Crank Wheel & VR Sensor
 

Installed crank wheel and VR sensor. I can watch the teeth get counted in TunerStudio, and I have a tach reading. Car acts like it wants to start (has spark and fuel), but just won't catch.

How can I tell if my VR sensor polarity is reversed? I know I can go swap wires around, just didn't want to damage perfectly good wiring with a a cut n splice if I didn't have to.

Also, the megamanual set up instructions differ from the Trigger Wheel Settings dialog box presented. I understand you turn the crank wheel such that cyl 1 is at TDC, then count the number of teeth between the missing teeth and the VR sensor. You enter this data for "tooth #1". I get that. But in the manual, they show using the tooth count. The dialog box in TunerStudio (and MT) look like they're asking for crank degrees. Input?
I tried both... Same result: No start.

http://img248.imageshack.us/img248/2662/38790736.jpg

Also how do I know rising edge or falling edge? I haz no scope. (I tried both w/o success, though one way--I forget which-seemed like it was more apt to start).

Ben, I can't help you with this one, but my only recommendation is that since you are walking in somewhat uncharted territory, you should in fact get a scope, even a single channel 10Mhz unit will do.

Jim

Talk to patsmx5. He's running the same sensor and wheel on his 99 with MS and it works good afaik.

not that I know anything about VR sensors, and it may just be the picture...but your VR sensor looks really far away from your trigger wheel compared to what i've seen on other cars.

Rising / falling is not adjustable via software insofar as the primary trigger is concerned. The signal is going into the hardware IRQ pin on the CPU, which is fixed at falling edge detect (as measured at pin 14 of U1). Only the second trigger, which you should no longer be using, is adjustable in software for rising / falling.

You want the sensor and decode circuit configured so that the trailing edge of the gap is a falling edge. You will need a scope to visualize this. It should look like the yellow trace below, assuming an MS1:

http://img262.imageshack.us/img262/7...nvertedfh3.gif


I haven't really used TunerStudio yet, but basically you need to enter four tooth positions plus an offset in degrees.

The tooth positions tell it which tooth to use as the reference for both TDC and BDC, as well as which tooth to use as the "return", which is the one it'll fire on during cranking if you have cranking set to trigger return. The offset in degrees, if I recall correctly, tells it how far away from TDC / BDC the teeth you declared as the reference teeth are.

This is the least incorrect set of documentation on the subject I know of: http://www.msextra.com/manuals/MS_Ex....htm#wheeldecr


Nice looking setup, incidentally. Where'd you get the hardware?

+1

I'm looking to do the same thing with my new motor. I figure since my new motor will be out, I should do everything the right way before putting it in.

I also want to do a writeup on the trigger wheel and MS-II construction for my 92 motor a la Brain's writeup so other people don't need to endure the pain.

Good eye sir. I took the picture before adjusting the sensor towards the wheel. It's now set with a gap equal to the thickness of a business card. TunerStudio has a logger where you can watch the MS read teeth, so I can confirm it reads the teeth and the missing teeth, and I have RPM registering on the dash tach.

http://www.megamanual.com/ms2/misstoothexample.gif

Joe, the MS2 set up is a bit different than the MS1 set up. All you need do is enter the number of teeth on the wheel, the number of missing teeth, turn the crank to cyl 1 TDC and count the teeth between the missing tooth and the sensor position*. The rest happens automajically.
*The part that's troubling me is the ms2 documentation show entering in this count in teeth, and in TS (Joe you should use TunerStudio; you will like it a lot) the dialog box asks for degrees. They call this "delay" teeth. So if I have 6 delay teeth on a 60 tooth wheel, do I enter 6 or do I enter 36 into the dialog box? Well I tried both, and neither starts the car. The same car that fired up immediately yesterday on the factory wheels.

The set up came from Frank's former Motec (now V8) car. I don't know who made the components originally, but they are very cool.

If I can get the damn thing to start, you won't have to do that.

At this point, I'd say copy Patsmx5. He is running a Ford EDIS system. Once bolted and wired, using the EDIS system is as simple as picking "EDIS" from the ignition type menu.

Bah. I don't want to buy a damn scope.

I reversed polarity on the VR sensor leads. No change.

Confirmed that you do enter the delay teeth value in degrees.

Yes, I'm that damn stupid. Ugh.

Car runs.
Batch fire FTW.

And if I'd have looked at your sig, I'd have noticed you're using an MS2. :retard:

Sorry, I've been helping another guy plan his MS1 VR setup, and have that set of code in my brain right now. Totally didn't even think to see if you were on MS2.

But congrats.

I do want to get into TunerStudio, I just haven't been home long enough to do much of anything recently. It will be necessary, as Jessica will be running full sequential when it goes back together.


BTW, buy a scope. Until you have one, you are half a man. :brain:

I have no issues with buying a scope. I just would rather pick one up when I can snag a deal on something cheap.

I take it by another post of yours that you don't think an NA CAS on the back of the exh cam for the 2nd trigger will introduce any meaningful error.

I don't think the acuracy of the 2nd trigger really matters at all. The ecu just uses it to determine what cycle it's in. As long as it gets the signal before the cycle starts, it should be ok. All the timing is still based off the crank trigger. At least that's what I think.

I use a CAS on the exhaust cam along with a 60-2 crank wheel with my Haltech to run sequential injection (on my Protege).

If you are that worried about error, you should be mounting the crank trigger on the rear of the pulley so that it's not "damped".

--Ferdi

Dammit Ferdi, don't do that to me. Now I'm going to lie awake all night worrying about timing inaccuracy due to the harmonic damper. ;)

I believe you are correct regarding the second wheel. There's a fella on m.net who modified an NA magnetic CAS to replicate the output of the NB cam sensor. I think I might give that a shot, but use both wheels to output one to each computer. That would enable me to run a cut cam cover and install adjustable gears. WOOT.

Ben,
You're not threaded in on the inner wheel? Only the out bit is rubber. I think you should be fine. I welded one to a friend's wheel, worked ok till he switched over to my little card.

Did you get up and running, then, with the stock sensor? You're running parallel for emissions? I would STRONGLY suggest using the stock sensor. Its there and ten minutes of wiring should have you working. Even if only for the cam signal.

One more thing: I haven't done it, but I know folks who have - you can get software oscopes - you use your computer's sound card, plug a headphone into the mic or line in port and cut the wires off. Should get two channels of at least 40 kz bandwidth that way. Enough for looking at crank teeth.

I'm not really worried about it. It's bolted to the crank pulley in 4 locations around the big nut.

It runs great with the stock sensors standalone. Parallel it ran great, but threw codes. The car must pass obd emissions, and I already had the crank wheel. It was easier than dealing with trigger circuits, as I don't have the supplies and possibly the ability to build them.

The 60 tooth wheel should be more precise than the stock wheel anyway, I'm cool with running it. I wouldn't mind being able to share the cam sensor signal one bit.

I think that's the way to go. Post more details (pics/source/mounting method) on the crank wheel. This is something I've wanted to do and been asked to help with several times!

The crank wheel is really bad ass. I have some more pics of it here
https://www.miataturbo.net/forum/t40676/

The sensor bracket mounts to existing locations on the pass side of the crank. The bracket boss is 1/2" thick and it came with replacement bolts that were each 1/2" longer than the factory bolts that came out of the block. The two allen cap screws allow adjustment of the sensor front to back to get it on the same plane as the wheel, as well as adjustment in and out from the wheel to set the gap.

The wheel itself has a center section that is concentric with the crank pulley, fits around the crank nut, and is secured by bolts around the crank nut. It too came with longer hardware to replace the factory hardware.

I wish I knew the origin of the unit, but as you can see by the rust, I got it used. It came off fmowry's car. It was on the car when he bought it, but maybe he knows more about it, or maybe he can point us towards the previous owner.

I think it was made by Vishnu.

--Ferdi

Dammit Ben, I want that setup! I'm going to PM fmowry and see if he can point me in the right direction.

Unfortunately it's been raining today. Have not been able to make forward progress on the car. :(

However, I did get the timing light on it today. I am extremely thrilled to report that timing is dead-nuts on. No scatter that my eye can detect to 4000 rpm. No hardware delay to 4000 rpm. Good stuff.

Tomorrow is supposed to be sunny, so hopefully I can get it buttoned up and take a ride.

No hardware delay, huh? That's interesting. Generally by 4 it'll start showing up, and by seven it's pretty obvious. I guess there IS no delay loop inherent in a VR pick up, whereas I use a fixed delay as part of the noise filtering.

Ok, I'm feeling like a retard. I pulled up my MSQ to see what my latency setting was (I remember that I did have to dial in some latency with my 36-1 VR setup) and for the life of me, I can't remember where in MT that selection is made.

Bueller?

Mine is under Basic > More Ignition Settings
about 80% to the bottom

Some googling confirms that what I have is a Vishnu TEC trigger kit.

Yep, after PMing the original buyer, I confirmed that as well. So disappointed I can't get one :cry:

I need to find another good way of doing it now. Joe, you want to make me a nice hub like you did for yours? ;) Course that still means I need to make a mount for the sensor.

Yeah, that's MS2. I'm going crazy trying to find it again in MS1...

Well, like I said in the other thread- if you have a hot, single sister in the 25-35 age range who is looking to get hooked up with one of your friends, then we can probably work something out. :D

Beyond that, I sold my lathe to Abe when I left CA.

It just so happens I do, but I'm not sure it'll work out. She bats for the same team...then again, you never said she had to be straight. :laugh:

I wish I had a lathe. It would be so easy to make the part with a decent lathe.

On the dyno today, one of the bolts that holds the trigger wheel cup to the crank pulley snapped. 2 others loosened up. The resultant banging noise sounded like I lost a rod. 1 bolt remained tight. We snugged the other 2 down. The broken one seems like it's going to back itself out all the way, which is awesome.

When I installed the wheel, I did loctite the bolts with blue. This weekend I'll remove them and hit them with red. Hopefully the broken bolt will fully extract itself and I'll find replacement hardware.

The rattling noise was a huge mystery. Sounded like bad engine damage, but we couldn't find what it was. We even pulled the cam cover and timing covers off, assuming it was the bearings in the tensioner or idler.

Wasted 3 hours of dyno time troubleshooting.

BOOO!!!

Ben, I'll bet dollars to donuts (whatever the f*ck that means) that there's nothing between your trigger wheel/adapter and the four little allen bolts that hold it to the crank pulley.

What happens is that, over time, engine vibration causes the bolt heads to chew away at the soft aluminum adapter beneath them. The bolts then lose preload, loosen, and then eventually snap off. As you've found.

Loctite won't solve it.

The solution involves opening up the ID of the aluminum adapter on a lathe so that you can drop in the factory washer-ring thingie between the four bolts and your aluminum adapter. BTDT.

There are little washers behind the allen bolts, but they have a very small OD, not too much larger than the allen heads.

Dammit.

Yeah there's not space in there for individual washers of adequate size.

Car is going to Orlando next week. As an interim fix, I'll remove each bolt individually and replace them with loctite red. I can also grind down an edge of a washer such that it has a short side, and put the short side against the inner edge of the aluminum cup. Hopefully that will spread the load out on a large enough surface to hold.

A solid, flat, thin steel disc may also work. Cut it to the ID of the pulley, drill it out for the 5 bolts, and put a thin coat of JB behind it.

The bolts broke at the head, or along their shaft?

This sort of thing was a great worry of mine initially, but it's all held up rather well. I do have a nice solid chunk of aluminum between my wheel and the flat, inner surface of the crank, but still, those bolts are awful long and subjected to quite a bit of stress.

Worst thing that's happened is that the wheel is cutting a bit of a groove into the face of the sensor. It's obvious that they've come into contact a few times. Can't quite figure it out, but I'm assuming I've got some flex in the sensor mounting bracket, which is quite long. Every now and again, the engine skips on me- almost always at high RPM and under high load. This could be indicative of the sensor coming into contact with the wheel- I assume that'd fuck up the signal enough to confuse the software...

A solid disc will work. The advantage of the factory washer ring thing is that you can also fit hex head bolts in there with it. More surface area than allen bolts and nicer for servicing.

1 bolt broke at the shaft, just aft of the head. Jay's description of the failure mode sounds accurate. Under 50 street miles on the set up, and a little time on the dyno.

Man it would be nice to have a lathe.

The disc I think is the route to go.

The 3 remaining socket cap bolts were tight after 20 miles. The original mazda hex bolts have a integrated 6mm washer. I removed each socket cap bolt individually and replaced it with a new 7mm washer and loctite red. I am going to give the loctite 24 hours to set this time, I failed to do that when I previously installed the wheel. Hopefully that will make a difference.

The 4th broken bolt seemed like it would extract itself, but so far has not. What a PITA area to get tools to, if it comes to that.

I keep driving the car, and it's still tight.