Adamw

Ok, from the looks of those last logs there doesnt seem to be a trigger issue anymore, I suspect that may only have been a contributor to the starting problem and not all of the difference. Have you tried adjusting fuel at all to see if adding or removing fuel makes it behave any different? Backfiring normally suggests timing is off but can also happen if mixture is far enough out to cause misfires. I have ran your G4+ map and G4X map on the engine simulator and get the same spark timing comparing both so I expect that is ok. As a quick/crude fuel adjustment I would just go to >fuel main and adjust the engine capacity up or down. Try adding 500cc and subtracting 500cc and see if it runs any better either way. Otherwise the G4+ log will be helpful to compare too.

Ok, the ecu logs for some reason we have two of the same log and 1 with nothing. But in all the triggerscopes with the new firmware the trigger looks like it is working fine now (it may just be luck that only good revolutions were captured but that is all I have to go on so far). Sorry to keep asking you to do more and more logs but it is the only way we will get it solved without being there in person... Can you do a PC log of another start attempt. Also download the matching ecu log afterwards (maybe delete all logs that are presently stored in there so you know any new log is the correct one). Perhaps you can also do us a pc log of a start up with the G4+ so I can compare if anything else is drastically different when cranking such as inj PW etc.

Ok, I see the problem with the trigger. Your cam has 3 evenly spaced teeth - 0deg, 240deg, 480deg. The EJ25 trigger mode that you are using is designed for a camshaft with teeth at 0deg, 180deg, 360deg. I know K3-VET normally has 0/180/360 same as subaru but I have never seen a KF-DET before. I will talk to the engineering team tomorrow to see if we have any other trigger mode that will work for your engine. Does this engine have VVT on the camshaft?

Can you down load the ecu log files too.

Log and tune please.

Your E-throttle relay is turned on for the whole log and does not turn off in this log. Can you get a log which shows the relay turning off? However, one problem I notice in you log it looks like there is something wrong with the trigger/RPM - it should not be a bouncy saw tooth like this. Potentially if you have a trigger error that causes RPM to quick drop to zero, then E-throttle relay will be turn off. Can you please do a trigger scope when the engine is idling and attach here, I will take a look to see what is wrong with the trigger. Can you attach your map too please (*.pclrx file)

Thanks, we still dont have a good explaination of what is going on. In one of your triggerscopes the firmware engineer can see the missing tooth gap has been mis-detected due to the crank rapidly speeding up right where the gap is which could explain the starting difficulty. The G4X does an extra validation check after the gap which the G4+ doesnt do which may be some of the difference. But that reason wouldnt likely apply once the engine was idling and the crank speed was more stable so we are not convinced we have found the real problem. I will PM you a special firmware to test that has a loser tolerance on the gap width that may help. Install this like any normal firmware update, just use the browse button in the firmware update screen to choose this file. It would be really helpful if we could get a triggerscope capture when the engine is "idling" or closer to idling - I dont know how likely or easy that is for you, but please give it a go. It may start easier with this new firmware. Can you also get a couple of triggerscopes when cranking (with fuel/ign on), with two variations of "Ignition Idle Target", It is currently set to 10deg, I want to try 0 and 20 to see if that moves the crank acceleration away from the gap a little. You will find this setting in >Idle speed control>Idle Ignition Control.

So you need to fix that first. It could be the leverage ratio of the pedal not well matched, excessive cable freeplay - or something like too much carpet restricting the pedal movement. But what ever the case, the crank on the throttle body should be hard up against the stop when the pedal is on the floor.

Also I noticed in the last trigger scope there appears to be some wiring problem. We can see a small reflection of trigger 1 signal on top of the trigger 2 signal. This suggests to me there is maybe a ground wire broken or missing. Did you change any wiring when you changed the cylinder heads?

No, I mean just check that the bell crank that the cable connects to cant be turned any further by hand - it should be hard up against the stops already when pedal is on the floor

Can you get someone to sit in the car with foot on the floor and then check in the engine bay that the throttle cable has pulled the APS bell crank all the way open. Something a bit funny with the voltage range on that aps sensor - almost looks like it is not moving as far as it should.

I've moved all your posts into this one thread instead of having multiple posts about the same subject. Can you Set E-throttle mode to set-up, start a log, and press the pedal to the floor a couple of times, then save the log and attach. The last log only shows a max of 78% APS so maybe your foot wasnt all the way down. Your "Sloppy throttle" is just your soft E-throttle target set up, you are only asking for 9.8% throttle when the peddal is down 30% and only 50% throttle when the pedal is 80%, so the first 3/4 of your pedal travel is going to feel very soft.

Ok, that cam sensor will need to be modified so that it only gives a single event every 720 crank degrees. The easiest would be to just cut off the pink bit so you have one long window. 4 teeth on crank is fine.

You will need a CANPCB cable and a CANF plug, this will give you something to connect your wires to at the ecu end. The CAN wires and the plug at the Haltech end you will have to look after yourself. Connection will look something like pic below.

Have you done the APS and TPS calibration? It doesnt look like it, in the log your fault is due to the APS (pedal position) sensors not matching each other.

Yes, there is a fee for the upgrade code, around NZD$200 at time of writing but it does give you a big step up in functionality. We dont have a very elegant way to "buy" this upgrade as it is not in the dealers "web store" and probably not many dealers have ever dealt with one. So find a dealer you like, they need to call or email their regional Link sales office and manually order a "400-0011". The sales team will need the serial number from the ecu and will email back an unlock code when sale is complete. The conversion of the G3 tune to G4 needs to be done manually by you or your tuner. I will attach a doc that explains the process a bit better. G3 to G4 Upgrade.pdf

Not for a 3TG, not many of those things around... However one of the 3SGE maps will probably be a decent starting point. I know there are a few out there running one our older ECU's but havent seen one for many years.

4-140ohms with a 100ohm pull up should give you a voltage output range of about 0.19V to 2.92V, so your range of only 1.3V suggests something in your circuit is not right. Also the RC filter with 18Kohm and 1000uf should give a 90% rise time of about 40 seconds. I suggest double checking the values of your components and rigging the circuit up on the bench with some jumper leads or breadboard etc. Video attached to show how it should work. Fuel level circuit test.mp4

Can you please do a PC log and attach a copy of your map.

One of the firmware engineers only had a very quick look before he left tonight and the first thing he said was same as me "crank sped up dramatically". But that was just based on the one latest scope and wouldnt explain the bad idle as crank speed should be more stable then. G4X in my experience is also far more tolerant to cranking speed variation than G4+ too so that doesnt explain why G4+ is happy. He is going to have a deeper look tomorrow. In the mean time a couple of other things that may be helpful if you have time: Set up ECU logging like my pic below, this will give us some more detailed logs than PC logging. Set ignition mode to off and do us a new triggerscope or two while cranking, that will eliminate any big speed changes due to firing or kick backs. After you have done that new scope then go to >logging>ECU log file download and download all files that are in there.

@RamEvo3 I have deleted your post as this forum is not here to support other ECU manufacturers customers. You can PM stevieturbo directly if you want to discuss your system with him but please respect the forum is here for Link ECU users support - even when the subject may not necessarily be specific to a certain ECU.

Thanks, you have caught what I wanted in the triggerscope this time so I can see the reset. I will have to pass this on to the firmware team for some thoughts as I dont see what is causing the lost tooth count. It looks like the crank suddenly speeds up just as the missing tooth is going past the sensor and it gets missed. But I will confirm when they come back to me. The 2psi when cranking was just a very short spike so looks like a small back fire.

From your earlier scope either mode should work. The first step is check if you have realistic and stable RPM during cranking (say 150-250rpm). If you do then it is likely just your offset. You will need to check with a timing light. Once you have an offset that gets your timing marks somewhere close, then if still no go, try changing the offset by 360deg.

Yeah there are a few mistakes in the default Motec template. It looks like they have the wrong divisor on that channel. In the Motec dash software, go to >Conections>communications setup, then follow the steps in the pic below to get to the DI 2 divisor, you will probably find it is set to 10, change it to 1.

The V9 fuel table will probably be the closest if you have no TGV. If you have TGV the the V10 fuel table will prob be better (will need 4D table as well with TGV). I've attached you map below with V9 fuel table imported and the boost cells removed. SubaruEJ253COASTVWEthrottlepossible V1.1.pclr