OEM 36-1 Wheel
 

I thought it would be a good idea to post some information from my build thread here so it would be easier to find.

More details on my build can be found here.
https://www.miataturbo.net/showthrea...205#post846205




I was originally fabricating my own 36-1 ignition trigger. This design used a wheel that brazes to the damper hub.

https://www.miataturbo.net/attachmen...ine=1332754515

https://www.miataturbo.net/attachmen...ine=1332754515

https://www.miataturbo.net/attachmen...ine=1332754515




Then while doing some research on various Mazda engines I came across a 1.6L ZM-DE engine in a 2000 Protégé. The engine block looked remarkably close to a B6/BP so I did some disassembly to inspect. What I found is that the ZM-DE appears to share all the same dimensions with the B6 engine, except it has an aluminum lower split-block design and two extra oil drains on the exhaust side of the block/head assembly.

I then came across this while removing the lower timing belt drive gear from the ZM-DE engine.

https://www.miataturbo.net/attachmen...ine=1332754515

I took and laid a 99’ crank trigger wheel I had over the top of the ZM-DE trigger wheel.

https://www.miataturbo.net/attachmen...ine=1332754515

https://www.miataturbo.net/attachmen...ine=1332754515

https://www.miataturbo.net/attachmen...ine=1332754515

O.D. and the center bolt pattern appear to be identical!

https://www.miataturbo.net/attachmen...ine=1332754515

https://www.miataturbo.net/attachmen...ine=1332754515

https://www.miataturbo.net/attachmen...ine=1332754515

https://www.miataturbo.net/attachmen...ine=1332754515





I contacted my local Mazda dealer to get a part number and pricing.

#ZM 01-11-408
$21.68

https://www.miataturbo.net/attachmen...ine=1332754515

Superb work!

Wow!! awesome!

surprised this didn't get more reactions. Seems like a very easy way to replace the 4 tooth wheel on a NB with a 36-1 wheel. Lots more resolution.

I have one and will be using it with MS3 :)
Haven't 100% figured out the input mods but it should be straightforward.

I plan to get one later on as well

Thread stickeyd, and added to the MS "Super Important" list.

Thanks for the clever detective work, Matt. This is really phenomenal.

Thread o the month.

Brilliant detective work! Figured I'd add this to the thread, looks like the same part and the numbers match:

http://onlinemazdaparts.com/parts/20...icleid=1363777

OnlineMazdaParts.com - 2000 Protoge 1.6ltr / ENGINE / ENGINE & TRANSAXLE / ENGINE PARTS / Crnkshft pulley plate, ZM 01-11-408 (#2 in diagram) $12.36

I'll probably just order one to play with and confirm, I've got a few other little parts I need to order as well ;)

Ordered one, and will report if the stock sensor reads it and communicates with MS3.

Received the part, diameter holes and slight offset as on original:

https://www.miataturbo.net/attachmen...ine=1336014762

Installed and sensor should be able to read teeth

https://www.miataturbo.net/attachmen...ine=1336014762

https://www.miataturbo.net/attachmen...ine=1336014762

While I was checking fitness, I noticed that the missing tooth passes the sensor ~90 degrees before #1 cylinder is at top dead center.
Being new to Megasquirt,can this be handled in software or should the gap be at #1 Top Dead Center.

I'll be doing some reading tonight, but I thought it might be of interest to someone to get this out and open for discussion.

:edit: It appears that Megasquirt likes to have the missing tooth 9 teeth before TDC of #1. For 36 tooth wheel 9 teeth is 90 degrees, so I think the gap is right where it should be.
hrk

Nice, glad to see some confirmation. What vendor did you source it from?

In for "explain it to me like I'm 5 Hustler" instructions for setting this up in my DIYPNP :bigtu:

Yeah you can set it in software, see:

http://www.msextra.com/doc/ms2extra/...Miss_Tooth.htm

Just work out how many teeth between the gap and TDC and then work out your angle to plumb into the software :)

Looks like this place has it for 15 dollars
http://www.jimellismazdaparts.com/pr...ZM0111408.html
looks like around 23 dollars after shipping, for those who don't have a local mazda dealership.
didn't see the other for 12, dollars, oh well

so is there a full mazda part number for this? and will it work with a NA8C engine?
Here in the UK the mazda dealers don't like ordering parts unless full part number of VIN given,

Dan

Sub'd for this, need spoonfeeding please. Looks like an awesome upgrade.

Wiring the sensor in -96 model.
 

Question about wiring to MS3 and sensor from -96. I believe my car has the original sensor from -96 donor, there is a bit of metal visible in the middle of the sensor as seen in the pitcure above and it appears to be magnetic, so I assume it is VR type. The wiring diagrams I have found so far show different color labes than the actual wires in the car.

Sensor has three wires at the connector behind 8" pigtail.
Red
Green
Black

https://www.miataturbo.net/attachmen...ine=1336654679

And they connect into
cars wiring harness as
Red -> Green/Black
Green -> Green Yellow
Black -> Black/green

According to wiring diagrams there should be Yellow/Black and Yellow/White and Black/Light Green wires coming out from this sensor.

https://www.miataturbo.net/engine-performance-56/na-nb-crank-wheels-same-34986/

Black/Light green appears to be the ground shield in both original car and wiring diagrams. Two other wires disappear into original ECU pin 4E and 4H.
From internets I found names for these pins.
4E Crank miss-fire sensor N/U
4H Timing ground for crank pickup

I understand that VR type sensor creates voltage on its own, and that would make sense per original wiring diagrams.

The real question I am trying to get to is how to get this sensor connected to MS3 or MS3X?

Currently I have it wired like hall sensor:
+5v from Megasquirt throttle output to the Red at the sensor
signal ground to the Black shield wire at the sensor
Pin 24 on MS3 to the Green at the sensor.

I think it is not correct.

I am using DIY wiring loom and it has nice shielded wire for crank sensor which has tiny white and black wires.

hrk

Your car's wiring puzzles me. Is this a US-spec car, or is it from a different region?

At any rate...

You do not need +5 from the ECU with a VR sensor. One of the VR sensor's active wires will go to the ECU's VR input, the other will go to the ECU's ground along with the shield wire. You will need to experiment to determine the correct polarity (which one of the green wires goes to ground vs. which one goes to the ECU's VR input.)

From your description,
"4E" pin is wired to the white wire inside the shielded cable
"4H" pin is wired to the black wire inside the shielded cable
Black/green can be connected to ground or to the shield conductor on the shielded cable

As Joe pointed out, the sensor polarity as used on the stock computer *may* be backwards for the MegaSquirt. You have a 50/50 shot. You need an oscilloscope to determine this ahead of wiring. Without a scope, you can wire it up and use MegaSquirt's toothlogger function to determine if polarity is correct. If wrong, just reverse the wires.

Thanks guys,

The car is US spec -96, and this part of the wiring harness should be as it came from factory.
I had Emanage Ultimate there for ~5 years and now rewiring the whole thing for MS3. I never had to touch this part of wiring. Of course there is a chance that I picked wrong connector from the bundle of wires now laying on the floor, but it is one of the last connectors in the loom, relatively close to TPS and idle valve connector, and it fits physically. It has been a long week in the garage.

I will try one way, I presume there no danger of damaging the sensor or with the small voltage VR sensor puts out while cranking.

Thanks again, I will report how it goes. I think it will be very easy 36-1 conversion for NB Miatas with stock parts, once the wiring is figured out and the sensor can communicate singal to software.
hrk

The sensor you are using may be a VR and Not a hall effect sensor, I was working with Joe in this thread of what needs to be done to use a VR sensor
https://www.miataturbo.net/megasquirt-18/looking-ms1-settings-36-1-trigger-65675/

Or as it seems you can use an OE NB sensor and use the same hall effect circuit already built for the CAS.

There is indeed a danger of damaging the MS if the output of the VR sensor is connected to an input pin other than the one configured to accept a VR sensor. These sensors are capable of generating voltages higher than Vcc, and the VR input circuit is configured so as to isolate this voltage from the CPU and other ICs within the MS.

If you're asking whether the sensor itself was damaged by the application of +5v? Probably not. I've never measured the resistance of a VR sensor's coil, but I assume it's high enough to prevent a significant amount of current from flowing through it, especially if the other side was not grounded.

Joe, I ment that with cranking speeds I presumed no harm to Megasquirt.
No power was turned on to the sensor. I wired black and white wires from DIY autotune harness into respectively green and red ones in the sensor. The shielding ground was connected to signal ground, but I think it might be better to ground it to engine block instead, at least in theory.

Car starts and idles with sequential ignition and injection. The gap between sensor and wheel needs to be checked when changing the wheel, the od appears to be the same, but in my case there were slight contact between wheel and the sensor at some teeth. Also the sensor is pretty sensitive, it will not see the targets if there is nmore than 2 mm gap.
So for ~$20 wheel you can get the resolution you want with factory parts and no bracket or wheel design and fabrication needed. I presume it will work as long term solution as well as the stock one. My sensor is quite probably original from -96.

hrk

Cool.
Apparently Flyin miata is developing a 30-2 wheel as well, so now there will be several options for increased timing resolution.

I think I now posses the knowledge to do this without guidance. :giggle:

Very nice find! Sherlock Holmes would be proud. Added to the to do list.

So, is anyone now running successfully? What was the angle offset compared to the nb wheel/trigger? I'd like to run the same on my MSM w/Hydra. The 4 tooth setup makes me feel like I'm flying an F22 with wooden chariot wheels as landing gear.

+1 on the spoon feeding. I have a 00 with a Rev built MS2. Can someone post a step by step of the settings that need to be changed to make this work?

I am also looking for someone to post the settings in ts with the new wheel

Another nice wheel option -> New FM 36-2 trigger wheel for Hydra ECUs - MX-5 Miata Forum

Thanks to from Soviet and Reverant about the cam sensor mods required to do this on an NB. To use this trigger wheel you need to grind off the dual sensor nubs by the "E" on the cam pulley. You also have to make sure that the cam pulley sensor the cam sensor is NOT over (or very near!) the single cam tooth on the cam wheel when the crank sensor is over the missing tooth on the 36-1 wheel. The crank trigger wheel can only be installed one way but the cam pulley can be positioned in three different positions on the cam.

https://www.miataturbo.net/attachmen...ine=1316890297

Has anyone with an NB tried this mod and repositioned the cam pulley? If so, how was the cam pulley placed and what settings in MS are required?

So does that mean the setting in the picture would be correct (with the two nubs ground off)?
Did Revrant say anything about changing any of the settings?

The pic just shows the cam pulleys at TDC. I don't know the MS settings needed nor how to reposition the intake cam pulley.

Wiggy. I guess it's all in the decoding, then? Cause, anything like that would make VVT more annoying. I guess not crazy, but grinding the cam feels.... permanent.

Think I'll be playing with this one later, in a couple months, when I gotta figure out how to support both VVT and 36-1.

I ordered this part.
Somehow I cannot get it to fit.
The wheel interferes with the plastic TDC reference point. That can be trimmed though.
Also the dished out part is not deep enough for my pulley. There is a gap between the wheel and the balancer. Quite a big gap. The wheel runs into the rubber part of the balancer, but even with that trimmed out, here is not enough room to get it to fit. How on earth did you get that wheel on the stock balancer ?

I think you have to use the later style balancer to get it right. To make it perfect, is there a way to upgrade the whole belt assembly to serpentine on the big nose B6 engine ? What parts are involved ? It would be nice to swap the alternator pulley, the waterpump pulley and the cranckshaft pulley to serpentine + ability to fit the triggerwheel. This would be the ultimate upgrade.

I have measured this wheel and the OE NB wheel, here are the angles:

Crank sensor is mounted at 110 degrees.

OE crank trigger plate:

First tooth is at 190 degrees on plate, so triggered 80 degrees BTDC
Second tooth is at 120 degrees on plate, so triggered 10 degrees BTDC

Per 180 degree period:
80,10

Protege 36-1 trigger plate:

First tooth is at 190 degrees on plate, so triggered 80 degrees BTDC
Second tooth is at 180 degrees on plate, so triggered 70 degrees BTDC
...

Per 180 degree period:
80,70,60,50,40,30,20,10,0,170,160,150,140,130,120, 110,100,90 (90 missing on 2nd period for reset)

The OE wheel has teeth that are 1 degree wide, and on the Protege wheel they are 3 degrees wide. With my ECU I trigger on the trailing edge of the tooth, so it will be 2 degrees out unless adjusted.

Would fitting one of these allow one to just use a single crank angle input with wasted spark without the cam position input. Making ms easier to set up and also preventing timing from being effected by cam belt stretch etc. Would this bolt onto a 1.8 bp engine?

Timing is not affected by cam belt stretch, as the cam sensor (as used on the 99-005) is only there to tell the ECU the current cylinder phasing, and not the crank angle.

Without a cam sensor, you can't run sequential injection or ignition.

Is there any real world benefit to running sequential injection and ignition. I'm halfway through building my MS2 and its my first foray into electronics so any thing that makes it simpler is a good thing. Also I have a standard stim not jim stim

Sequential injection will make the car idle better, have lower emissions and snappier throttle response at low revs.

Bringing this back from the dead. I am going to install this on my 1.6 engine, using the 2000 miata crank sensor. Does the 1.6 oil pump have the hole for the sensor bolt there?

Good question... it may be there but its probably not tapped.

https://www.miataturbo.net/attachmen...ine=1391488641

Ah you just reminded me, I've got an aftermarket pump for 1.6 laying around:
https://www.miataturbo.net/attachmen...1&d=1391489336

https://www.miataturbo.net/attachmen...1&d=1391489175

It looks like you are right about it being untapped. M6 is the correct size I'm guessing? I hope that hole is there on my OEM pump to make things easy.

Has anyone tried this, I did and now my car does not run good. I'm not sure what went wrong during the change. My old tune didn't work after replacing the wheel. Im running MS2

Yes. Works great.

I pulled this from an old email, maybe it will help.

CAS is a cam angle sensor that's mounted on the exhaust camshaft at the back of the head on 90-97 miatas. The exhaust camshaft has 2 notches on it where the CAS fits. Inside, are 2 trigger wheels, one with 2 teeth and one with 4. The 4 teeth is the crank signal and the 2 teeth is the cam signal (one tooth is bigger than the other).

By simply cutting off some teeth inside the CAS you can turn it into a "1 signal per cam revolution" that Megasquirt needs.

None of this is relevant to you. For NBs the crank and camshaft signals are already separated. The cam signal I believe is 2-1 pulse. That is, on the camshaft wheel there are 2 "nubs" on one side and 1 nub on the other. So it goes

_||_________|_________||_________|_________||_____ ____|_________||

If you grind off the 2 nubs you will also end up with the "1 signal per cam revolution", e.g.

____________|____________________|________________ ____|___________

Now for Megasquirt settings. First, I would recommend getting the car to run with JUST the trigger wheel and add the camshaft sensor later. You only need the camshaft sensor to run full sequential. So you will *have* to run wasted spark/batch fuel to begin with. If you already wired for sequential, megasquirt 3 can still handle it - just use "Wasted COP" mode.

For ignition settings
- choose "trigger wheel"
- "Single wheel with missing teeth"
- 36 teeth, 1 missing
- the missing tooth offset, I think, is 90*

Once its running confirm the timing with a timing light and confirm you have the rising edge/falling edge correct. If the timing is not right adjust your missing tooth offset a bit.

After its running in batch mode, to utilize the camshaft sensor, simply enable "Use CAM signal" option. Then you can switch to sequential.

good luck,
-JB

Thanks a lot guys. I will try and see

I talked to DIY a week or two back, they said that the 36-1 + VVT support had been pretty much abandoned due to a lack of perceived gains with doing so.

While I'll be the first to admit that 36 teeth might be more than is needed.... has anyone looked at their jitter with the different wheels?

Spark jitter on the NA CAS was pretty bad at high RPM, but from what I've seen it's pretty minor on the NB.

The only real justification I've ever heard for 36t over 4t was from Jason C hypothesizing that a high tooth count wheel will exhibit slightly less predictive error during moments of extremely high transient ∆-RPM.

Interestingly the worst case for timing error when you have few tooth is at low RPM. You can think of it as because the sampling rate is low at low RPM. On my AEM FWIW which only effectively times off of 2 teeth with the factory trigger wheel, the error was on the order or 5* when blipping the throttle off idle. When I had 6 effective teeth, it was ~2*. With a 12+1, it was unreadable. Idle got noticeably smoother going from the factory to the 6 or 12 tooth wheel.

The worst case at high RPM I think is when you do a hard upshift and the clutch drops RPM very quickly. The rapid crank deceleration would cause the timing to be over advanced with few teeth.

I would think going from a 12+1 to a 36 tooth wheel would be very return-diminishy.

How about for the 36-2?

There is no difference between the stock 4T and the FM 36-2 with an MS2/MS3.

The spark angle is rock-solid (with a timing light) at idle and when bliping the throttle, up to 5K RPM.

Did you try the test even at say, a fixed 15* advance? I've found that it's significantly worse when the requested spark event occurs just before a tooth (which is at 10* BTDC for the factory trigger wheel).

Im thinking about going back to 4T. I either don't know what im doing or MS2 is not liking 36-1. Does anyone live in the dallas area that can help? I will drive to you

Yeah - that's interesting. I've always had a gut feeling that 12 tooth is the way to go - more than that and you're spending all your time worrying about predictable things anyway. Mine's been reasonable at OEM 4 tooth, but I haven't really beat it up. The CAS was dismal which I blamed on the belt.

FM 36-2 with NB cam trigger is supported. 949 Racing 36-1 with NB cam trigger will be supported in the future (when Emilio sends me data captures).

I have a FM wheel on my car. I am unable to perceive a difference in spark timing as compared to the original 4t wheel, nor am I able to perceive improvement in dyno results. I do believe that there is gained resolution with the valve timing with the extra crank teeth however, as I noticed much error in VVT Target vs actual after the install, and had to retune the PID. The Miata's VVT cam trigger is pretty low resolution.

Why not a 12+1 with VVT? I have posted the timing diagram before.

Re: 4-tooth accuracy. Have you tried blipping the throttle off of idle with the timing set to 15* BTDC?

There are only so many wheel patterns that it makes sense to support. We already support 2 other patterns with a 3rd coming. Why do we need 4 different Miata wheel decoders? My human eye can not detect a difference between stock and 36-2 crank triggers under any condition.

Only because TSE sell a 12+1. I was under the impression it was popular.

Was this blipping the throttle off idle? Was it with the timing set to 15* BTDC so that the requested spark even is *just before* a tooth arrives?

I believe its the cat's pajamas with the AEM.