Ignition output/dwell issues MS3Pro 2JZ VVTI

[IS300Turbo]

Hello, I'm hoping someone may be able to shed some light on an issue I'm having. Engine is a 2JZGE VVTI. I'm running COP using LS coils from a GMC Envoy. From my research these coils are supposed to be pretty hardy. The issue I'm having seems to be related to spark output. Let me preface everything by stating this is a new build. I just started breaking the motor in and tuning when I ran into this issue. Motor starts right up, idles smooth. All sensors are reading correctly, no sync issues as far as I can tell. If I slowly rev the engine up and hold a steady throttle, at 3000rpm it starts to breakup. Below 3000, no problem. If I goose it a bit, it will rev past 3000 but still breaks up noticeably. At first I thought it was just an AFR issue, when it starts to breakup, wideband reads lean. I adjusted the VE's and removed the lean spot and it helped a tiny bit but not much. I double checked everything trying to fix the issue, couldn't get it to go away and then finally out of desperation I started playing with the ignition dwell and voila, that made a difference. I originally had my dwell set to the recommended 3.5ms. Once I lowered the dwell to 2.5ms, the problem disappeared. Motor revs up clean, all better I thought. Took the car for a drive, started tuning the VE table, everything seems great. Start getting into boost and winding out to higher RPMS and then started running into some breakup. AFRs are good, 99% sure the spark is blowing out. Makes sense because the dwell is set so low. With these LS coils I should be able to run 4.5ms no problem. My plug gap is set to .028. I'm only running 5psi of boost. Problem is, if I raise the dwell, then the car breaks up at low rpm. With the dwell at 3.5ms it breaks up at 3000rpm, if I raise the dwell higher, it breaks up at an even lower RPM. I thought perhaps the coils weren't getting enough current, but when I check the primary voltage at the coils while holding 3000rpm where the motor is breaking up, voltage is good. I hooked a coil up to a spark plug externally and put the output in test mode so I could observe the spark. I raised and lowered the dwell and observed the spark get stronger and weaker accordingly. I set the dwell to 3.5, set the spark frequency to the same as if it were running at 3000rpm and everything seems fine. I opened the gap on the plug way up and even at 1ms dwell it was still jumping the gap no problem. I thought perhaps there is some kind of bug with the specific 2JZVVTI ignition code so I switch it to toothed wheel and setup all the other parameters, same exact thing. I tried changing from COP to wasted COP, same thing. At this point I'm out of ideas. I've checked all my wiring, all grounds are good. Everything is wired as per the manual. Coils are powered by their own relay. I've checked my physical timing, it's spot on. Ignition timing is spot on. The ignition coils came from a running motor. I really don't think there's anything wrong with them. However, I haven't checked all of them in test mode as of yet, only cylinders 1 & 2. Spark plugs are NGK V power, 7 heat range. I recorded datalogs while holding the motor at 3000rpm. I can't see anything that looks wrong. I also recorded a composite log and there is one strange thing on page 8. I'll attach all these logs and my tune. If someone could take a look I would really appreciate it. If the injector timing looks weird, it's just because I was messing with that seeing if it made any difference. It didn't apart from altering my AFRS. Thanks in advance! I hope someone can help. I've emailed DIY autotune but all they suggested was setting dwell to 3ms and lowering spark plug gap.

[nathaninwa]

Don't lower the spark gap. From other forum reading and some one well versed on the ls coils tightening the gap kinda creates a short for the coil in a way. I'm running .032 and 33psi with no issues. Don't run these coils over 3.8ms dwell as they can break down early causing advanced ignition timing

How are the sensor return grounds wired in on the coil?

What springs are you using between the coil and plug?

Just helped a buddy get his vvti 2j running well and his sensor returns were wired funny, unhooked them and now clean as can be. We also discussed rf noise from not having a spark plug wire

I run a capacitor on my main coil power lead, not that I needed it but I had my own syncing issues in the beginning

Could be a bandaid till you get it figured out, but ms3 has a coil dwell table based on load and rpm you can new with

[IS300Turbo]

Ya, I couldn't imagine lowering the gap more would be necessary. I did read about them auto firing if the dwell was set too high. When I originally wired the coils, I used the oem pigtails and made my own harness. I ran one wire for the +12v and soldered the coils to it, did the same for the ground and the logic ground. The coil signal wires are all crimped. The logic ground for the coils is crimped to the ECU wire along with all the other sensors. The ground for the coil goes to the head to the same place as all the other ECU grounds. What did your friend do that was funny with the sensor returns? I reused the three oem springs that came off the stock lexus coils, and three more that came in pack of random springs. I was wondering if the needed to be something special. I did find that a few of my coils weren't actually sitting down on the plug far enough because the seal on the boots was airtight and would push them back up. I drilled a small hole and now they sit down all the way. The actual coil contact is maybe 1/4" from the tip of the spark plug. If I push down I can feel it bottom out. That's a good idea about the dwell table. I will enable that and see if that helps. I might also just rewire all the coils if nothing else works. Thanks for the help!

[nathaninwa]

He had mixed some of the ms power grounds and tied them into sensor ground. Honestly, you don't really need the sensor ground hooked up, it's unclear what gm uses it for but some peoples reverse engineering of it shows they use it for trigger return to verify the plug fired, hence we get a misfire code

I'd cut them at this point till it's sorted. Mine are wired in, but on an ls install I did I left them out intentionally. Both installs work great

Also try some new springs. I bought new stock Toyota springs and at 30lbs I wore them out in 6 months. Would misfire under boost anything over 14psi. I have some that are 50% larger in diameters now with over 1.5 years

I get a bracket to hold the coils inplace for sure and push them down as far as you can

[IS300Turbo]

Oh that's interesting. I'll definitely give that a try. I'll see if I can find some better springs also. Making a bracket to hold the coils is on my list.

[IS300Turbo]

So I disconnected the coil logic grounds, that made no difference. I also updated the firmware on the Megasquirt, no difference. I enabled the dwell table and set my lower rpm/load dwell to 2.5ms and my higher rpm/load dwell to 3.5ms. That made it worse. Car breaks up before It even thinks of getting into boost. Basically, it appears anything over 2.5ms dwell and the coils are taking a s***. I decided to check the resistance of the ignition coils. I found specs for LS1 coils on another forum. According to that person, the resistance between the positive and negative terminals should be .4 to .6 ohms. Well, when I set my multimeter scale to the 200 ohm range, I get nothing, i.e. infinity. I had to move my range all the way up to the 200k ohm scale, where all my coils read about 52. Which I assume means 52 thousand ohms? Does this mean the coils are all shot? Seems unlikely that all of them would read the same if they were shot, or some were shot. The other specs given to check for are between the ground terminal and the output of the coil. Supposed to be 5000 to 7200 ohms. Mine all read infinity, regardless what scale I put my meter on. I also checked the resistance between the signal terminal and the logic ground terminal, all my coils read about 9.8k. And between the logic ground and ground they all read about 145k. Not sure what all of this means. I want to think these coils are just smoked but I find it hard to believe they would all give consistent values and operate fine at lower dwell values. I think I'm just gonna buy another set off ebay and try them.

[Matt Cramer]

Are these the heat sink type coils? I'd try a set of coils without the heat sinks; they don't have the auto-discharge "feature".

[IS300Turbo]

No, they're not. I did just buy a set of the heat sink ones off ebay though. Waiting for them to arrive. I double checked the resistance readings I got today with my friends much nicer voltmeter and it was the same. There must be something wrong with these coils, the values I'm getting are so far off from what they're supposed to be, I don't know how they're working at all.

[IS300Turbo]

Got my new coils in today, swapped them in, same exact s***. This is really pissing me off now. Anyone else have any idea what could be going on?

[radial]

I have some 5'cents you might try;

Verify timing with timinglight, and test again with static timing of 10*.... does it rev any better? Test with 15* etc.
If it revs better now - reduce timing on your table. It brakes up when you are in the 30-35* area of 40kpa, that amount of timing sounds alot...perhaps too mutch?

Also, try to add some fuel...get it down to AFR 11,5-12ish. from 1500rpm and above, makes tuning a little easier in the beginning. 14.7 is better to dial in when it runs well.

Also, try testmode and physically check your spark performance.
Weak spark usually indicates wiring issues. A poor ground will easily kill coil performance as it won't get enough power to charge itself up.
I'm running AWG12 cables for power to my coils, and all the way to gnd at engineblock. Fused 20A and a small cap for noise supression. All connections was soldered as my cheap ebay coil pigtales was extremly poorly crimped from factory...could pull the wires out of the crimps...lol
My Delphi D514A's could probably kill a horse with sparks above 3,5ms - So the LS coils are for sure very powerful when wired correctly and sufficiently. Mine have the 4ms limit, will advance timing if above 4ms.

[IS300Turbo]

Thank you everyone who tried to help. I finally figured out the source of the problem. So, when I originally got the car running, I set the ignition options/wheel decoder to the specific 2JZVVTI selection that is in the pulldown menu for spark mode. Ignition input capture was set to falling edge. Running fully sequential injection and COP. The motor fired right up, verified the timing was on with a timing light, did tooth logs and everything looked good, no sync loss or anything. The motor seemed to run pretty d*** great except for my strange ignition breakup issues that I could alleviate or exacerbate by adjusting dwell. Well, I put new coils on today with no luck, I then rewired all the coils just to be safe, no luck. Then just for shits I started playing with more ignition options while the car was running. I changed the ignition input capture to rising edge, revved the motor up and amazingly it no longer breaks up. However, when set to rising edge, and spark mode set to 2JZVVTI, the motor will not start. So I changed the spark mode to toothed wheel and rising edge and it starts up, revs up good, ok sweet. Then I noticed that my VVTI is not working anymore. It was working perfectly. So then, looking at the trigger wheel arrangement pulldown, I kind of figured out the problem. I had it to set to dual wheel missing tooth. The problem is, on this motor, the crank wheel is a 36 tooth with 2 teeth missing but the cam is a 3 tooth evenly spaced trigger. It appears there is no option to run a crank wheel with missing tooth and secondary trigger without missing teeth. I think the 2JZVVTI spark mode must be only for the GTE motors, like the Supra, that are factory turbo. I believe they have a different trigger setup on the cam. So without being able to use a cam sensor at this point, I am forced to run wasted spark coil on plug and non sequential fuel injection. Kinda lame, but the good news is that it's finally running like it's supposed to and I can finally rev to redline under boost without issue. I hope someone from Megasquirt can chime in on this issue and verify what I seem to have discovered, and also if there is a way for me to use my 3 tooth cam wheel so that I can enable VVTI again. Thanks!

[Matt Cramer]

The 2JZ-VVTi mode can work with both turbo and NA motors. What are the odds you've got the cam sensor wired backwards? Wiring the sensors backwards can have the same effect as swapping ignition input capture.

[jsmcortina]

I had a look at the logs in the first post.

The composite log looks really clean, that's a hall-effect type cam sensor I guess?
The tooth log shows that the crank sensor is wired backwards. Swap the wiring or swap the input capture setting.

Please use ENTER in your posts to spread the text out and make it more readable.

James

[hobieboy]

The tooth log shows that the crank sensor is wired backwards. Swap the wiring or swap the input capture setting.

James

Hi James,
Sorry to hijack the thread but how does one tell from tooth log that the crank sensor is wired backwards? Is it because of 2 longer bars instead of 1?

thanks!

[jsmcortina]

Sorry to hijack the thread but how does one tell from tooth log that the crank sensor is wired backwards? Is it because of 2 longer bars instead of 1?

Yes, see the manual:
http://www.msextra.com/doc/pdf/html/Meg ... .4-48.html

James

[IS300Turbo]

They are both VR sensors. I read the manual but was unable to figure out how to tell polarity from looking at the tooth logs. Obviously the crank must have been wired

backwards because that was the fix, but when I have the Spark mode set to 2JZVVTI, the engine will only start if set to falling edge, it won't start on rising edge. If I switch

to rising edge while the engine is running, it keeps running. If I set the spark mode to toothed wheel and rising edge, everything is good. That's what I don't understand. And

on toothed wheel mode, my VVTI is acting all screwy, I'm pretty sure I tried switching the capture mode on the cam to no avail but I will try again.

[radial]

I dont think toothed wheel supports vvti. Always custom wheel modes on vvti.

[Matt Cramer]

I dont think toothed wheel supports vvti. Always custom wheel modes on vvti.

Toothed wheel does support VVTi, but only with the following configurations:

1. A one tooth cam wheel.
2. All equally spaced teeth on at least one edge (GM 4X, for example)
3. Certain modes that have dedicated cam decoders (Ford Coyote or Duratec, Chrysler Hemi)
4. Certain equally spaced / missing tooth layouts.

2JZ-VVTi mode is none of the above, however.

[jsmcortina]

If I switch to rising edge while the engine is running, it keeps running.

That's because the setting is only checked once at power-on.

The composite log most certainly did not look like the signal from a VR sensor on the cam.

James

[IS300Turbo]

I don't know, the cam sensor seems like the same type as the crank, it has a magnetic head. It's wired to the CMP+ and CMP- wires. Ya it definitely won't work with the toothed wheel modes as the crank has missing teeth but the cam doesn't. I can't use the 2JZVVTI mode because the motor won't start if the crank is set to rising edge, and ignition breaks up if set to falling edge. I have one of the 10k resistors wired inline on the crank. Could that be causing any issues?