VN V6 Corolla - Haltech can't get to AFR of 14 without stalling

[stradlater]

I've been trying to get my mates VN V6 running properly with an aftermarket computer (haltech Sprint 500) and an Innovate LC-1 wideband O2 sensor.

Now I know this is a toyota forum, but the engine is in a corolla.

I have configured and tuned the sprint 500 on my engine quite well, I'm therefore confident that I know how to get the thing going. I have a wideband o2 setup too, but I don't have the same kit he has.

My setup seems to work pretty well and the O2 signal comes through relatively smoothly, no great fluctuations or anything of the sort.

Now, we have his engine up and running with the Sprint.

However, it's a biatch to start and even with Deka 600cc injectors, it seems to require a BUCKET load more fuel than my engine does to run. Whilst I'm not too worried about that at this stage it does beg the question.

I was playing with the tune yesterday and found that I could get the AFR to hover around the 11 mark ok and the engine seemed to idle OK. But then when I started to lean it out to try and get it closer to 14, it started to run REALLY badly and almost stall. At the same time the AFR would then start to fluctuate between 12 and 17, almost within an instant. WHich I thought was weird.

So, all this leads me to - I can't get the AFR to match a comfortable running state on the motor.

Theories I have are:

- the o2 sensor setup isn't wired quite correctly and is therefore reading incorrectly (bad earth perhaps)...

- The engine is blowing oil smoke, perhaps this is confusing the o2 sensor? (blown turbo seal suspected)

- There is something crazy amiss with this engine such that is actually can't run in the stoich region and needs to run rich, can't think of any reasons why an engine would do this though...

- Vacuum leaks did also cross my mind, but it's MAP sensed, so I can't see why that would make any difference to the AFR's.


Anyone got any other ideas we could stab at?

[oldcorollas]

what is the injection parameters set up like? doesn't have something odd like 1 injection per 2 cycles or anything?

fluctuate between 12 and 17 is probably missing, and leting a bunch of air through which reads as lean.

is possible the sensor is not happy as well. check wiring.. and check with narrowband to see if it really isn't getting past stoich.

it could be that one or two cylinders have issues, and the O2 readings you see are averaged, but not indicative of what is happening on good/bad cylinders?

[stradlater]

It's Semi-sequential firing, which in Haltech terms is basically like wastespark ignition only for injection. I use this on my engine and it works a treat.

Regarding the fluctuation - It does it SOOOO quickly that it seems hard to believe a sensor could possibly see something that quickly and convert that into a signal, and then back again. Like how could you possibly bounce betwen roughly 12 and 17 MULTIPLE times within a second? Now the numbers aren't precise, it ain't exactly 12 and 17, but you get the gist.

It IS however running 1 O2 sensor on one bank only. I wouldn't have thought this would be a problem, but perhaps it is?

Would oil effect the sensor?

Interesting thought on the cylinders. Perhaps some spark plugs are not doing what they should....

Will check over that too.

[oldcorollas]

"wastespark" injection should be fine.. i run that on the MS too

narrowband can fluctuate very quickly, but wideband should be a little slower.
could be interference with the O2 sensors ground, or signal wires.. could even be the unit or the sensor itself (like if it has cracks in it perhaps)
would be worth scoping to see what wires are doing what

if everything is mickey mouse, 1 O2 sensor on one bank should be fine, but since it is heated, it could go where the exhaust joins too..

[stradlater]

I thought the WB02 had to go somewhere within about a meter of the back of the turbro?

[oldcorollas]

it depends on temperature.
placement is usually to limit the max temp the sensor sees.. but before the cat (if there is one...)
doesn't say specifically here, but gives lifetime for <300C exhaust temp
http://www.bosch.com.au/content/lang...r__Cat_WEB.pdf

WBO2 say around 1m.

it can be longer... otherwise the lambda readings of tailpipe sniffers would be totally and utterly wrong...

[thechuckster]

WB02 is heated - so as long as it's before the cat and not downstream of any leaks, it'll read accurate. There will be a miniscule lag between ignition event and sense time, but for your engine, that will be of minor consequence.

Oil would eventually clog up the cell barrier in the WB02 but those units are internally heating so minor oil vapour will be burnt off.

I'd also look at injector timing/wiring - I'd hazard a guess that the injectors are banked incorrectly and are firing after the valve closes. Fuel starvation combined with high duty cycle could also be:
- wrong impedance
- wrong fuel pressure
... to name a few.

Definitely pull the plugs to see how the thing is combusting.

note: you can also slow the sampling rate for the innovate down.

[stradlater]

So the location of the o2 sensor aint' the thing...

I will go back and check the injector timing, though I think it's all OK, I will still check.

The impedance is more than likely right as I've checked it a few times, but the fuel pressure is low, more specifically it is STOCK, which may not be good enough for the Deka's.

More investigation requried!

[oldcorollas]

it could just be that the combustion chamber and fuel mixing is so por that it will only idle nice with 11:1 mixtures...
have you mucked around with the timing at the same time?
if it is hard to start, then is it lacking fuel? or spark at the right time?

i know with the MS on the 4K that it likes to run rich at idle (but.. cam..) but with timing control it would be easier to make a nice idle zone

when you say a bucketload more fuel.. is that from the PW in the table? or actual amount of fuel?
i'm not familiar with Deka injectors, but maybe they have a slower opening speed?

does the injector impedance/resistor match what the haltech thinks it is?
they need more than 40psi? what si stock VN pressure?

[MWP]

LC-1 WB02's are actually quick enough to read the rich/lean pulses from individual cylinders.
So that fast moving 12-17 will be cylinders not firing causing a lean spike.

Is this just idle you are having problems with?
Stoic is just a nice number.. in reality when tuning it doesnt mean much.
If the engine needs to run richer to idle properly, then run it richer... there is no harm in doing so, and it wont use much more fuel.
Also, whats the ignition advance at when idling? Again, set it to whatever is required to get a smooth engine. Sometimes it can be as much as 20deg adv.

[thechuckster]

most engines will idle at all kinds of AFRs and timing - tho a smooth/stable idle requires more finesse.

Personally, i'd put aside the AFR readings for now and go back to basics:
- rail pressure should be 35-40psi (engine off)
- injector wiring/batching needs to suit the crumpledoor engine's firing order
- does it have good spark?

Are you using the factory trigger wheels on the V6? or a timing ouput from it's funky ignition box (that hides underneath the coil packs).

[stradlater]

Stock VN fuel pressure is ~36psi ish.

The injectors are low impedence but we have injector resistor booster boxes there to boost it, I will however review that config to make sure we have the right numbers in there, we may have it cross wired or something.

The running we are having at all stages, infact at the moment it doesn't have enough juice to power itself anywhere, so I suspect that there's somethig more serious at play.

As for ignition advance, at hte moment I"ve bumped it up to ~35 degrees to see what happens. It runs OK but not well enough for that to be the 'problem'.

That said, the VN harmonic balancers dont' have timing marks on them, so we are timing it based on what we have measured, we could be a few degrees out in terms of timing alignment - doubt at this stage that would make a huge amount of difference.

[stradlater]

Chuckster:

Rail pressure - gauge hasn't worked, we need to go back to the drawing board on that one. It's stock at the moment. Will check again.

Injector wiring SHOULD Be right, but again, time to check.

Spark also SEEMS to be good, pulled each lead and checked with another plug.. Again, will check again.

We tried to get the stupid crumpledoor ignition box thing to work, and it's f'ing useless to say the least. So we have ditched that and modified the standard trigger wheels to get it to work, and that has been a dream compared to anything we may have tried with the DFI module.

[oldcorollas]

not something stupid like air temp sensors or TPS not connected? (i did that last weekend )

[stradlater]

Nah, the air temp corrections tables are not in use yet, and it's not TPS driven...

The MAP sensor shows some interesting results, but appears to be reacting appropriately. (0PSI on cranking seems a little odd to me.)