Like the title states, why would the timing move? I could be vague as f*** and stop here, but ill explain what my dilemma is.

The car is a 90 miata with a newer 1.8, running 3.4.2 MSII Extra on a DIYPNP. I was/still am having some sputter up top in boost, around ~5500, and the tach would start doing some little dance bouncing back and forth. Thinking that it might be related to timing latency, I took my timing light out to check the timing across the entire rev range.

Once I got to 3500, the car was going full retard. The timing light was showing ignition timing at TDC. I looked at the loggers in TS to see if anything came up, but either
a) nothing was showing up
or
b) I don't know how to use it

So to save us all from stating some common problems, i listed my troubleshooting logic flowchart with answers.

Could this be a sign of a bad CAS?
Don't think so, I've been using a optical CAS and i switched it for a hall effect from a 94 with no change.
Could it be signal interference from the toyota COP thats installed?
I moved the harness while reproducing symptoms and it did not change anything.
Could my dampener be slipping?
Don't think so, I verified TDC manually with a brass rod and the spark plug out.
Timing belt deflection?
Maybe, not sure what the symptoms of this are.
It could be firmware related.
Yes, it could but i have not seen anything here or miata.net about 3.4.2 causing issues.

What could i be missing? I have a short catalog that has me opening the throttle until it starts retarding the car. I put my msq up as well.
Thanks

 

This log is for this tune? Why is the timing stuck at ten degrees for the whole log?

This log isn't for this tune, or you forgot to turn off fixed timing before making that log. Try again?

 

Correct me if I'm wrong, but it looks like he's saying that when he sets timing to a fixed 10° it moves to 0° by 3500RPM.

Dano - I'm assuming you've verified your timing at idle to start. Does timing continue to retard after 3500RPM? Have you adjusted your timing latency? Your OP wasn't specific on that point.

 

Maybe?

Did you wire the megasquirt to power the tach, or are you using the toyota cops? If the latter, I think you lost coils around 5500 rpm.

 

Thanks for the replies.

The log is for the tune, i intentionally fixed the timing at 10* but the timing light would scatter to TDC.

Timing light is verified 10* and consistent at idle-3000 rpm. The engine would clear up after that and retard at a later rpm, around 6k rpm.

? The tach is wired by the output on the toyota cops, as this is how the early NA setup was IIRC. The MS fires signal to the coils’ igniter and sends the tach signal to the tachometer.

 

The answer is trigger latency.

 

Trigger latency to my understanding is for a gradual retardation. This stays right on 10* and then goes to 0*

 

And I've tried trigger latencies from 0 max(550?) and everything in between.

 

I would try a new CAS.

It is odd how there is a specific place that it retards.

 

Another one? I've tried two: my 90 with is an optical and a 94 that's hall effect so I'm not sure that two could be bad

 

I noticed my timing error is all over the place. So MS is seeing a different timing that what it expects and goes nutty?

 

timing error means nothing.

 

does the log actually show 0° timing, or are you seeing this on your timing light?

 

The log is at 10* fixed but the timing light will show tdc

 

So i tried taking a video. Best I could do hold the phone, timing light, and operating the throttle.


I don't know how to 'embed' the video.

 

then what @Joe Perez said.

Increase Timing Latency until the timing no longer decreases as you increase RPMs.

 

Based on his updated description of the problem (that it jumps from 10° to 0° without passing through the range in between), I'm not so certain.

Thinking more along the lines of soft rev limit. I don't have TS installed on this machine, so I can't check his MSQ.

 

I also checked my spark outputs just incase i might have wired the cops in wrong. Coils 1&4 fire at the same time, as do 2&3. I took a composite log, does this look right?







It doesn't look like this one: