All,

From everything I've researched, there are 2 ways that the MS-II can be set up to control the stock leading & trailing ignition on the 2nd gen RX-7. To summarize, option 1 involves using 2 VR input circuits (i.e., need to add 1 more VR input circuit to the MS-II) OR option 2, which needs just the existing one VR input circuit on the MS-II, but requires you need to mod the stock CAS by grinding off 2 opposing teeth from the 24 tooth wheel.

What I have not been able to find is a comprehensive discussion of which implementation is "better". I realize that some of the pros & cons for each method are subjective, based on personal preferences (i.e., some guys might just be more comfortable modding a CAS than adding electronic circuitry to the MS-II). What I'd like to instigate is a discussion on what are the pros & cons of each method are, in terms of: (a) Performance capabilities & limitations of each implementation; (b) Ease/difficulty in tuning the completed system.

Cheers,
Pete

 

I doubt there's a difference when it comes to "ease" of tuning. The ecu gets position information either way. Running dual VR should have a higher tolerance to interference because of having two positional inpuits opposed to one.

I'm running dual VR on my daily driven TII and have not had any issues.

 

Personally, I mod the crank angle sensors instead of building the circuit. I have had absolutely NO problems with this whatsoever, and modding the CAS is soo simple and straight-forward, as well as setting the CAS. As with building the VR circuit, I personally don't know of any benefits. However, I am open to learning anything beneficial from using the vr circuit as opposed to modding the CAS.

Brian

 

These are exactly the kinds of performance issues I'd like to explore.... Obviously the dual VR implementation is how Mazda ended up doing it -- doesn't necessarily mean it's better, but there must be some good reasons for doing it that way. They could have just as easily designed a CAS with the missing teeth from the get go, and it probably would have saved them some production costs to boot.

That's good to hear, I'll be starting my '89T2 MS-II project after the holidays. How did you end up implementing your 2nd VR circuit? I was planning on doing mine on the proto area of the v3.00 board using the write-up on this thread:

https://www.rx7club.com/megasquirt-forum-153/2nd-vr-circuit-build-v3-ms-board-690000/

 

I fit my 2nd vr circuit completely on the proto area of the v3 board, its not hard to do. From what i've seen, the 2-trigger (stock CAS, 2nd VR circuit) setups are more reliable and less prone to interferance than the 1-trigger (ground CAS teeth), but either can be made to work.

 

This would be a good thread to illustrate the different spark settings for each type of code (ms2 vs. ms1Extra) as well as stock cas/ dual vr circuits vs. modded cas w/ single, if anyone has this info. The FAQ really could use a "freshening up"when it comes to this stuff. Lots of good info on settings for stock cas, but very little for modded cas. I would provide some input of my own, but my laptop took a dive on me. Not at all happy about that.

 

On ms2/extra, the "dual wheel" setup could potentially have more accurate timing and dwell control than the missing tooth setup because it uses every tooth on the wheel to schedule dwell and spark timing. When you remove a tooth, any even that would've been scheduled from the tooth that's now gone has to be scheduled from the tooth before... at low revs, a quick change in rpm could result in later start of dwell or later spark than ideal.

It's also a lot easier for the code to sync with the dual wheel, and therefore the dual wheel can be less prone to noise issues.

Ken

 

Mazda did not design it, Denso did. That same configuration is used in other vehicles. Mazda simply adopted it in the design of their engine management. They very well could have adopted or designed otherwise, but sourcing from outside manufacturers significantly cuts down on costs.



That's exactly how I did mine.

 

Ken - This makes perfect sense from a mathematical perspective. With a missing tooth, you're basically increasing the uncertainty of the measurement, and at low revs that uncertainty would be greater. At low RPM's the CAS shaft takes more time to rotate between teeth than it would at high revs, thus the timing error/uncertainty of any event the code is trying to schedule between teeth would have to increase.

Well I think I'm sold on the dual VR implementation based on this alone. Does the current extended MS-II code support the "dual wheel" setup?

 

Yep, also, as far as syncing goes, the missing tooth decoder does everything it can to find the missing tooth (fairly involved in itself), which means it waits until 1 tooth after the missing before it fully syncs (I wrote that code this way mostly for V8's with high compression... the engine speed changes dramatically while cranking, which can throw off more simple sync code). This means it syncs slower.

With the dual wheel, it's not looking for a missing tooth, it's just looking for the second trigger. As soon as it sees a second trigger, it resets the primary trigger count, and declares sync.

MS2/extra 1.0.2 is working great for rotary guys as of right now.

I have a couple of bugs to fix to make it work right for the 2.0 betas... but a fix will be in the next beta.

Ken

 

I've decided to convert to dual wheel instead of modded CAS. It's a good thing I had a spare stock cas lying around. I suppose I could sell my modded one in the FS section