Supra Forums banner

1 - 10 of 10 Posts

·
Registered
Joined
·
762 Posts
Discussion Starter #1
Hi guys, I'm just getting to the end of a project car for a guy. Built engine, forged rods, SP63 full kit, 720cc RC injectors, Lexus AFM, SAFCII, Walbro pump, all the usual stuff for a big horsepower buildup.

I have the engine running now, and the car is ready for paint, but I am having problems getting the thing to idle. I have a brand new narrowband sensor at the top of the downpipe, and my AEM wideband at the bottom of the downpipe. Timing is spot on at 10BTDC, and fuel pressure is 38psi with vac line disconnected. The car is trying to idle at 17:1 a/f, off-the-scale lean on my wideband. If I constantly blip the throttle the accel pump function will deliver enough fuel to get the gauge blipping into the 12:1 - 14:1 region, and you can hear the engine "clear its throat" and start to run properly. As soon as I come off the throttle, the engine note goes real fluffy and the car dies.

I have checked the cambelt timing is spot on, no intake air leaks, idle valve is working correctly. I have NO error codes showing up in diagnostic mode, the SAFCII displays normal numbers for the KV signal from the AFM. I have tried swapping the black box on the AFM for a known good one, I have also tried swapping the lexus AFM body for a standard one. I have tried going to +50% enrichment across the board on the SAFCII, raising the fuel pressure 10psi, AND running the standard AFM (all at once). With 720cc injectors, this should be enough to fill the cylinders with fuel or at the very least have big plumes of black smoke coming from the exhaust. Instead, the WB gauge just flickers onto 16.5:1 occasionally and the engine struggles along at 600rpm for a few seconds before dieing again.

I am pretty sure that this is not a mechanical problem with fuel delivery, as the car has enough fuel to run nicely if you rapidly pump the throttle, and it will need more fuel to run well at 2500rpm than it does to tick over at 800rpm. I also put a Walbro pump in this car last year, and have inspected all the lines to and from the rail. I also tried disconnecting the return and running it into a bucket, and then running the fuel pump, it's definitely flowing plenty of fuel out of the tank.

My plans for tomorrow are to see what the Vf signal is doing, take the SAFCII out of the loop (ie run the AFM signal direct to the ECU and unclip the SAFC), and maybe swap in a replacement ECU. I have also just realised that I can check injcetor duty cycle on the AVCR which is fitted, anyone know what it should be at idle? I think my own car (AEM equipped) runs about 0.8% duty cycle with 680cc injectors, does that sound right? Has anyone else got any other ideas?
 

·
greasemonkey
Joined
·
3,126 Posts
prolly too low a signal at idle whe nthe fuel is scaled down. try raising the idle to 750-800rpm. my isc is blocked and im at a 1k idle, runs great.
 

·
Premium Member
Joined
·
5,301 Posts
Bump the idle. Stock ECU has a really hard time with injectors of that size. AFC can't compensate for it, so just bump the idle. With 680's on my car, I idled at about 1000-1100 and it ran and drove 1000 times better than when I tried to leave stock idle and let the ECU control it.
 

·
Registered
Joined
·
762 Posts
Discussion Starter #4
It won't idle even if I crack the throttle open and try and hold it at a constant 2000 revs. My own car used ot idle fine with 680's and lex AFM and no fuel corrections at all.

However, I think I may have figured it out. I've never used an AVCR before. so I set it to 6 cylinders, ascending throttle, and then left it alone until the engine is run in. However, I tapped into the injector wire to read duty cycle, and I've just read in the AVCR manual that number of cylinders must be set to ONE if you wire it in this way. So I guess maybe the injectors are only firing 1/6th of the time or something weird?

Either way, the AVCR/injector wire interface is the only thing different on this car to other Mk3's I've built and tuned, so that is where I shall concentrate my efforts tomorrow!
 

·
Registered
Joined
·
6,475 Posts
hmm

the ecu should be able to handle it as the open close signal will be the same as a (usually minimum of 2 ms) regardless if the injector is 100 cc or 1200cc. So even at the smallest injector pulse width time. 2ms is never gone under. This sounds like the injector might be having trouble STAYING open. Have you tried to see what happens when you raise the base fuel pressure?
 

·
Registered
Joined
·
46 Posts
don't do anything. your car is supposed to be that lean at idle. you want typically 16:1 or 17:1a/f at ilde. you said when you snap the throttle it goes between 12a/f and 14a/f right. thats becauses when you start to give it gas then your pulse width is longer thus giving more fuel to match the air. don't get bigger injectors or nothing like that. also a good idle is between 750 and 900rmp
 

·
wastegate hose is pulled
Joined
·
4,586 Posts
TOCA0015 said:
don't do anything. your car is supposed to be that lean at idle. you want typically 16:1 or 17:1a/f at ilde. you said when you snap the throttle it goes between 12a/f and 14a/f right. thats becauses when you start to give it gas then your pulse width is longer thus giving more fuel to match the air. don't get bigger injectors or nothing like that. also a good idle is between 750 and 900rmp
:wtflol:

Adam W:

Remember, when the engine is misfiring, the wideband will give a false lean indication. It may just be running too rich.
 

·
Registered
Joined
·
626 Posts
Would the narrowband react to misfires in a similar manner? Apparently, the TCCS thinks the car is running lean, because the Vf is pegged at 5V. My wideband indicates it's lean. Wether those are false readings or not has yet to be determined...are misfires always obviously audible?
 

·
wastegate hose is pulled
Joined
·
4,586 Posts
Yes, a narrowband will react the same way since it is still only reading the oxygen content of the exhaust and inferring the A:F from that, not the actual A:F.

If you are familiar with how your engine runs, you ought to pick up that the car is missing right away. An exhaust leak before the oxygen sensor can also cause a false lean reading.
 
1 - 10 of 10 Posts
Top