Timing errors with AEM EMS-4 (crank sensor)
 

Car: 1994 w/ VVT Swap
ECU: AEM EMS-4 Series 2

I'm getting timing errors all over the place and the logs point to the crank sensor picking up teeth that aren't there, this causes a nice jolt as the car misses and has been a pain in my ass. AEM techs have advised me to put a pull-down resistor on the signal wire, and then an RC circuit in an attempt to filter it; however, still getting timing errors and the RC values are getting ridiculous (lol .5 milliseconds).

I am curious if I should just do Abe's NB Crank circuit (Schmitt trigger) and then just fuck with the AEM to make it either rising or falling edge. I have a track day that I'm trying to make in 2 weeks, but I've been fucking around with this timing error shit for almost a month.

Fun aside: the IAC also seems to be wrecking havoc (with the IAC plugged in, I get way more timing errors), so I've just unscrewed the idle screw and I'll just let it idle high.

I'll try to post with any requested information.

You shouldn't have to jump through hoops to fix your problem.
Sounds like a grounding problem. How is the sensor ground pin connected to the sensor ground in the car's harness?

Spliced the 94 CAS wiring into the individual crank and cam sensors. The power and ground are shared between the two sensors.

The signal wires from the crank and cam need to be shielded, ground the the shield to the engine block near the sensor.

I can move the drain to the valve cover real quick; it's currently grounded to the chassis. BRB.

EMS-4 is really sensitive about the crank and cam signals.

Just moved it over to the valve cover, still getting a shitload of timing errors with the IAC plugged in. I do get timing errors with the IAC unplugged; however, it seems like it isn't as violent; I just cranked the idle speed up to 1200rpm on the TB in the mean time.

Right now I'm running 16awg shielded to both the cam and crank from the factory 94 wiring and the shield drain is connected to the valve cover (like I've said about 5 times in this post) currently.

I have some 18awg, 3 twisted pair shielded cable I can use. Tie the +12V and GND on one pair, and then put the signal and a spliced GND to another twisted pair; tie the shield drain back to the valve cover.

Can you drive with the IAC unplugged to see if the timing errors are really gone, your IAC may bad or wired incorrectly.

The timing errors still increment with the IAC unplugged, but it's a lot less frequent.

I want to see what the twisted pair will do, and go from there. (fuckin' 5th time creating this harness shittttt)

Is the cable running near the coil pack or spark plug wires?

Nope. This is a bit outdated, but it's completely over the intake. The route is pretty similar (blue)

https://www.miataturbo.net/attachmen...ine=1364246990

Well, moving to twisted pairs didn't work. Still getting timing errors.

Isnt one of the sensors that came on VVT engines known to have problems (cam/crank) or was that the 99-00? You have me worring because I just wired mine in, and used the stock NA cas wiring. The NA cas wiring is not shielded and the wire I used to extend the crank sensor wiring isnt either. I didn't think shielding was really required with hall effect sensors.

I'm doing that Schmitt trigger circuit. If that doesn't fix it, I'm out of ideas.

Now to pay 5 dollar shipping on a 48 cent part.

I'm battling the exact same issue as you right now! I've been battling the timing issues for a few weeks now. We should compare notes...

I have a 95 car with 99 head and I've spliced in the wires exactly like you did. I used to have a shit ton of timing errors. I flashed my firmware to v24 and it actually got a lot better but I still have timing errors every few seconds. According to my logs the ems seems to be picking up extra crank teeth where there arent any (sharp drops in crank tooth period time). I thhought that might be due to noise so I've been all up and down the chassis searching for faulty grounds or ground loops and I can't find any obvious error. I'm in the process of running shielded wire all the way back to the ecu for cam and crank, but somehow I doubt that will fix it. So far the ems4 seems extremely finicky with these inputs.

What is this Schmidt circuit mod? I may have to try it as well...

I just fixed this (i think). Do the Abe NB circuit.

I don't have the trigger in yet, but the passive filtering (RC circuit tied to ground with the +12V pull-up) got rid of all of my shit. Now to bitch out Sav..

Can you tell me exactly what you did? Quick diagram would be cool if possible. I tried a. Rc before but I might have used all the wrong values. Or just wired it wrong.

Which crank sensor and trigger wheel are you gays using?
99+, or 94+?

I'm using the oem 36-1 wheel.

OEM by Ford?

Which crank sensor?

Oh sorry it was the 2000 1.6l protege wheel 36-1 or something that was stickied a while ago and I have a nb crank sensor and nb can sensor.

Crank sensor: OEM 2002 Miata
Cam sensor: OEM 2002 Miata
Trigger wheel: TSE 12+1

The circuit that I used is the NB Abe circuit but no Schmitt trigger.

Okay I searched around and found this thread.

Is this the circuit you made? I made some modifications to his picture based on what I understood from your previous post.

https://www.miataturbo.net/attachmen...1&d=1364598557

That's it: I did it for the cam and the crank.

I didn't have too much time to putz around with the IC since I was trying to make a dyno appointment; however, I found this to work. The guys at AEM were saying to use a pull down resistor, and if that didn't work, try a low-pass. They suggested a max of .01uF, which kind of set me on the wrong path seeing how I was getting RC values of ~.5ms.

Wow they told you to do exactly the same things as me. They must have copy pasted the replies to us haha. They were super nice though at least haha so I can't fault them too much.

I can't wait to try the circuit as soon as I get back from the track. Unfortunately running my slow rx8 backup car haha.

Ps: where are you located?

How bad were your timing errors? I only tested mine during cranking with the injectors turned off before I did the first start of the engine and I was only getting a couple. I didnt have time to go back and look with the car running to see if I was getting any. I think the most I ever saw with the cam/crank sensor diagnostic box pulled up was 3 while cranking.

Is this diagram a little weord or is it just me. I bought a 2.2k ohm resistor for the one labeled 220k right?

I don't really get any before startup, but afterwards it happened every like 5-10 sec. Kinda inconsistent and random, not related to rpm at all. sometimes it would be like a minute wthout any, sometimes one every 2-3 sec.

Basically i think the circuit pulls the signal up to 12V before it filters it. That's all.

I think 220k means 220k...you might have to look in the 1/2 watt resistors.

:dealwithit:

Seconding what reddragon1234567890 was saying. If it was attached to engine speed, load, or anything it would have been fucking awesome; however, it was intermittent as fuck and messing with the IAC and shit made it sometimes better but also worse.

As for the diagram, it's 220k ohms in parallel with the capacitor (this is an RC circuit/low-pass filter) and is pulled to ground.

I only get 1, maybe 2, timing errors and it's during cranking. A 15 mile drive would net me north of 100 timing errors. It wasn't pleasant.

Lol.

Hey Gamo,
I just tried it and things are WAYY better. I got 1 while cranking, and 2 shortly after starting, but i've had the car on for like 20 min now and no more errors! Is that what you're seeing as well? Also do you have your crank set on Hall rising or Hall falling? People say it doesn't matter but I was wondering what you're running. I have Crank Falling and CAM rising.

Set both of them to the same edge detection, there will be a delay if they are not the same. Also, check the base timing because you might be off a few degrees.

I usually get one or two at cranking, but that's it. I'm glad it's working for you.

I just did this and the car won't start now. Feedback?

I only did the crank for now.. Not sure if it matters.

Just FYI, 2-3 timing errors during cranking are totally normal.

After installing this circuit on my cam and crank without changing any settings I get no startup. Just counts errors and won't fire up. Without the circuit it fires up and will randomly error while at idle and cruising.

Does tooth number increment?
You probably screwed up and the AEM is getting no signal.

Did you get a 2.2k or a 220k ohm resistor?
Is the crank count incrementing?
Is the cam count incrementing?

I went and got a 220kohm resistor. It does count. Just doesn't fire. Just counts errors.

This might be dumb but check if you have the input and output reversed?


Gamo, quick thing, did you use the sensor ground as a ground or a separate chassis ground somewhere?

I tried switching it. One way seems to count teeth normally the other way counts very slowly.

http://img22.imageshack.us/img22/9524/photo1gb.jpg

http://img22.imageshack.us/img22/9524/photo1gb.jpg

Red is 12v, Black is ground, white is the signal wire.

The resistor connected to the red wire that has heat shrink covering the band is the 3.3k. The Band that is covered is the GOLD one.

So from left to right:
GOLD-RED-ORANGE-ORANGE
GOLD-BROWN-BLUE-YELLOW
GOLD-YELLOW-RED-ORANGE
4.7uf CAP

I may have just noticed a problem. The circuit calls for 4.7nf not uf.

I was just about to say that the cap might be wrong =). Get one of those dinky looking ceramic capacitors. They work good. 0.0047uF to save you the conversion.

Hmmm I'm still getting a few errors (5-6 total at most) before the car is completely warmed up, but after that it doesn't seem to get any more. I can rev it all i want after that and I don't get any errors...any ideas?

P.S. I found out today that the alternator doesn't charge if gauge cluster isn't plugged in. May save someone from panicking and rummaging through wires in the engine bay. lol.

Yeah, the 4.7uF will be an issue since it'll take forever to charge up. Radioshack had them (shocking, I know) for me locally. It'll be a small brown thing, and make sure that you get 4.7nF (.0047uF) because I was off a decimal point when I got the caps the first time.

Well - I'm an idiot. I had never used the "NF" measurement before... Usually it's 1 Farad. I'll swap it over tonight and let you know the results. I timed it and I'm getting 11 errors in about 60 seconds of idle without a filter (cold car).

SI Prefixes:
milli = m = 10^-3
micro = u = 10^-6
nano = n = 10^-9
pico = p = 10^-12

You will almost never see whole Farad capacitors because it is an insane amount of energy. The formula was designed before there were any practical applications and you'll see most capacitors in the micro range. They are never labeled nano, so you have to either convert from micro or from pico.

To put it more simply, AEM told me to not go above .01uF; you were at 4.7uF (500x). I used a 4.7nF which worked, which is 1000x smaller than the 4.7uF.

Orders of magnitude can break things fast.

Best of luck.

e: to quote wikipedia...

Thanks for the breakdown- whole farad comes into play with voltage regulation and car audio components :)

What type of exhaust has capacitors on it? :party:

Swapped it out with the 0.0047uf cap and shes perfect. Previous test showed (at idle) 11 timing errors in 60 seconds. New test shows 0 timing errors... ever. Success. Thank you GAMO

I wonder who designed the circuit... ;)

;)

Looks like I need to build this circuit. I was getting about 70 timing errors per minute and pretty much constant sync errors counting up to 3 and resetting. Every time it would hit 3 sync errors it would miss something fierce, like hitting the brakes practically.

Let me know if you need any help.

That was easy. Now I get about 10 timing errors per hour and sync errors just barely flickers between 0 and 1. No audible miss fires.
Black = Ground
White = +12
Brown = Cam sensor
Orange = cam ecu
Purple = crank sensor
green = crank ecu

https://www.miataturbo.net/attachmen...ine=1365880579

https://www.miataturbo.net/attachmen...ine=1365880579

https://www.miataturbo.net/attachmen...ine=1365880579

https://www.miataturbo.net/attachmen...ine=1365880579


What a nest.
https://www.miataturbo.net/attachmen...ine=1365880579

You should be getting zero. Did you run shielded twisted pair for the crank/cam wires?

No I used the stock wiring for the cam and just extended the wires to the crank with some milspec wire. The NA CAS isnt shielded. If it becomes a problem or if I ever rewire the car I'll run shielded all the way.

To update this thread. I added the intake manifold side ground because I appeared to be missing it. Now I get 1 timing error on startup and sync errors still flicker from 0 to 1 rapidly. HOWEVER I have what might be a bigger problem. With the car idling at 1krpm and the ignition timing locked at 10deg the timing marks are perfect, but if you rev it to ~2500 the advance slowly climbs to 12deg and we couldn't really see with the timing light as the revs went higher. This is an issue, is it caused by something changing in this circuit be it the pull-up voltage or the ground offset voltage or something. Could some electrical issue cause this circuit to do this? Is this a downfall of the circuit? or what? Car currently does not knock before redline or at peak torque.