Cold Idle Blues

[93_Clubman]

No a/c fitted. Lights on, maybe a 50rpm drop if that. Depressing clutch drops the rpm about 100. Engine behaves perfectly when warm.

...which as you say suggests some sort of a problem with the cold start system, which seems to check out ok...

you mention you "got a '91 mx5 recently with cold idle issue" - has the car had this problem since you got it, & can you check with the previous owner to see if they experienced the same problem.

only other thing that comes to mind is that when cold you have an air leak which seals when warmed-up.

[Locutus]

No a/c fitted. Lights on, maybe a 50rpm drop if that. Depressing clutch drops the rpm about 100. Engine behaves perfectly when warm.
Thanks
Mark

i thought idle speed should increase in both of the above situations?

sounds to me like an electrical problem. what does your idle speed do when the engine is at operating temp and your cabin ventilation fan is switched between settings 1 & 2?

[PUR157]

sounds to me like an electrical problem. what does your idle speed do when the engine is at operating temp and your cabin ventilation fan is switched between settings 1 & 2?

This sounds just like my problem I've been having intermittently for the past half year.... it started doing it again just last weekend...

Sorry for the hijack squid but what would the above symptoms point to locutus?

[Rotary]

Engine needs to be extremely rich when starting, and cant be too rich easily, as most of the fuel will stay liquid if there is no heat to vaporise it
the standard warm mixtures, will not be enough to run a cold engine becuase of the lack of vaporisation, and this can cause it to run as you say
if it runs fine when warm then it may be related to the Coolant Temp Sensor,

But it could also be a general lean condition (cold and warm) due to another sensor issue or lack of fuel flow, and may not be bad enough not to upset it when warm, but any leaning when cold may stop it working.

I'd try the Temp sensor first. if connected properly try dissconnecting it and see how it runs(when cold and warm)

[Locutus]

This sounds just like my problem I've been having intermittently for the past half year.... it started doing it again just last weekend...

Sorry for the hijack squid but what would the above symptoms point to locutus?

there are various electrical switches/triggers which tell the ECU to alter the idle speed. the ones that i know/heard of are:
- neutral switch
- clutch switch
- ventilation fan switch (activates at fan speed 2 or above)

there are probably a few more. i think these are all wired in parallel so the ECU knows to raise the idle slightly to prevent stalling. my guess is if there is a malfunction somewhere in this circuit then the base idle can't be set properly or the ECU is unable to detect extra load on the engine at idle, resulting in a car that stalls intermittently.

[Rotary]

Good chance it could be related to what Locutus is saying, and could be just a heavy carbon buildup in the Throttlebody near the openings for the Idle adjust operation.
2 Days ago I cleaned a half mm layer of carbon buildup in mine and its done a relatively low 68k, after i started it the idle went up to 1900rpm without touching any idle adjustments

While this is possible, you should still be able to keep the engine runing between 1000-1500 with light throttle, rather than have to keep it above 2000 as you say

[Rotary]

The engine needs a very rich mixture at startup because some of it settles as liquid and lowers the Fuel Air ratio, hence the need for Cold Enrichment. over richness wont be your problem here, if it was then it would get worse as temps rise not better.

Doubt the carbon is an issue if you cleaned it,
You said it runs fine when warm, right, or does it stall occasionally?

when you start it(no throttle) does the idle go up above 1k temporarily then stall, if so your idle adjust circuit is not totaly offline,

easy way to see if its fueling related is get a can of starter fluid from your Auto shop, spray a good amount past in the Throttle with butterfly wide open, before you start it, and if it stays on longer than before, your problem is fueling related, if not then probably something else.

[dab]

Another way to check if it is fuel related is to clamp off the fuel return line before you start it and see if it runs any better when cold. If it does this can indicate to poor fuel flow/pressure. Which could mean a weak fuel pump, dirty injectors, coolant temp sensor for ecu, ect...

Also disconnect the vacuum line to the fuel pressure regulator before you start it, and again once its running and just after it stalls. You want to look for any fuel residue in the vacuum line. FI their is any present it means you will have a faulty fuel pressure regulator.

Where did you find your information on the ISC and AV as it would be handy to know. thanks. (edit: don't worry found the link in your first post)

[93_Clubman]

Rotary, I believe you're right with your call on the coolant sensor. This was a common problem which you hear little of these days. JBT & myself responded to a post about it sometime ago, & JBT found/ took a photo or two as there are two coolant sensors on the NA6, one, IIRC, specifically for cold start, at the back of the head, while the other is just below it. It may be that in diagnostic mode that the former sensor is not bypassed when the engine is cold. The thread is quite old so it might be in the archive, but I'll see what I can find.

see last post page 1 & then page 2 - unfortunately JBT's photo's not available anymore, but the sensor in question (Water Thermosensor not the temperature guage sensor) is at the top rear centre of the head (see just over half way down miata net cooling system page for a photo).
viewtopic.php?f=29&t=16716&st=0&sk=t&sd=a

http://www.miata.net/garage/CoolingSystemProblems.html

[Rotary]

That makes a beter case for it, I haven't come across a faulty one(in my limited experience), but they're built quite sturdy and Thermocouples can usually take a beating
I think the main thing in this case is the car runs perfectly while warm, so the chance of it being another sensor is significantly reduced, but never totally ruled out with so many variables.

Fortunately its an easy one to check with a multimeter or just by removing the connector and sending the ecu into limp home mode(rich mixtures)

[Rotary]

5) Using a pair of pliers, pinching the fuel inlet line to the regulator almost shut, increases the idle to self sustaining (can control by the pressure on the line!). Pinching the other return end of the fuel rail kills the motor. Looks like its running too rich.

The fuel line connected to the regulator is the return line.

The regulator can only regulate the fuel coming toward it(not passed it), and that must be the side the injectors are on(The Rail),
so the fuel inlet to the Regulator is the fuel rail itself.

The pinch test shows the fuel pump is working well so rules that out

[Rotary]

Actually with the pump there still may be a slim chance its running weak, with a lower overall pressure, the only way to rule it out is by pressure testing it with a gauge, and should be within specs in the workshop manual.
Although Lower pressure is usually from a blocked fuel filter.

As I mentioned earlier there may be an overall lean condition occurring, maybe not enough to affect it when warm, only cold where its most sensitive. Could be the fuel Filter, Fuel pump, regulator, or something else

If your certain it’s the right Water temp sensor according to what 93_Clubman was saying, then the resistance is good

I'd change the fuel filter, but one thing that pops out a bit is you said no change at all when the vacuum hose is removed from regulator.
Removing the regulators vacuum hose at idle should set the pressure at a higher level than it should be for idle and should richen the mixture slightly, try this again from cold start without the vacuum connected, then see if it stays on longer.
Its only a very blunt way to check it but if it does stay on for a long time then the Reg may be ok.
Again I think sticking a fuel pressure gauge on may be the only way you know what’s going on, because they(Reg & Pump) may appear to work but may still be below the required specs

Also not sure if you mentioned sparks, but if they’re close to a year old then I would change them as well

[93_Clubman]

If your certain it’s the right Water temp sensor according to what 93_Clubman was saying, then the resistance is good

Just for clarity the centre top sensor is the sensor concerned as per the photo caption. The gauge sensor is the one below it & to the right.

image009cropped.jpg

[Rotary]

WRT to the filter causing the pressure drop, at idle the demand is not that great surely, compared to warm (and cold) high speed running (where its fine). I would expect the pressure to build up ok at idle if the regulator allows it. It does sound like a marginal pressure condition affecting the cold idle though...

That’s what I believe, only mention it because the fuel pump can cause some strangely varying symptoms, and the filter being a service item is worth changing first(if old)

Like the thinking but not sure making a gauge would be accurate, It shouldn’t cost too much just for a quick pressure test, if you can’t get a loan of one. It will tell you if it’s the pump, regulator or neither, would save buying parts you don’t need.

[Rotary]

If the test didn’t work, it would have made better case for being faulty, but that’s only half the operation of a regulator, the other is to set a base pressure which must be tested for spec with a Gauge.

Still possible that the issue in this case may have developed progressively, and the feedback system (closed loop) with oxygen sensor compensated for it during warm idle, but not during Warmup (open loop),
So still the chance of sensors working a bit out of specs (Air Temp Sensor, AFM)

fuel injection systems have so many variables. So once narrowed down the only way to know for certain is by testing.