try this.

To get the CAS to work correctly I used a 470ohm 1/4 watt resistor and soildered one end to pin 3 on the map sensor (5v) and the other to the CPU pin 11. Then I ran a wire from CPU pin 11 to X11 which goes to pin 25 on the DB-37 connector.

I also took a 1.2k 1/2 watt resistor and soildered one end to pin 28 on the connector and the other to pin 24.


Run 12v, ground, and the Y off the two cas signals to the megasquirt.
Start the car and see what the tach does on the megasquirt.

All right, I shouldn't have missed that, it was early on in the Miata v Megasquirt thread. I went too far down and followed the link.

OK, so I made another big step. Did what you recommended, added the 470 ohm resistor and 1.2K ohm resistor. Wired the white CAS wire to pin 25 of the DB37 connector. The stock tach works again, and MS reads 1/2 the rpm of the engine. I still get the 20 warnings with Aussie's .msq file, but I loaded it anyway because it wouldn't run with mine. It started, and if I flickered the throttle carefully I could keep it running. Not correct, but more run time than I've had yet.

Maybe I'll build the second MS to this point to rule out a damaged board?

Still kicking,
Kyle

Your probably have it set for a 8 cyl if it reads 1/2. Goto the config page and make sure its set for a 4cyl. You should check the log file to see what the warnings are. Finaly I would reflash the megasquirt to the newest extra and start from scratch. You can manualy copy the setting out of the other log file.

The warnings were a units thing. I changed my units to Celsius to match the .msq file and the warnings went away. All my settings match the settings on the settings sticky exactly. Yet I have to change cylinders to 2 to get correct rpm in MS.

You could go ahead and set it up with the board mods to run both CAS signals in wheel decoder mode. I really don't understand why you would be reading half RPM.

Al is right you should be running both CAS inputs into the megasquirt and running in decoder mode. That will fix your 1/2 rpm problem.

So there are additional board mods to be made for decoder mode, or do you mean the input board mods I already made and just check decoder mode?

No, the mods you've done are all you need. Just connect your secondary tach wire to the secondary CAS signal.

I'm assuming the outer (blue/yellow) CAS wire is primary and it is hooked to pin 24. If the inner (white wire) is the secondary, it is hooked to pin 25. I reloaded MSnS last night. Nothing changed. Should I be posting to the MSnS forum? Will the board mods I made confuse the situation? Man, I can't believe this. The longer these problems take, the more you know it's going to be something stupid. I'm looking through the assembly checking for things that can change based on application. Did you guys use diode D8? Any other ideas?

Dem wires are backwards. White goes to primary tach, blueyellow goes to the secondary tach you added with the input mods.

Where do the terms primary tach and secondary tach come from? Is that MS terminology? Regardless, white to 24, blue/yellow to 25? It's going to be another one of those "I can't wait to get home to try this" kind of days! Thanks so much. :bowdown:

Al, I was going to say the same thing. There are 1/2 as many holes in the inner wheel as in the outher 2 vs 4. Hence 1/2 the RPM. Fulltilt you should
be ready to rock once you get the wires swapped.

So the translation of that last email was, it runs on MS! Pretty good actually. Slight stumble, but not as bad as I would have thought, and it's not tuned yet. Now I'm on to datalogging problems, but there are a lot of things over on MSEFI.com that I can try. You guys have been very helpful, thanks so much! :bowdown:

Kyle

If you have your 02 sensor hooked up. Check out the autotune feature
in Megatune it works well for everrything accept Wide Open. That or an LC-1
would get you spot on.

I got autotune to work and it roughed me in. It was great! I am using a DIY WB. Not long after, something started to go wrong. My A/F spikes to 30 or 40 and stays there. I am suspecting the WB, though certain changes in MT tend to bring it back in temporarily. I'd drop the $200 for a LC-1 if it wouldn't hurt my pride so much. I spent months building that kit years ago!

Since I'm close, I may just put the NB back in and tune it through logs. What bums me out the most is that I had just gone out on the OEM ECM to find out what the stock A/F was at WOT. Does anyone know that? I need to get to a dyno and tune it right, but right now I am just using the standard target A/F chart.

My logs show 16-17:1 real briefly on accel. I dialed in a bit more AE. Turns out I was using RPM based AE, probably because I started with Aussie's .msq. The help file in MT says only 99.9% of all applications will use RPM based accel, yet discussions seem to suggest that it's more accurate. I guess since it runs pretty good on it, I should just run with it, right?

It isn't lean on WOT, I think I'm almost ready to run it on MS at the racetrack!

As long as its around 12.5:1 at WOT you shouldn't have to worry.

I can't decide if this is a dumb question and I'm missing something fundamental. If I'm using a NB sensor and the VE table is set to run my car more rich than 14.7:1, is the NB really doing anything, or am I just always in open loop?

The NB is only accurate around 14.7 but the MS treats the input the same as a wideband. So, you could get yourself in trouble tunning WOT with a NB sensor. Its fine for crusing and idle though.

Contrary to the popular belief, a narrowband can be very accurate and repeatable at the top of its' curve and can be used to hold an AFR of 12.5 very nicely. It is a chemical reaction and the curve is not nearly as steep at the top.:bigtu:

This is not necessarily true of "NB Simulators" provided by some WBO2 manufacturers.:ugh2:

http://mymiata.paladinmicro.com/Miat...tm#O2SensorHow