coolant sensor and open loop technical question

[Morgenster]

Hi,

my idle gets worse with temperatures dropping this winter. That is up until the engine warms and then it just works fine-ish. I've tried to pinpoint a vacuum leak but can't find any by spraying brake-cleaner (which is a carb cleaner AFAIK) in a very general way being cautious.
- Is there any place I really shouldn't spray on or can I just run wild with the can?
- From reading another thread I saw someone suggesting the coolant temp sensor to be at fault. How does this relate to the problem in any way?

[markus]

if the coolant temperature sensor was bad then it would throw a fault code. however you can check the resistance of it if you have a bentley manual. It most likely sounds loke a vaccum leak and id check the intake boot and spray around the intake manifold gaskets as well. Do you have an M20 or an M50?

[Morgenster]

It's a technical thing I'm really curious about: how does the coolant temp sensor play into open loop operation (assuming closed loop relies solely on O2 sensor)?
Since it's a euro car the classic stomp test will just not work on mine. I really need the code reader and don't have it.

I'm thinking this:

Suppose fuel mixture and ICV opening are controlled based upon MAF input, coolant temperature sensor input, throttle potentiometer input, etc.
If a coolant temperature sensor fails to let the computer know that the engine is still cold would it adapt the mixture/ICV opening to 'warm' levels? And so cause the mixture to be insufficient and idle crappily low which in turn illicits a signal from RPM sensor saying this to the computer. The computer then picks up the pace and adjusts mixture and ICV (causing a slight surge (never over 1000 RPM)). The subsequent drop in RPM after correction illicits a new response from the RPM sensor and what you have is a low-surging rumbly idle?

I have no reason to suspect the ICV since it works just fine when warm.

Another train of thought is this:

Suppose there is a slight vacuum leak near the block. It could be 'open' when the engine is cold and 'close' by heat expansion when the engine warms. That would point to PCV connector and hose or intake gaskets?

It'd be really sucky if the MAF is at fault, but if it dies doesn't the engine run really supercrappy?

The symptoms are classic but I don't feel like swapping numerous bits and pieces like many before to wind up frustrated about that.
It starts fine but idle immediately drops to some 500+/- RPM and then after applying throttle the RPMs dip real low only to surge/dip/surge/dip and level out to +/-500RPM. Rarely it stalls but 've had it happen and find it embarassing. again: the surges never reach higher than 1000RPM.
When warm the engine idles at some 750RPM and is steady.

[Morgenster]

bump for DME knowledge

[Bellicose Right Winger]

Does this problem occur on gasoline or LPG?
Unlike carb cleaner, brake cleaner is usually non-flammable, so it's effect on locating vacuum leaks is questionable.
My '93 has a slightly different engine (M50TU) but I had a cracked intake bellows and the smaller of the two PCV hoses had broken where it connects under the intake manifold.

I wouldn't rule out ICV because it's ok when warm. The oil and sludge that collects in the ICV becomes less viscous as it warms up from engine heat.

Cold engines need more fuel then warm engines. Coolant temp sensor adjusts mixture based on engine temp, in both open and closed loop modes. Same for intake air temp sensor. Use DMM and check resistance of coolant temp sensor as Markus said.

Paul Shovestul


It's a technical thing I'm really curious about: how does the coolant temp sensor play into open loop operation (assuming closed loop relies solely on O2 sensor)?

Since it's a euro car the classic stomp test will just not work on mine. I really need the code reader and don't have it.

I'm thinking this:

Suppose fuel mixture and ICV opening are controlled based upon MAF input, coolant temperature sensor input, throttle potentiometer input, etc.
If a coolant temperature sensor fails to let the computer know that the engine is still cold would it adapt the mixture/ICV opening to 'warm' levels? And so cause the mixture to be insufficient and idle crappily low which in turn illicits a signal from RPM sensor saying this to the computer. The computer then picks up the pace and adjusts mixture and ICV (causing a slight surge (never over 1000 RPM)). The subsequent drop in RPM after correction illicits a new response from the RPM sensor and what you have is a low-surging rumbly idle?

I have no reason to suspect the ICV since it works just fine when warm.

Another train of thought is this:

Suppose there is a slight vacuum leak near the block. It could be 'open' when the engine is cold and 'close' by heat expansion when the engine warms. That would point to PCV connector and hose or intake gaskets?

It'd be really sucky if the MAF is at fault, but if it dies doesn't the engine run really supercrappy?

The symptoms are classic but I don't feel like swapping numerous bits and pieces like many before to wind up frustrated about that.
It starts fine but idle immediately drops to some 500+/- RPM and then after applying throttle the RPMs dip real low only to surge/dip/surge/dip and level out to +/-500RPM. Rarely it stalls but 've had it happen and find it embarassing. again: the surges never reach higher than 1000RPM.
When warm the engine idles at some 750RPM and is steady.

[Morgenster]

Does this problem occur on gasoline or LPG?
Unlike carb cleaner, brake cleaner is usually non-flammable, so it's effect on locating vacuum leaks is questionable.
My '93 has a slightly different engine (M50TU) but I had a cracked intake bellows and the smaller of the two PCV hoses had broken where it connects under the intake manifold.

I wouldn't rule out ICV because it's ok when warm. The oil and sludge that collects in the ICV becomes less viscous as it warms up from engine heat.

Cold engines need more fuel then warm engines. Coolant temp sensor adjusts mixture based on engine temp, in both open and closed loop modes. Same for intake air temp sensor. Use DMM and check resistance of coolant temp sensor as Markus said.

Paul Shovestul

It happens with gasoline.
I'll check resistance on the coolant temp sensor and get something more flamable than brake cleaner to do the testing. As for the ICV: I get your point, so I'll try to get the sucker out without scraping off my delicate skin :) .
If I find wholly flamable carb cleaner, where do I definitely NOT spray except the MAF?

So does this mean that a consistently low idle speed with low surges doesn't necessarily mean no unmetered air is entering? Cause I thought that would lead to higher surges and a runaway idle.

[ryan roopnarine]

unless you want to kill your catalytic converter or oxygen sensor, use something that says "converter and o2 sensor safe" on the can. afiak, i've never seen a can of brake cleaner that says this :D carburetor cleaner works fine for the work you describe.

[Randell]

i currently have a similar problem.. when starting first thing in the morning or in the evening after work, the engine 'chugs'. The tach needle stays steady but the lumpy idle rocks the car... it got to the point where it was still crap even after it warmed up. Off idle it seems fine though

turns out the injector in cylinder 6 wasn't firing properly, replaced that and now the car is good when warm but still a bit funny when cold.. by 'cold' i mean 20+C mornings in the aussie summer! By the time the temp needle gets off blue the idle is basically good

i tested the resistance on the coolant temp sensor, both hot and cold readings are within spec as per the bently.. ICV is fairly new and clean, engine mounts brand new, replaced the intake boot that had a crack in it... i haven't done the carb cleaner thing yet i never get round to buying some! The little plastic mounts for the AFM are all cracked in half but it's held steady by the rigidity of the hoses, so i don't think that would cause a problem. I don't know how to test the AFM but the flap moves freely and there was no gunk in there

i have the car in today at a different mechanic getting a new prop shaft installed, i asked him to have a look at the engine after he's done that, i'll post if there's any new info

[Morgenster]

Hi,

my idle gets worse with temperatures dropping this winter. That is up until the engine warms and then it just works fine-ish. I've tried to pinpoint a vacuum leak but can't find any by spraying brake-cleaner (which is a carb cleaner AFAIK) in a very general way being cautious.
- Is there any place I really shouldn't spray on or can I just run wild with the can?
- From reading another thread I saw someone suggesting the coolant temp sensor to be at fault. How does this relate to the problem in any way?

OK so here it gets spooky. last tuesday when I got back home from a nice highway drive I parked and revved a few times to see what would happen in the changeover from idle to higher RPM's and back. Then all of a sudden I hear an electric pumplike noise every time the revs drop to idle and after two or three times the noise is gone.
Since then My idle is spot on perfect. Cold, hot, you name it, idle stays at 1000 rpm when cold and drops to 750 when warm whereas before it would be more like 600 when cold and 700 when warm. And it would start bouncing and sometimes stalling when cold once I applied throttle and let go again.

So now I think the ICV was the problem all along and it somehow got 'liberated'. The question now is how long it will last...

Also couldn't find any vacuum leaks when spraying intake cleaner, but that's no longer an issue is it?