My car has been running quite rich, especially in idle, so i'm going to play around with this and see what I can get. I would imagine that having a rich idle (in the 12s) right before shutdown would leave a bit of fuel in the cyls.

 

It was said somewhat rhetorically. Inertia causes the engine to continue rotating for a bit after you switch if off. Initially, I suspected that maybe this problem was related to people with lightened flywheels, but that seems not to be the case, at least not exclusively.

No. It's a stupid design. And it's not something the MS is doing intentionally. The problem is the circuit that we all add to drive the spark outs- the one that involves soldering pullup resistors to the LEDs and running wires from there to the connector. It's a pull-down design, meaning that anytime the circuit is energized but the CPU is inactive (during bootup, or during a reflash) the outputs go high. A better design would make the driver outputs follow the state of the CPU pin. Pin high = output on.

 

well i don't really understand electronics but i'll accept that as a solution, can this be reasonably changed?

 

It would be built in the proto area, ignoring the LED drivers.

I really should get off my *** and prototype this. Since I've still got LostSoul's MS here (we're installing it tomorrow) I'm gonna run down to RadioShack and pick up some PNPs right now, and build this thing. Data to follow.

 

Cool, keep it up!

 

God I wish I understood electronics better. I hope this goes to plan, if so I'm pulling my MS out tomorrow and re-doing it!

Nice one Joe

 

FYI, leaning out my idle seemed to fix the issue so far. 3 starts so far without a pop, it would pop every single time before. I also seem to have my cranking settings just about nailed

 

i don't know what we'd do wihout you joe!

i would love not to have this happen ever again

 

I've changed the design to use an NPN transistor instead of a PNP, since there are already NPNs on the board and therefore the MS distributors have them in stock. It's working on the breadboard. I'm going to build the design into LostSoul's MS, and I'll stim test it later this evening, and hopefully we'll run it on his car tomorrow.

Updated schematic and build instructions to follow.

 

Well, I've done both channels of Lost Soul's MS, and it's up and running on the stim. A teaser:

 

What are us noobs looking at here...?

 

<< Also a newb, but those look like alternating high voltages on an oscilloscope, which I'm assuming would be our spark outputs since we run wasted spark.

 

So, you've changed that they do not go high during bootup but work normally after that of do you need to change the coil wiring as well?

(I changed my major the minute they brought this stuff up)

 

Those are the SparkA and SparkB output signals. Not HV, just the triggers.

The setup is now double-inverting (rather than the normal way we do it, which is single-inverting) so the output follows the logic state of the CPU pin. You'll need to change SparkOutputInverted to YES, but everything else remains the same. With this setup, the outputs remain off during powerup, and will also remain off during a reflash.

It's blindingly simple, actually. Here's the MS schematic, with the circuit we normally build added in black:

As you can see, when the driving transistor is inactive (not conducting) the spark output is on. The CPU has to assert a voltage on the pin feeding the base of the transistor (through the 1k resistor) in order to turn the output off. It can't do this when it's inactive, as during bootup or a reflash, so during those times the outputs naturally go high.


Here's the modification I've done:

It's blindingly simple, really. All I've done is to tap the circuit at the same point we normally do, but instead of directly feeding it out, I run it through another inverting transistor stage. This way, when the CPU is inactive (its output pins are not on) Q6 and Q8 remain off, and their outputs remain high, however this in turn causes the new transistor to turn on and conduct, pulling the output low. When the CPU does assert a pin high, Q6 / Q8 turn on, which turns off the new transistor, which causes the output to to go high.


So far I've only bench-tested this. We're going to install it on a car and run it today. I'll have results, along with more detailed scope captures that clearly illustrate what actually happens at powerup, later.

 

Success!

The MS is installed and running in LostSoul's '92. We just drove to BW3 on it, so I'm writing this while we're waiting for our food. Sidebar: why the hell do they have free wi-fi at a wings place? (Not that I'm complaining.)

I'll try to get this documented shortly.

 

looks simple enough, but I don't have the pooping issue, so I probably wont bother.

 

I have the pop but will need alot of documentation to get this done.
(Only to find out the pop is from the heated WBO2?)

 

Indeed it runs, and runs sooooo much better than before Joe is the MF'n man!! Datalogged a few runs, and didn't run into cut until the last one, so I'm going to start the basic VE tuning from that one. I'll send you the log Joe.

 

Full writeup and build instructions: https://www.miataturbo.net/forum/t33964/