Ok guys, the FSM isn't really clear on this one and I don't recall what ground truth is, so I'm looking for verification. Below is a photo of my OEM CAS trigger wheel; in the yellow circle you can see one of the 2 timing mark notches, and the one in the light blue circle the 2nd timing mark notch. I painted each notch white to make them easier to see with a timing light years ago. Per the FSM, one of these notches should represent -20* BTDC and the other -5* BTDC when lined up with the pointer on the front cover. Based on the clockwise direction of rotation when running, I think the yellow circled mark is -5* BTDC and the light blue -20* BTDC. Please confirm or correct me - need to be 100% sure as I'm calibrating the base timing on a new ECU.

 

The time mark is at 12 o'clock with the white strip. When aligned with the timing pin:

• Trailing side: 20°ATDC (-20°BTDC)
• Leading side: 5°ATDC (-5°BTDC)

 

What CarlosIglesias has said above is correct. There is only one timing mark on 13B-REW's. The mark indicates -20 BTDC on rotor 1 trailing plug and -5 BTDC on rotor 1 leading plug. The thing you circled in blue is nothing of importance.

Not sure what ecu you have but you typically you will need to lock the timing into either -20 or -5 then test using the correct corresponding spark plug lead. My Fueltech has a special mode to lock the timing to a set position to set the base timing. Some ecu's you have to temporarily set the ignition map to -20 or -5 to perform the base timing calibration.

Why Mazda this timing mark as the indicated index makes no sense to me. Why choose a timing value that the engine will be very unhappy running at? Seems like choosing 20 BTDC trailing/5 BTDC leading to indicate on the pully would fundamentally work much better. Its a bit of a pain to keep the engine running when you set the timing lock at -20. This is what we are stuck with though if still utilizing a factory CAS system.

 

Agreed! Mazda's OEM timing marks SUCK when setting base timing on an aftermarket ECU. At any rate I got it done thanks to Carlos for confirming the marks! I'm setting up a Link G5 NEO4, and their procedure for the 13B-REW involves temporarily setting the entire trailing split map to -20BTDC, turning off fuel & use the PCLink tuning software to lock the leading timing at 0*BTDC. Then crank away with the timing light connected to the T1 plug wire - if timing marks don't line up with the OEM -20BTDC mark, tweak the "timing offset" setting till it does & save when complete.

That 2nd mark (blue circle) on my OEM wheel LOOKS like it came that way from the factory but I can't be 100% sure - it's a divot on the inner part of the wheel that looks just like the -20BTDC mark, and if i run my engine and let it suffer through idling at a locked -5*BTDC, that mark will almost perfectly line up with the front cover index mark when I connect the timing light to the L1 wire.

 

not to be trusted, given how critical timing is on a rotary it’s why I bought an Xtreme Rotaries TDC tool …

 

How does that tool work?

 

It attaches to the dowel pin positions on the rotor housing side of the front iron and is machined to fit and lock the e-shaft front rotor eccentric in the TDC position on the front iron. Which along with front cover and e-shaft timing pulley in place allows you to verify the TDC position. So this has to be done with the keg torn down. From there the other marks are easily verified with some simple math or angular measurement. They do sell another tool for easily marking out/verifying those other positions, but the cost for it wasn’t worth it to me. Even the cost for the TDC tool has increased quite a bit since I purchased it, along with the shipping tariff now depending on where it’s purchased.

https://m.youtube.com/channel/UCxt31VQj3AvK76bz_LwOIBg

There is a procedure for determining TDC on an assembled wankel engine, but not as accurate imo. There tends to be some slack on the exact eccentri position on an assembled engine due to clearances etc. It’s not as locked in as a piston in a reciprocating engine. You can google for that procedure if interested.
.

 

Interesting. How far out have you found the factory -20 mark to be? I guess I'm not understanding why Mazda would not use similar or better means of marking the factory pulley.

 

^Yeah, that tool wouldn't help with the question posed in my original post. Tearing down a perfectly good 13BREW just to verify TDC is a bit extreme. Verifying that the mark on my OEM Mazda trigger wheel is in fact at -20*BTDC is good enough for my purposes.

I'm wondering the same thing... Are we talking whole degrees of error with the OEM timing mark or some fraction of a degree? Suspect it's the latter, and also suspect that a fractional degree error in the TDC mark won't matter much with either an aftermarket ECU or the OEM ignition system, since physical resolution of crank position is ultimately limited by the # of trigger wheel teeth.Whether you're using a VR or Hall type sensor, a 36 tooth wheel gives you 10* of resolution (360/36 = 10), the OEM 12 tooth wheel gives you 30*

 

I haven't tried this on a rotary, but I've helped add a timing mark to a piston engine. We would find TDC by slowly rotating the crankshaft while using a long piece of welding rod in the spark plug hole to see when the piston changes from moving up to moving down. That's probably a two-person job on a rotary, because someone would need to hold two welding rods (in the leading and trailing plug holes) and watch both of those welding rods carefully to see if the face of the rotor is lined up or not. I would try it on the rear rotor first, since the front spark plug holes are obscured by the power steering lines and the wiring harness loom. Once you have your TDC mark, you can do the string trick or wire trick where you cut a length of string to the circumference of the pulley and then divide that in half enough times to get marks at 180 / 90 / 45 / 22.5 / 11.25 degrees. Or just take the 30 degrees distance using the crank trigger teeth and divide that by 3 to get 10 degree marks.

I've been running my car for years after using a timing light to sync the ECU so that -20 degrees in the ECU software lined up with the factory -20 mark. I think I set the leading timing to 0 degrees and the timing split to 20 degrees, to keep the math simple. I've thought about adding more lines on the pulley to show things like +10 degrees, but it runs OK so I've never bothered. On a new install, I would check all four coils to make sure that the leadings and trailings are configured and wired correctly. If you only check one coil there are three other coils that could hurt your engine if they happen to be wired wrong or configured wrong.

Only tangentially related to standalone ECUs and ignition timing, I've never tried to go aggressive on the ignition map but I did see some knock sensor activity at low RPM low boost in some cells that I had too-low values for trailing ignition split. Changing those cells back to the commonly-recommended 8 or 10 degrees of split made the knock sensor quiet down again.

 

I just don't get how the tool actually achieves anything. The crank pulley is indexed so it will only install one direction. That would mean that the factory would have had to incorrectly mark the pulley for there to be an error. Is this tool meant to be used for custom/aftermarket pulleys that are not indexed or indexed in the traditional orientation? That would make sense. Hoping TeamRX8 chimes back in not that I am curious.