Misfires under boost, ecu hardware?

EMU PRO
1

Hey all,

I have an EMU PRO8 running on my supra with a R35 COP ignition conversion, running a 2jz GTE style trigger setup. 36-2 crank gear and 3 tooth cam pickup through OEM toyota sensors.

I’ve been running this car and setup for about two years as a daily problem free now. Since a few week i’ve been getting misfires under high boost, first at 1,6 bar, then at 1,4 bar and now even at 1,2 bar.

The coils that came with the kit werent genuine hitachi ones so I just swapped those out with genuine ones and new sparkplugs. The ground for the coils is on the engine block and the 12V feed is from a ignition switched relay from the fuse box with a nice fat cable.

I now keep getting misfires, even with the new coils and plugs. I have run this car unchanged for 2 years now and even though sometimes I had a tiny stumble at 1,6 bar, the car is now undriveable even at 1,2bar under boost conditions. With the boost controller switched off while driving on 0,9bar wastegate spring pressure its still OK.

I have checked literally everything, even changed out engine bay side wiring harnesses that contains all the +12V,

High RPMscope_2026_0903_1510.emuproscp (2.7 KB)

Misfires and LC2026_0818_1324.emuprolog (6.1 MB)

scope_2026_0903_1509.emuproscp (3.0 KB)

ground and signal wires and replaced coils and plugs, made new grounds and new +12V feeds. However the misfires remain unchanged.

An intresting detail is that from day 1 coil 3 and 6 have always had a high coil dwell time error. They are often at 0,043ms while all the other coils are at 0,001-0,006. Because of this im starting to suspect a hardware failure in the ECU, but at the same that would be strange since the ECU only provides the firing signal to the coils…

Attached are a log with the misfires and two scopes, one at idle and one at higher rpm. Both stationairy free revving so there where no misfires during scope.

I am completly at a loss and would highly appreciate your help and guidance.

2

Can you provide time when the misfires are in the log? There was 3 launch control events, were those on purpose?

3

Ok seems I fixed it…After making a whole new +12V feed by temporarliy tapping into my fused fuel pump feed coming directly from the battery, I got a major improved. I was able to trace everything back to what I suspect was a wonky Deutsch connector. I rewired the +12V from the coils and took out the Deautsch connector, did a load of testing just now and the misfires are gone!

The dwell time errors do remain thoug, but they disappear entirely under heavy, that still stumps me…

4

Dwell time error is an effect of an uneven trigger pattern or a crankshaft speed change.

Dwell error is the difference between what the ECU expected and what actually happened, based on the crankshaft sensor.

If you enable the view of edge distance on the scope, you can see that the distances are slightly off around the gap.

obraz
obraz584×605 15.3 KB

When you look at the dwell time of coils after the gap in the scope, you can see they have different values compared to other coils. And those are cylinders 3 and 4.

But that doesn’t match the log, which reports a higher dwell error for cylinders 3 and 6.

I suspect we may have a numbering inconsistency between the scope and the log, likely due to the conversion between cylinder and firing order.

We will look into that, but overall, the dwell error you have isn’t concerning. If the crankshaft sensor is a VR sensor, swapping the sensor’s connection polarity could help achieve more consistent edge distances.

5

Thanks! This of the one of the reasons I love this forums so much! Being able to converse with developers and true experts is very unique and awesome! I fear there is not a lot I can do about the trigger edge gaps, I swapped around the wires during the first few start ups but how it is right now its optimal.