RAV 4 cuts out/low tickover

[krusty]

Hello,

My 1998 2.0 Toyota RAV 4 has an AEB/Leo controller and a Bigas vapouriser.

Since installation I have never been able to get it to tick over correctly/consistently.

When it starts it's fine but when it warms up the revs drop to around 700rpm (should be around 1000rpm). After a run at high revs it often stalls when returning to tickover. When sitting in traffic the revs seem to progressively drop to around 200rpm - a quick throttle blip restores the revs for about 10 seconds and then it progressively drops again. Sometimes it will tickover as expected. It is quite frustrating. I've endlessly adjusted the idle bleed screw and tried various settings in the software but have never managed to get it working properly.

Has anyone got any ideas? Am I missing something obvious?

Cheers

[Gilbertd]

Do you understand what the settings in the software are showing you? Blindly twiddling the idle screw will make things worse rather than better unless you happen to strike lucky. You need to get the actuator opening the same at idle as at open throttle with the lambda flicking from green to red all the time. If it is sitting at one end of the scale all the time, possibly with the actuator against the limit, then it needs setting up from scratch. This all assumes that the lambda sensor is good and giving sensible readings too.

[Tubbs]

I'm sure there is something odd about the idle air control valve on them. Damned if I can remember what it is though....

[krusty]

Thanks for the responses.

I have a fair understanding of the aeb system as i've used it on a couple of range rovers and a jeep and it works perfectly on those. I'm familiar with the requirement to have the actuator moving between rich and lean constantly.

I've never been able to tune it so that it does this at tickover. It drifts over to the rich side and the actuator drops. I have set a default of 85 on the actuator and it makes no difference. Most of the time the vehicle nearly stalls when coming off the throttle and it often seems to choke up when ticking over for a while. But not always.

I'm not sure if the system is suited to the rav 4 as it has coil packs? Or is it something else? Could it be the lambda, it seems to work correctly when the throttle is open?

Cheers

[Gilbertd]

Open the idle limits to 255 upper and lower then see if the lambda starts flipping at idle. Assuming it does, that should give you a rough idea of where the actuator needs to be at idle. Then rev the engine and see where the actuator sits out of idle. The two want to be roughly the same. If it's going rich at idle that suggests that the idle bleed is screwed too far out. On an R90E the idle is normally always left screwed fully in and I would expect the Bigas to be similar. A rich mixture at idle is the first sign that the vaporiser is reaching the end of it's life too.

[Tubbs]

Where is the idle air control valve ? It's bugging me !!

[krusty]

Open the idle limits to 255 upper and lower then see if the lambda starts flipping at idle. Assuming it does, that should give you a rough idea of where the actuator needs to be at idle. Then rev the engine and see where the actuator sits out of idle. The two want to be roughly the same. If it's going rich at idle that suggests that the idle bleed is screwed too far out. On an R90E the idle is normally always left screwed fully in and I would expect the Bigas to be similar. A rich mixture at idle is the first sign that the vaporiser is reaching the end of it's life too.

Should I remove the default lock currently set at 85?

[krusty]

Where is the idle air control valve ? It's bugging me !!

Wouldn't that affect the running on petrol as well? It runs perfectly on petrol.

[Gilbertd]

Default lock just stops the system from setting the default itself. If the actuator opening at open throttle is wildly different from at idle, then it will adjust to suit one and be miles out on the other. Leave it as it is at the moment or you'll just end up screwing everything up.

Idle air control valve will affect both, not sure what Tubbs is on about. The only thing you want to be touching is the idle bleed on the vaporiser.

[rossko]

Idle air valves feed extra air bypassing the throttle disc, that's their job. If the LPG mixer is fitted to the throttle body, depending where the IAV picks up its air supply from, it may feed fresh air rather than LPG/air mix. Obviously, idle mixture would be rubbish. It is sometimes necessary to re-route the IAV feed.
I can't remember if that is needed on this engine. It has a gert big inlet plenum/manifold thing, which will exacerbate any idle stability issue for mixed LPG. Throttle at the bottom? risk of LPG pooling down at the throttle end if you overdo the reducer's idle bleed.

[Gilbertd]

That would make perfect sense, and explain the problem, if the IAV is bypassing the mixer. A picture of the intake should give us a clue.

[Gilbertd]

What engine does your RAV 4 have? My daughters Toyota has the 1ZZ-FE engine so I've got the manual for that. That shows a standard throttle body mounted IAV that wouldn't bypass the mixer.

[Tubbs]

Or maybe it's a 2 or 3 wire IAV stabiliser you need. See how you get on turning the Leo 255/255. Leave the default lock on. And see where the stepper actually hovers. As Gilbert suggests.
As I say. I'm sure there is something odd about that early rav 4. I wasn't trying to be cryptic ! Honest !!

[krusty]

Only just seen your messages. Thanks for the responses. On the way home last night I removed the default lock and set the idle steps to 255/255.

The stalling has become worse.

The default moved up to around 115. When it returns to idle it eventually falls to around 45. It takes anything from 5 - 20 seconds to get down to 45 and the lambda is indicating rich the whole time. Once it gets down to 45 it then starts flicking between rich and lean and the car ticks over normally.

If i press the clutch at speed the actuator stays around 115 and the lambda stays rich until i've coasted to an almost stop. It often stalls and at best hovers at around 300 rpm on the verge of stalling. Could this be an issue with ram air into the intake or hot air as the intake is behind the radiator (i've fitted a k&n as I've gone with all my other lpg vehicles)?

The engine is the 2.0 litre (no idea about model numbers).

I tried setting the default lock to 85 and leaving the idle setting at 255/255 on the way to work this morning. It didn't make much difference, the car still stalled frequently from speed and took a long time for the actuator to drop to 45 ish.

I'll take some pictures of the engine at lunch time. It has a square plastic box on the air intake pipe with a sensor throught the pipe if this helps.

Thanks guys.

[Gilbertd]

If the default moved up to 115 while it was being driven, that's correct for open throttle. The fact that it drops, and the lambda starts flicking, at an opening of 45 at idle explains the stalling. It's running so rich when you come off the throttle that it's drowning in fuel. With it idling and the actuator at 45, screw the idle bleed screw in. The actuator will start to rise. When it reaches roughly 115 (or, and this is where Tubbs and I agree to disagree), slightly higher. You want it to sit at idle at around 120-125. If the idle bleed screw is fully screwed in and it still doesn't get that high, screw in the main bias until it does. Then give it some throttle and see where it sits at open throttle. If you can't get the open and closed throttle actuator positions to match, the vaporiser is toast.

[krusty]

Here are the pictures of the engine.

http://s1219.photobucket.com/user/pieea ... 6.jpg.html

http://s1219.photobucket.com/user/pieea ... 3.jpg.html

I had a quick fiddle with it at lunchtime. Having wound the screw in several turns and the idle bleed in completely. It seemed very difficult to get the tickover to settle anywhere near the default. It moved from high to low and back very slowly. It would go lean for 20-30 seconds and then go rich for the same and back again. The actuator would vary from 180 ish all the way back to 40 ish. I set the the idle steps limit to 20/20 and it would stop the actuator but it would stay at the limit at each extreme.

Very confusing

[Gilbertd]

Part of the problem is going to be that huge loop of pipe and the position of the actuator. The actuator needs to be as close to the mixer as possible so that any changes have an almost immediate effect rather than causing it to overshoot and then have to hunt back and forth. Can't really see where the IAV is from those pictures, maybe someone that knows that engine better (Tubbs?) will tell us where it lurks.

You're probably being too rough with it, you would usually adjust no more than a quarter turn at a time then let it settle. You said before that the idle would settle at around 45 with the lambda flipping. Get it running at idle with the idle bleed screwed all the way in and see where it sits. If it is low, screw the main bias in by a quarter turn and see where it moves to. Keep doing it like that but let it stabilise between each tweak.

[rossko]

I set the the idle steps limit to 20/20 and it would stop the actuator but it would stay at the limit at each extreme.

That's a Good Thing. It is often misunderstood that in a system with perfect mechanical (mixer/reducer) setup, the actuator would not move at all between idle and full throttle. It's only there to fine tune on top of the mixer aerodynamics doing the donkey work.

If there is a lot of lag (time delay) between actuator and lambda sensor - long dangly pipes, large plenum volume, long exhaust downpipes, elderly/cold sluggish sensor - then at idle (low air speeds, increased delay) you will need to restrict how far the actuator wanders in the time before any effect gets measured. That's why they provide those idle step limits.

[krusty]

I've now managed to shorten the pipes and further adjusted the mixer and it is much improved. It still dives lower than when on petrol on returning to tickover but I think this may be down to the age of the sensor etc. This is likely to be as good as it can be. It doesn't stall anymore and seems more responsive.

Thanks for all your input guys. Much appreciated.

[Tubbs]

Well that's good. Just take the idle steps under default to 0. And make sure it's rich just there. It should tick over perfectly at that.