On my rebuild, I'm making a new wiring harness for the motor and I'm trying to figure out the best way to manage it on a megasquirt.

So to all the megasquirt guys, if you were in my position and starting from nothing, would you stick with the stock cas or run an EDIS trigger wheel? I've heard a horror story or two about noisy cas signals, on rotaries and miatas. on a rotary, bad ignition can ruin your day (and apex seals) pretty quick. I know rx8's run on EDIS and i know its a more accurate system, but is it worth it in your experience?

also, unrelated less important question. I have a set of high impedence red tops, what is their plug style, and can I buy new electrical plugs for them anywhere?

thanks.

 

Properly shielded cable and good wiring connection and routing practices are mandatory when fabricating a bulletproof CAS install. A noisy EDIS signal will have the same effects as a noisy CAS signal. RX-8 doesn't use an EDIS wheel - it's 36-2-2-2. Your injector plugs might be found at http://diyautune.com. If they don't have them they may be able tell you who does. I think they're a Bosch product but am not sure.

 

We have more or less solved the noise issues... As long as you use good cable and follow the howto documents and make sure your grounds are nice and clean and attached to the correct places, you'll be fine.

I wouldn't use EDIS as then you don't get control of split, and you lose out on the ability to run things like negative split, ignition-cut revlimiting and launch, etc...

As pmrobert says, the rx8 ignition wheel is a strange one, 36-2-2-2... It's not even close to being anything like EDIS either as it is a coil-on-plug leading and trailing system. It's actually not that much more accurate than the stock CAS due to the placment of the missing teeth parts of the wheel... they're right where you'd normally want to do ignition. With the teeth missing the engine can accelerate more between teeth and ignition can be further off.

I've been using the stock CAS in my installs for 4 years now with no issues once the grounds are good and the 2nd trigger conditioner is configured properly.

Ken

 

thanks for the info. I'll probably just stick with the stock CAS then.
but how does running a trigger wheel prevent ignition cut?

 

The EDIS module decodes the trigger wheel pulses and passes rpm and TDC data to MS. MS then sends data back to EDIS (SAW, Spark Angle Word) telling it which angle BTDC you want it to generate a spark pulse at to the EDIS coilpack. The problem is that you can't command EDIS to not spark but still send RPM and position data. If you don't send SAW data, EDIS goes into "limp home" mode and defaults to a 10 degree BTDC timing regimen. You also lose control of split timing (as in always at 0) because an EDIS/rotary hybrid fires in wasted spark mode, i.e., L&T for rotor 1 fire as a wasted spark pair and L&T for rotor 2 fire 180 degrees later as a wasted spark pair. In an NA engine that's perfectly fine - I spent some dyno time a few years to try to quantify how much trailing spark actually contributed in the NA case - there was realistically no difference in power or EGT at varying amounts of split and even no split at all. I'm sure there's differences at part throttle, low load situations where the trailing plug and proper split help with emissions, etc., but don't have hard data on that. Mazda didn't put it there for looks and I'm sure it helps clean up HC emissions under low load conditions.

 

ah i see. I ignorantly thought that EDIS was just the name for a system using a crank trigger wheel. Thanks for clearing that up.

Sorry about that!

What about using a crank trigger wheel with some kind of aftermarket coil setup, or coil-over plug setup, or I guess you could still use the stock coils? From what I've gathered though, any gain in accuracy of crank angle wouldn't really make much of a difference though, would it?

 

A crank wheel scheme removes the small amount of gear lash inherent in the stock CAS. I'm not sure that the increased accuracy would translate to any real world improvement. Ken would probably be much more qualified to comment on this issue. You can mix and match lots of items such as coils if you understand how things interrelate. For example, I use 4 GM L1 coils firing in FD mode. I've used stock FC coils for several tens of thousands of miles without complaint as well. The LS1 coils were chosen to give me more flexibility in mounting as things are getting crowded in the engine compartment. I fire 2 LS1 coils off one logic level output for leading spark, the leadings in traditional form.

 

Actually mike, that might've just been your engine. The engine we tuned for the GRM article lost power at full throttle as soon as you went more than about 2 or 3 degrees split. It only lost a few horsepower, but it was repeatable every time. We didn't try going more than 2 or 3 degrees of split though since we saw it start losing power there.

This might have something to do with the higher-compression rx8 rotors or something of that nature.

Ken

 

If you wanted to, you could drive the leading coils in COP mode too.. I wrote that to work on the Renesis engine.

The gear lash in my experience only translates to fractional degrees of jitter. You could potentially gain accuracy under very fast acceleration by going to a crank-mounted wheel with more teeth though. 36-1 for example has a tooth every 10 degrees (except where the missing one is). The stock CAS has a tooth every 30 degrees.

The code uses every tooth on the wheel to schedule spark, dwell, etc... At low revs, the engine speed can change a LOT more in 30 degrees than it can in 10... I've seen this translate to 2 or 3 degrees of lag in spark during very fast acceleration. Using prediction can recover some of that but it won't be better than having more teeth.

Ken

 

the stock cas wiring is shielded, it looks like braided metal wires, does that get grounded?

 

Shields typically get grounded at one end only, closest to the ECU. Yes, shielded wire should be considered mandatory for the CAS signal.