Air temp sensor ( racing type ) too much variation at idle leaning when very hot.

R

Rev

Member +
From what I have read the air temp sensors can override all other sensors which is what I seem to be experiencing.
As you know NA has air temp on air box where as Turbo has it in the inlet manifold after the throttle plate.

I think the idle heat from the pcv plus heat in manifold and hot bay air through the idle ups are too much for the quick reacting sensor and its much shorter design picks up direct heat from the aluminium walls of the manifold.

So do you think it is ok to move the turbo air sensor before the throttle plate say on the silcone coupler similar to this ?
http://forums.corral.net/forums/superchargers/1412945-lightning-iat-heat-soak.html


20130318_163930_zps6f3750f1.jpg
 
Last edited:
R

Rev

Member +
Thanks, as you say gives partial improvement may be worth a try but this sensor is so close to the wall compared to stock and so sensitive
I am thinking it is heating up a lot from the first hot layer of air not just from the metal contact. Maybe I can extend the sensors length a little as I like your idea to improve things with the sensor in the stock position.
 
Last edited:
S

Skalabala

Member +
Have you tried to plug the sensor out and then drive with the car?
No money for stand alone? :p You can make your own air temp correction graph.
 
R

Rev

Member +
Unplugging the sensor ( to 20degC default ) works to some extent but this is a step in my grand plan to have the idle actually work with the A/C on.
I have thought maybe 2 sensors, stock and racing sensors switched by TPS but if it is safe to just use one sensor just outside the throttle plate that would be my preference especially as the stock one was heat soaking as well . The car runs superbly apart from idle so not thinking standalone at present.
 
Last edited:
S

Skalabala

Member +
Rev said:
Unplugging the sensor ( to 20degC default ) works to some extent but this is a step in my grand plan to have the idle actually work with the A/C on.
I have thought maybe 2 sensors, stock and racing sensors switched by TPS but if it is safe to just use one sensor just outside the throttle plate that would be my preference especially as the stock one was heat soaking as well . The car runs superbly apart from idle so not thinking standalone at present.
Click to expand...

Do you want to cheat the management?
What does the air temp sensor do in a stock car? Makes fuel richer as it goes colder and takes timing away when its colder?
 
R

Rev

Member +
Skalabala said:
Do you want to cheat the management?
What does the air temp sensor do in a stock car? Makes fuel richer as it goes colder and takes timing away when its colder?
Click to expand...

I would be ok cheating management if it is wrong.
eg.
> The fast sensor is picking up too much heat from manifold casing that won't go in the engine because of thermal layering that occurs inside the manifold so management is wrong.
> The fast sensor may pick up so much heat it exceeds the ecu saftey threshold and dump fuel in management also wrong.
> PCV valve heat stream touching the sensor showing extreme intake temps before the stream mixes properly also wrong.

I think this may be fixed by moving the sensor to just before the throttle body .

Not sure whether bay temp is added to the manifold by idle ups so the management will be partly wrong this time as bay heat will not be effecting the engine cooling to the same extent.

My idle setup is a worse example than what most people experience so I have a big hill to climb on this one basically with too much heat the ecu leans too much and as revs drop knock occurs .
 
Last edited:
S

Skalabala

Member +
Does any one know how the compensation graphs look for the air temp? Think we might never know :(
Yes you are up to something bud. Just move original one. Idle up heats it up real bad yes.
CAI is also important and works good :)

So you are saying when sensor leans out after hot idle and you give it a little boost then you get knock?
If so then there is something else wrong bud.
 
R

Rev

Member +
Moving the sensor seems to fix a lot but so much experience on this site I thought I better ask before attempting anything.

This is only an Idleing knock issue say mainly 750-950 rpm - on very hot days, after hard boosting and when oil is lower than 80% full. ( Worse with AC load but will try to fix that once this fueling issue is resolved ) Once I hit 1300rpm knock is usually zero. Hot condition, Lean burn will advance flame ignition probably giving similar effect to A/C load slowing the piston so spark arrives too early.
 
Last edited:
S

Skalabala

Member +
For a car to knock on idle the fuel or timing must be so far out that it won't even idle bud.
If it knocks of to lean mixture on idle it won't idle for sure.
 
R

Rev

Member +
Ok so lets say it is idle roughness rather than knock.
Most effected by the temp sensor starts gets rougher with excess heat and gradually worse as revs drop.
What I should explain is I have manual ecu with auto tran. Auto needs more timing in drive at idle for gearbox load. I am on the limit so any variations made by the ecu especially to fuel make a big difference.

I am thinking if the fast acting sensor is more stable before the throttle plate maybe I should try 2 fast actings one before for idle and one in stock position ?
 
Last edited:
R

Rev

Member +
Skalabala said:
For a car to knock on idle the fuel or timing must be so far out that it won't even idle bud.
If it knocks of to lean mixture on idle it won't idle for sure.
Click to expand...

You have got me there @Skalabala I checked the fuel and safc2 was a little lean -37% at 800rpm my injectors are equivilant -32.4% at idle fuel pressure. It seems the temp sensor is very sensitive to leaning together with hot intake temps it goes much better with -24% at 800.

Also I hot wired the idle up from the cabin and can hold better idle with AC on but the behaviours are still present ie . dropping revs after about 3 secs stationary and intermitant big drop of revs goes up and down as if hot air touched the fast acting sensor for a second.

More research needed on this -
I have to get AC gas filled to tune under full load.
Also I noticed when hitting the brake the idle up goes higher than my idle and into a lean part of the map maybe this is triggering the ecu to reduce timing in the heat which is why revs are dropping after a few seconds or maybe its just a loose connection somewhere ?

Thanks for help so far everybody this has just fixed a large chunk.
 
Last edited:
R

Rev

Member +
Ok good to know I have that too. The delay I mentioned was delay in revs dropping after the braking idle-up revs up.
I will do some tests and see if I can tune the rev drop out, the Ac and idle up are now booked in next week so will know more after that.
It would be good know whether leaning is definitely due to the fast acting sensor or not before changing positions etc.
 
R

Rev

Member +
I haven't tried that yet have to make the cable connection etc. Have had no time to do things properly just trying to keep things running in the heat.

Skalabala you are right now would be a good time after the repairs to get some timing up on screen and see if it is the ecu changing timing or not. I'll give it a go when I have weekend available in a few weeks time.
 
Last edited:
R

Rev

Member +
Have not yet run software connected to car. I have seen timing move with leaning when using timing light previously. If the connection works my plan is to check 2 ecus, the stock one and another identical ecu that has been chipped.
 
Last edited:
You must log in or register to reply here.
Top