Ignition Timing drift/limit issue on a 91 Toyota MR2

[NL22]

Vehicle: 91 Toyota MR2 
Mods: Vehicle was originally a non-turbo model but was swapped to a 3rd gen 3SGTE, stock internals, factory ignition system, 1000cc injectors, AEM 340 fuel pump, upgraded pipes and intercooler, OEM manifold mated to a t3/t4 external wastegate adapter with a garret gt35 turbo, downpipe, Berk exhaust system and on e85
Engine Management: PnP Link G4X ECU

It's a long read but bear with me, just trying to get in as much information as possible. The issue I am dealing with is that the vehicle is having a major power loss after 4700rpm and all the way to 7000rpm. The vehicle was tested on a Dynapack Dyno. When I looked at the logs, it'll build boost up to 14-15psi on wastegate spring at around 4700rpm and hold all the way to 7000rpm, but the horsepower and torque will dip drastically after 4700rpm on the dyno graph. Did multiple runs with various changes and still the power and torque graph would consistently fall off after 4700rpm. Vehicle doesn't smoke, doesn't misfire, doesn't overheat, idles fine, sounds healthy, and drives fine at partial throttle. Ignition timing was kept very conservative around 12 degrees at 4700rpm and peaks at 15-16 while addressing the issue. IAT was around 110deg at redline. I checked if maybe ignition timing was being pulled by some kind of trim, but ignition stayed flat at 15-16 till 7000rpm as well. AFR was around .760-.780 lambda, target was .780 to be on the safe side. I don't think it's a software issue as I have used the same settings in the PnP Link unit on a few other Mr2's with good results so I was leaning towards some kind of a mechanical issue. Did a compression and a leak down test just to confirm that engine is healthy, and the cylinders were good. Checked the exhaust system to make sure there wasn't any restriction and that was fine. Checked belt timing, that was good as well. Checked for boost leaks and did find a few different leaks at the couplers but nothing too major that I think would cause such a power loss. Fixed the leaks and even gapped the plugs down to .024" just for the sake of it. Mechanically everything that can be checked, checked out as it should. Throw the vehicle back on the dyno and still the same results, same issue after 4700rpm. The last thing I could think of is if there is some timing drift that is causing the major power loss at the high rpm. So had someone hold a timing light and watched the timing mark at the crank pulley to the timing cover while I made a pull and sure enough, the marking on the crank pulley would get to about 10-12degrees peak and fall back down to 5degrees or less while the ecu is commanding 15-16degrees of ignition timing in the logs. Did a few more runs to verify the issue and it was consistent with each run. I've swapped to another good known distributor, adjusted the ignition timing delay, trigger voltage thresholds, and still having the same issue. I've looked through all the settings in the Link and I don't think that I may have missed anything that maybe limiting/trimming the ignition timing at high rpm, positive pressure. Just trying to gauge some ideas of what maybe causing this ignition timing drift/limiting issue. Below is the calibration and 3 WOT logs. Let me know if there is something that needs to be changed, or I may have missed in the Link settings. Thanks.
 

91MR2_Ignition-Issue.pclx wot1_decreasetriggervoltage.llgx wot2_increasetriggervoltage.llgx wot3_basetriggervoltage.llgx

[Adamw]

I dont see anything ecu related in the logs, dwell and engine speed runtimes are about as stable as they get with a distributor trigger so the ecu appears to be happy with that.

You generally dont need a dyno to confirm timing drift - it is usually purely RPM based, load has no effect on triggering so the first thing I would do is lock timing and confirm there is actually timing drift by holding it steady at a range of different RPM's. 

I would expect you would get edges crossing at some point and a big trigger error if one of the sensors was wired incorrect polarity, but a couple of trigger scopes at idle would be a good idea too.    

Your fuel table and inj PW peak at 4700 and drop about 10% by to redline so some of it is airflow related, but 10% doesnt seem extraordinary considering stock manifold etc.  

[servicesoon]

Long ago on earth, I locked timing at 10 deg and used a timing light to compare ecu ign timing to mechanical timing.

At idle speeds they matched perfectly. When I revved the motor to 3,000 rpm, physical timing dipped 3 or 4 degrees. After sitting at 3,000 rpm for 1 or 2 seconds, physical timing settled on 10 deg.
 

My conclusion, if you are using the distributor for crank and home signals, it is quite normal to ‘lose’ ign timing. I’ve heard from others testimony that it is not uncommon to lose 4 to 7 deg. 

[Vaughan]

15 minutes ago, servicesoon said:

Long ago on earth, I locked timing at 10 deg and used a timing light to compare ecu ign timing to mechanical timing.

At idle speeds they matched perfectly. When I revved the motor to 3,000 rpm, physical timing dipped 3 or 4 degrees. After sitting at 3,000 rpm for 1 or 2 seconds, physical timing settled on 10 deg.
 

My conclusion, if you are using the distributor for crank and home signals, it is quite normal to ‘lose’ ign timing. I’ve heard from others testimony that it is not uncommon to lose 4 to 7 deg.

Did you do this on a factory ECU or on an Aftermarket ECU? If the ignition delay has been set clearly the physical timing should always match the specified timing in the Link. Some trigger setups can have scatter and slight movements on sudden transients but not a delay of 1-2 seconds

[servicesoon]

11 hours ago, Vaughan said:

Did you do this on a factory ECU or on an Aftermarket ECU? If the ignition delay has been set clearly the physical timing should always match the specified timing in the Link. Some trigger setups can have scatter and slight movements on sudden transients but not a delay of 1-2 seconds

The test was on an aftermarket ecu to determine appropriate ignition delay.

Is your advice that NL22 needs to adjust ignition delay?

[ATSAaron]

Hook up a timing light. Lock timing at something you can see. 15 is good on a 3SGTE. Check at idle, rev to 6000 and see how much it retards. Adjust ignition delay appropriately. 

Aaron