SW20 3S-GE: Bad first idle (Almost stalls straight after starting)

[tiberis]

Hey All,

I'm fairly new to the toyota scene (after buying an ADM 1990 MR2) and i have been enjoying it in general so far.

The story goes that i'd bought the MR2 a few months back from a lady who didn't take... the best care of it, but it was fairly cheap and the GF liked the look of it.
When it was bought it had trouble starting at all when cold (you had to keep your foot on the accel pedal to start it and keep it alive for the first min or so), not long after buying it i notice it developing an oil leak under the distributor.
So i buy a seal kit for it and some stuff to do a general service. i do the oil/ oil filter, new plugs (old ones were gapped too large), fix the oil leak and reset the timing (it was ~12 deg advanced). After this it now managed to start by its self, but the issues are not over.

Basically what happens is when you start the car, it revs up to ~1300 as you might expect, then it quickly drops to like 400rpm, as if it's going to stall before rising and sitting at ~750rpm.
From the limited testing i have done, i know the ECT sensor is good and the cold start injector sprays fuel, but i am stumped as to where to look next.
I've heard rumours that the cold start injector can be held on by the ECU? is this true/ the source of my problem?

Any info would be greatly appreciated.
Cheers.

[Punknoodle]

Just unplug the cold start injector to see if it makes a difference. Can you smell fuel? Do you have little black spots coming out of the exhaust? That would indicate a rich condition, possibly caused by the injector.

Did you adjust the timing with the computer in the timing adjust mode? If not the computer will be compensating for whatever changes you made anyway.

One more simple thing to do is check your fault codes, then pull the EFI fuse, leave it out for 30 seconds, then put it back in. This will reset the ECU.

If all that still doesn't improve it, I guess the next step is to make sure one of the injectors isn't leaking while the car is off. Does it only do it first time? Like if you start it, it nearly dies, then it comes back up to idle, then you turn it off straight away and repeat does it do the same thing?

[tiberis]

Thank you for your prompt reply.

From memory, when i tried unplugging the CSI last time, it made the car very hard to start, but that was before i had done the service/ reset the timing. It used to have black spots coming out back then as well, not so much any more.

Yes, i followed the service manual and made sure to bridge the correct pins while setting the timing, it made a huge difference to the idle.

Again, i checked the fault codes before the service i did and all i got back was code 21 (i think it's 21) or which ever one meant o2 sensor. As of yet i have not looked further into this (though i should reset ECU and check if it still has the error) mainly due to the car only having an issue while it starts cold, at which point the o2 sensor isn't heated correctly so the ECU ignores it any way correct?

It does happen every time as long as it is cold. the car can be left for up to an hour or so and it will start perfectly, but any longer than that and it has cooled down enough to have the same issue. Not sure if it is worth mentioning here that it has no high idle when the engine is physically cold. If i trick the ECU into thinking the engine is permanently cold (placed a 2.2k resistor in the ECT socket) then it still has the same issue when starting, but once it is warm, it will high idle (at ~1300rpm) like it should be doing when it is cold.

[RT40Corona4AGTE]

your 02 sensor is referenced and used to correct AFR's under low load conditions (cruising, idle, etc). I would suggest fixing this problem (new sensor or a good used one) and possibly even replacing your temp sensor. also you could try opening your ecu up and replacing the capacitors, these old ecus suffer a lot from leaky caps, and they can cause all sorts of problems when they fail.

[tiberis]

I was under the impression that the o2 sensor data was not valid until the sensor had reached operating temperature, hence i had not considered it to do with first cold idle. It is in the pipeline to be replaced though, just haven't gotten around to it yet.
I have also verified the ECT sensor as working according to the data from the service manual, so i don't think i need to replace it.

I do know about leaky caps though, might be worth a look.

Thanks for the info so far guys!

[thechuckster]

O2 sensor won't be used for a cold idle - the sensor has to heat up and the engine be at operating temps before it's used.

It sounds more like the aux air valve (or idle speed control valve) is borked or blocked. Given that you say the car wasn't well maintained, it could be just blocked with crud (either oily shit the air path thru it or crud in the heater water jacket on it).

http://www.toyodiy.com/parts/p_G_199...MZFQ_2211.html

[tiberis]

That is what i had previously thought.

I was hoping it wasn't the idle speed control valve (i figured that it would idle ok after a few seconds, so the ISCV must be working (also, if i trick it into thinking it's cold, when the engine is hot, it idles high).
Is there any trick/ secret way of checking/ cleaning the ISCV without breaking the coolant system? I'm happy enough to check and clean it to try and solve this issue, but the coolant system seems quite happy at the moment and i have heard the horror stories of having to bleed the system in an MR2.