 2 revolutions (all cylinder have fired) the cam sensor triggers. The cam sensor is set up to trigger the reset on the counter. The counter goes back to zero, and we start over.
You should be able to trigger as many coils as the counter has output states like this.


So the only tricks here are
1)you might need to delay the ign signal to the counter input. You don't want it to switch until AFTER the coil has fired. The propagation delay of the TTL devices might be enough though.

2)you need to "tell" megasquirt that it needs to send 6 equal ignition pulses for each tach pulse (I assume you could mod the code to do that if you had to?)

Anyways It's amost 2am now. I will make a schematic tomorrow. Yes I am mentally unstable. Don't you guys think it'll work though?

Matt Dupuis
09-13-2005, 08:19 AM
Yeah, why not?

If you get around to that, it could also be applied to sequential injection, correct? Go batch fire until the cam position is known, etc. I too have a cam position sensor on my engine, and could use a little less wasted fuel.

linuxman51
09-13-2005, 09:45 AM
****, i didnt hit submit..

anywho, why not run wasted spark and just use 3 ignition outputs? Sequential ignition is very cool (and i think with the coming of ms2's can capabilities not too far off) as is injection, and wouldn't require trying to keep track of extra circuitry that may or may not be easily modified. anyway, as per the tach signal from ms, as long as you set it up for a 6 cyl with 1 spark output, i would think it would generate the pulses you're looking for out of the box without any kinda extra or new code.

Captain Bondo
09-13-2005, 11:53 AM
Yup! You could get sequential injection the same way!

In terms of wasted spark, the only problem with that is that at 6,000rpm, the coils are going to think they're doing 12,000 rpm, and I don't think they'll be terribly impressed with that. They might be fine but I just don't know. These motors seem to like smoking coils as-is. That might work though.

Other things I realized:
the cam trigger magnet can't be at tdc. you need it to trigger far enough BTDC to be able to ensure the counter has reset to zero before the ignition pulse for #1- do you have 120 degrees between ignition events on a six, so you might put the trigger at 70* btdc. That will ensure the circuit is stable to at least 60 degrees total advance.

This could be interesting. The logic gates aren't worth more than maybe 10 bucks....

Captain Bondo
09-14-2005, 03:20 AM
Aww, crap. Won't work really. By the time you have a 6cylinder at 6500rpm, you run into having dwell overlap. ba$tard. So, you can't run it all from one ignition output.
However, you could set up three outputs like kenny suggested, make MS think it's running wasted spark, but divert each output alternately between two opposite coils. Opposite coils won't overlap.

Dwell overlap is really interesting and really helps explain why 6cylinder and especially 8 cylinder high-rpm motors pick up power using COP or CDI.

So basically it can be made to work using three "miniature" versions of what I described. Same goes for using it for sequential- as soon as you need to power two injector at the same time (ie high duty cycle) you're screwed.

Matt Dupuis
09-14-2005, 08:20 AM
Good point on that. TFB.

Matt Dupuis
09-14-2005, 10:47 AM
Actually, for the ignition, couldn't you build in the dwell feature to the "distributor", which then just takes the leading (or trailing) edge of the dwell pulse from Megasquirt as the signal to fire the coil?

I can't think of a good solution for the injection function, just yet.

BB-Q
09-18-2005, 02:38 AM
Yup! You could get sequential injection the same way!

In terms of wasted spark, the only problem with that is that at 6,000rpm, the coils are going to think they're doing 12,000 rpm, and I don't think they'll be terribly impressed with that. They might be fine but I just don't know. These motors seem to like smoking coils as-is. That might work though.


I was thinking- how about motorcycle coils? They're 12V, you have a choice of coil in head or wasted spark and even the wasted spark coils will run at 12000rpm (effectively 24000) all day long. Most turbo bikes use Dyna coils with no problems at all- even the ones around 500bhp.