turbofan

iTrader: (1)

Just tried it. No dice. Same issue. Next stop I'll try readjusting the crank sensor and try again, just to be sure. It did seem to rev maybe a little higher this time...

patsmx5

iTrader: (16)

What motor, VVT or 99 head? What ecu? If MS3X, how are the pots adjusted, and do you have a resistor on any of the sensors? Are every inch of the wires between the ECU and sensor shielded? Are the wires extremely close to the ignition cables? Does the problem occur before the engine is heatsoaked? What is the distance between the crank and crank wheel? Is the crank wheel wobbling at all?

aidandj

iTrader: (5)

VVT head.
MS3x
Pots unknown (thought I remembered adjusting them. Need to check)
No resistors
No shielding
Happens cold
Distance unknown. I told Ed to use a business card.
Wobble unknown.

We can get a scope on both sensors and rev to the moon in my garage. That will happen after simple fixes are checked.

If I wasn't also trying to run my turbo Miata at the track 800 miles from home on a cracked exhaust manifold I would have checked most of that by now. I was a little preoccupied

patsmx5

iTrader: (16)

I would shield the wires, just good practice given what they are. Definitely check the pots, that's likely the problem they are adjusted wrong. I use a .040" business card on my motor to set the crank sensor gap. And check the things you listed as unknown of course. Having a Scope is nice.

turbofan

iTrader: (1)

It's not wobbling.

The composite log clearly shows that it's the cam sensor that loses sync, but it continues to register the crank sensor. I'm a little confused Ian, since the cam sensor has less marks than the crank sensor so it's easy to tell them apart in the log.

Perhaps I'm not understanding something.

And Gordon -- thank you! I appreciate it. It was a lot of fun being out on track with you guys.

I am very, very pleased. This was the first time driving this car in anger, EVER, and it was a gutted rolling shell just 60 days before the event. The last time I tracked a Miata was last September, over a year ago. Early Saturday I was struggling to put down 1:56-1:57 lap times, but got into a groove and dropped a lot of time quickly after that.

Can't wait to get back there and try again.

turbokitten

The logs are "events the ECU saw", not "a history of voltages seen on the pin". The cam sensor won't show in the logs if the interrupt masking drops the edges, and the way the code is written (kinda hacky at a low level) can make it unintuitive to diagnose when it goes wrong. We spent a long time staring at it while diagnosing Ian's setup.

I have a hunch it's aggressive filter settings... If you post a screenshot of Ignition Settings => Noise Filtering I'll have a look.

codrus

Basically what I'm saying is that after staring at the source code for a while, I don't trust the composite log code to always tell you the truth. The code is a mess.

Back when my car was throwing sync errors every 15-20 minutes last year, the composite log never showed anything strange. I've come to the conclusion that the log is a useful tool but should not be assumed to always be definitive. As Reagan said, "Trust, but verify", and in this case repeating the 6K RPM test after fixing the crank pulley was verification.

--Ian

turbofan

iTrader: (1)

I. am. So. Pissed right now.

my camera recorded ONE session. Even though I set it to record the last 3 sessions each day, it got only one. it looks like a sunday session; praying it was the last one. Not sure yet. Ugh. sooooo disappointing.

turbofan

iTrader: (1)

Doesn't seem like there's much here, but here's the screenshot.





Also, for the video... the session it got was the last session with the best lap! Uploading to YouTube now...

turbokitten

Theory's busted. An excessive filter might cause it to drop out as you rev, causing the next pulse to fall into the window, but you have it all turned off. I suggest hitting it with a scope next to see if it's a sensor output or ECU input problem.

aidandj

iTrader: (5)

Pots adjusted. Car hits 7k with ease. My fault. Thought I adjusted them all.

EO2K

iTrader: (37)

Yes, make Ed even faster, this is what we need

Glad it was something simple enough

aidandj

iTrader: (5)

It was the ms3x pots. I adjusted the main board pot but forgot about the ms3x zero crossing pot.

18psi

iTrader: (76)

why didn't you tell him to do this from the get-go?
when scott messed up on mah pots we discovered the issue after like 10 min of troubleshooting.

tisk tisk aidan

aidandj

iTrader: (5)

I actually did. One of the first things I said when we saw the cam signal was disappearing. But we were at MRLS and it was dark at the campsite, and I didn't feel like pulling the ECU and doing stuff.

I also thought that you had to measure voltage when you adjusted them, instead of just turning them a set number of times.

aidandj

iTrader: (5)

See. I wonder who came up with the pots suggestion

18psi

iTrader: (76)

haha I also don't understand the pots "ritual"

1) turn 2 counter clockwise
2) turn 3 clock wise
3) sacrifice a goat
4) ????
5) big power

aidandj

iTrader: (5)

The potientiometers are bias resistors that are used to set the zero crossing threshold of the VR conditioner chip. Takes a 0-5 volt signal and converts it into on and off.

EO2K

iTrader: (37)

I feel like my MS3-Basic does not have these magical pots. I'll have to check next time I have it open. I don't remember having to adjust anything, or instructions to do so.

I brought my Fluke and a 12v wall wart out to MRLS so we could play with Eds MS, but I think we got distracted.

aidandj

iTrader: (5)

90% sure they don't. Turns out we didn't need those things. I was mistaken.

Jealous of your fluke.