FatKao

I'm building a MS3 on a 3.0 board for my new (to me) 2001 motor using the 1997 NA sensors (CLT/CAS/TPS) in a 1997 body. VVT left unconnected at the moment waiting for DIY's board to come out.

Board was built by the megamanual assembly guide directions and referring to the MT build thread when there was decision to be made. I added the spark inversion circuit right from the start and that appears to work fine. The only change I made to the daughterboard from the instructions was to move the filter and pullups from JS8 to JS10. (http://i29.tinypic.com/10mn9eq.jpg) JS10 and MS3X were the only options in TunerStudioMS for the cam signal input so I matched my board to what it let me do.

The problem I'm having is that I can't sync to the 4g63 signal from the JimStim (DIP 2-3). I get a RPM that bounces from 0 to some number that varies depending on the RPM *** location. I can sync to the 99-00 signal no problem (DIP 2-3-6). I have the 2nd tach input running to I1A, 5v pullup on 2nd trigger, 12v on primary trigger.

Now, if I invert the second trigger (Not true to the Miata CAS) and pull the 2nd trigger jumper on the stim I can randomly get the MS3 to sync to the signal when I reapply the 2nd trigger jumper.

I have a 2 input scope available to me for testing, but I only have one test lead in my possession. The motor is also on an engine stand currently so I could pull the CAS and spin it by hand if anyone thinks that would be helpful. From what I saw on the scope the Jimsim seems to be outputting a true to life signal (Can't check the phasing of CKP and CMP though since I only have one input working on the scope).

Any ideas? The fact that 99-00 works makes me think that electrically everything is okay (Pulling the cam signal makes it lose RPM sync). MS3 firmware is ms3-beta-0.27 and TS is the latest version.

Braineack

iTrader: (62)

you are using the MS-I input mods. you need to do the MS-II or MS-III input mods...


why are you doing a CAS setup!?!

FatKao

Autox/TT classing. I have to use OEM parts. Glad to see there was something obviously wrong. I'll get everything reconnected the correct way tonight.

As a side note, is that referenced to in the guide? I thought I looked at it pretty carefully and I'd feel pretty dumb if I missed that.

Braineack

iTrader: (62)

it's on the miataturbo.wikidot.com page, not here, it should say it's MS-I only.

so why not use the stock NB sensors? fitting a CAS is going backwards.

FatKao

My motor didn't come with any sensors so the NA stuff is what I had on hand. I'm guessing that the NB sensor gives better precision than the NA one due to the differences in the signals?

And wow I feel like an idiot for missing that one.

Braineack

iTrader: (62)

yeah NB sensors would be very precise over the CAS since it's driven off the timing belt.

FatKao

That did it, syncs to the signal no problem now. Thanks for the help.

Braineack

iTrader: (62)

glad to be help.

webby459

iTrader: (8)

We are having this same problem. Car is a 90 with wasted COPS, CAS with a 99 motor. The plan was to run it with the exact stuff the car has now and then upgrade to 99 crank/cam trigger, then full seq. It runs on MS1 V3 board and hires. The MS3X is a totally separate unit, my friend built it, basically using the full assembly instructions in the manual. We used an interceptor board to run the car off the MS1 while the MS3 was watching the inputs.

All the input mods were done as per the pics in this page, our MSQ is attached, we used the 4G63 mode, didn't matter if we switched to falling or rising edge. It also didn't matter if we switched from MS3X cam in to JS10. The rpm showing from the MS3 would fluctuate when the car was idling, zero to random number as FatKao reports.

And yes, we are now switching to 99 crank/cam input gear and probably giving up on the CAS. But atm we are sort of curious what we were doing wrong? Also, I haven't searched yet and it might be well reported, but on a totally tuned CAS spark map will I need to use spark latency if I'm running 99 trigger wheel? Anything else I should know?

Ben

iTrader: (33)

I use the VR inputs with pullups for both NA and NB and don't have any problems after dialing the pots in. This is the way that is suggested in the manual.

Going from NA to NB sensors should not affect your latency setting. Depending on your coils, 30 to 60 is common.

WestfieldMX5

The VR circuit on the V3 board is slightly different than that on the MS3X (different opamp), does it make sense to modify the V3 circuit to that of the MS3X?
Or do you use the 'standard' V3 circuit as is for CKP and the MS3X circuit for CMP?
Pots all the way counterclockwise on both circuits?

Edit.
V3 circuit has no pullup. Do you add a 2K2 like on the MS3X?

Ben

iTrader: (33)

You're over complicating it Frank.

Go to:
http://www.msextra.com/doc/ms3/ignition.html#tachint
Follow all steps under "V3.0 board - VR Input + pullup"

That will address CKP input. CMP input wires directly to the Expander.

WestfieldMX5

thx Ben, found that page right after posting. Forget to edit
I guess the stock circuit with 1K pullup will do just fine.
You can expect my MS3X order any day

Ben

iTrader: (33)

Yup, I suspect pretty much any reasonable resistor value will work. Last one I built myself used a 470 ohm because I had an extra one on the table already. Advanced thanks for your order.

WestfieldMX5

Order arrived. It was in belgium on sep 23rd (3 days after ordering) ... and was delivered to me this morning, after 19 days in customs and a $150 bill.
Can start building

WestfieldMX5

Car is up and running. Both VR circuits (V3 and MS3X) are working fine with pullup. Very easy build.

WestfieldMX5

Out of curiosity. Does this work on a NA as well? CKP and CMP from CAS through the 2 VR circuits (MS3 and MS3X) with a pullup?
If so, why are folks not using the VR circuit on the NA builds (or building a 2nd opto circuit)?

Braineack

iTrader: (62)

I believe so; although, I've still yet to have issue with the 2nd trigger going straight to JS10 with a pull-up.

Hell, even the NB CKP signal should be able to go through the inverted VR with the NB CMP go directly to JS10 with a pull-up, or if not, through the onboard opto circuit...

Joe Perez

iTrader: (8)

Yes.

From an electrical standpoint, the sensors in the NA and NB are virtually identical. Any circuit that works on an NA "should" work on an NB, and vise-versa.

The primary reason we started designing different circuits for the NB is that it was found that they tended to have somewhat noisier signal paths than NAs. The "simple" circuits commonly specified for use with the NA on an MS1 have virtually zero hysteresis and noise-rejection, and were never a particularly good idea in the first place, they just happened to be extremely simple to implement.


Because years ago, somebody who wasn't very good at designing circuits said "wire it this way," and everyone did. Old habits are hard to break.

The opto circuit on the MS was never designed for logic-level triggering in the first place. Its original purpose was to interface to the coil (-) wire on old V8 engines with distributors, way back when MS was a fuel-only controller. In that environment, you need the kind of isolation that only an optoisolator can give.

For modern logic-level triggers, the circuit specified as the "VR" circuit is far better, as it employs both low-pass filtering (resistance to noise) and hysteresis (resistance to change.) It is not the absolute best design possible, but it is highly superior to the instructions specified in most of the common MS1 writeups (including the one on DIY's website), regardless of the year of the car or the version of the CPU.