3SFE - Caldina ST215 not starting / idle problem

[inno]

Hi All,

Great forum, been reading for a while but this is my first post! I have a 97 Toyota Caldina 2.0 Auto (GT). ST215. Since a few weeks
it's having some strange idling and starting issues. (done 180kkm orso). All is stock standard.

I should mention that the first time I had the problem is when I thought I ran out of petrol.. but I don't believe I did after all.
Just mentioning it anyway.

Basically, first time it just did not want to start. engine cracking, but not firing. after 30 secs orso cracking engine
would slowly start firing and finally running. Got petrol immediately and that was it I thought. well.. not so..

So. symptoms are:
1: Car generally starts fine when cold (i don't believe it ever had problems cold starting)
2: After some time idling, at least 15 mins, sometimes never, the idle speed drops suddenly almost stalling the engine.
Revs go up to 2000rpm and stay there for a while
3: If I would turn off the engine, at that point, it won't start.. just cranking. Playing with the throttle mostly
starts the car after at least 30secs cracking (sometimes it takes 15 mins..). It does not suddenly start, but slowly
starts firing/picking up and idle running rough.
4: As soon as you drive.. all is gooooooood.... even idling seems ok after a bit of driving.

It's been doing this for about a month now.

Tried so far:
1: Keep on driving and hoping it goes away
2: Trying to connect OBDII / VCI connectors with no luck. Had these from previous cars.
3: When not wanting to start. take out spark plug and check for spark.. spark = present!
2: Read out error codes. error=21 (O2 sensor).

Now, looking at the error code symptom description, it does not sound like my problem at all (even though the sensor is
probably kaput.). From what I know, an O2 sensor problem would result in bad fuel consumption, but not never wanting to start?

I've obviously searched the forum, and google, but have not found a similar case. I'm hoping the symptoms ring a bell to
someone here..

O2 sensors are generally not cheap. Happy to replace it of course if it solves the problem, but not if it doesnt. Will
get my multimeter out and see if I can test the bastard... any hints?

Oh.. sparkplugs & ignition leads replaced about 15kkm ago.

Cheers
J.

[RS13]

I've never seen a failed 02 sensor cause issues like those, aside from the increase in fuel consumption the only other symptom I've seen was a mild hesitation in the mid-range.

The 3SFE seems very sensitive to bad EFI temperature sensors, to me it sounds like its' overfuelling when warm. Next time it starts running like crap or not starting, unplug the electrical connector to the EFI temp sensor (from memory, looking at the coolant manifold that the top radiator hose goes to, its' the sensor with the green plug receptacle on top of the manifold) and see if that makes a difference. If it does, I'd check it with a multimeter then google the resistance numbers you should be getting at a certain temperature.

[Lloyd]

Unplugging can generally make them harder to start if anything. Most of the older setups defaulted to about -40 degrees if they picked up that there was no signal, so the default was to make them run super rich for safety. Pulling it may make it start, but it may also be a coincidence if it does and send you down the wrong track.

In saying that, they do give trouble. Just about any Toyota with the right shape green plug will do the job as a replacement too. If you want to check with a multimeter then you should have about 3000 ohms with the engine cold and 300 with it hot (ball park).

[RS13]

Fair enough, but I've have success using this method in the past. I guess if the signal is massively out of limits then going into "safe" mode could potentially improve running? They're cheap enough to replace regardless and I've found are often the cause of rough running in more than a few Toyotas of that vintage that I've worked on.

[inno]

Thnx for the reply guys.

I'm not too convinced about the 02 sensor myself, but I guess will have to replace it regardless. I did always find the fuel economy to be bad so that might have something to do with it. I'll try the temperature sensor some time this weekend and report back. Just wondering.. if it was that, would it not be reporting this as a fault code?

Btw.. the ABS light flashes an error code as well I believe it's 49. I have not looked it up yet but doubt it has something to do with starting problems (I'd imaginge the ABS light error codes are solely related to ABS errors...) ABS works fine btw..

Cheers
J

[Crucible]

Temp sensors (thermistors)do go out of range and if not open circuit will not bring on the check lamp. Just means that it will have incorrect resistance so ecu cant determine the correct temperature ie high resistance when hot and overfuelling. As far as checking an O2 you can check voltage on the signal wire. It should fluctuate from around 200 to 800 mv back and forth at idle.

You can check response by causing a lean or rich condition, rich it should jump to near 1v, lean it should drop to near 0v. If it doesnt respond it is faulty.

[inno]

Hi All,

Ok.. here an update. Finally got around to having a look at it again.

I did a bit of calling around and found 02 sensors are worth their weight in gold (if they work. . thas is . So decided to do a proper test.

I found out on the net there are 2 failure modes for the O2 sensors.. there is a heating element, and a sensor element. So I disconnected the sensor, and put an ohm meter across the pins for the heater element. I measured open circuit. I can't seem to find what it should be, but I have some electronic background and knew this is possibly not right. I ducked down to the pick-a-part, mainly because all the wreckers close around 1pm and found some engines with the same sensor. Interestingly, most of them were already disconnected, and all of them measured open circuit. Clearly someone has been there before Anyway. I did found one car in the end that had the sensor still connected and it did measure a resistance (400ohm from memory). So I unscrewed it and took it home after paying $14.

Connected it up, and reset the fault codes. Also I measured the temperature sensor, and measures 3kohm.. which from what I can find is good. Started the car and..... did not start.. and eventually it did.. interestingly, the radiator fan started to run immediately (cold engine) which I thought was weird.. until I realised I forgot to plug the temp sensor connector back in.. doh

Btw.. the error code 22 came up after I disconnected the temp sensor. (which means temp sensor error.. ) lol..cleard the error codes again, and after a testdrive still no error codes. Ran smoothly since, so will have to see if the issue happens again.

I'm not convinced the O2 sensor was the problem yet.. to early to tell. so will give it some time.

Will keep you updated.. thanks for all the hints..

Cheers
J

[inno]

Oh.. and one other thing I thought was worth mentioning. I found that unless I have disengaged the start-inhibitor (it kicks back in after 20secs orso on my car) I don't get any error codes out (does not even blink). Something to keep in mind for next times.

[sergei]

Here is how you test the O2 sensor (typical narrow band that is):
The O2 sensor generates voltage (it is not resistance type) so measuring it with Ohm meter will lead to misleading results.
What you do is set the meter to 2V range (digital, do not use anything below 2MOhm input impedance).
Reset the ECU
start the car
measure between Ox terminal and E1 terminal, once the sensor is warmed and car is in closed loop (hold the revs to about 2000rpm to help with that) the voltage should fluctuate between 0-0.2V and 0.8-1V at rate of about 8 times per 10 second.
If you get the voltage fluctuation at the sensor you can check the fuel trim compensation (do it at ~2000rpm) by measuring the voltage between VF and E1 (set meter to 20V range).
The output should be ~2.5V (0 compensation to the base map).
The VF output goes between 0V and 5V in 5 steps:
0V -> compensating (what ECU thinks) very lean condition, or open loop operation.
1.25V -> compensating lean condition
2.5V -> no compensation
3.75V -> compensation rich condition (ECU is leaning out the mixture, one of the sensors is out?)
5V -> compensation very rich condition (leaky injector?)

If the voltage on O2 sensor is not fluctuating, it means the ECU is in open loop. If it is 0V the sensor is probably clogged or faulty.
If the voltage is above 0.5V and stuck there then most likely something else is wrong (ECU is in open loop, and running rich).

The heater will not dramatically affect the operation of the sensor, only on cold start up.

If you have access to butane/propane torch, then it is possible to ressurect a sensor that is clogged. Additionally you can use butane torch to test the sensor at same time:

Remove the sensor from the vehicle.
Hook up the signal wires of the sensor to the multimeter set to 2V (the heater wires will have a few Ohm of impedance, between 2 and 20, use other wires for this )
Heat up the sensor end (make sure you do not overheat the porcelain/wire end) with the torch until the cage is red hot.
If the sensor is reading 0.9V to 1.2V, while the gas is passing the cage, the sensor is alive. You could further check by extinguishing the flame and just supplying the gas to the sensor while it is still hot, it should read ~1V (until it cools down).

If you are not sure which are signal wires are, trace them from diagnostic connector to the sensor connector with Ohm meter, the Ox wire is the "positive".

[inno]

Thnx for the feedback, I might give that a quick check.

Fair enough about the heater element only being used on a cold start, hence I'm not convinced it's my problem. In any case would be good to check the operation as you say.

I've had a hard time finding easy access points to the terminals, without having to disassemble or cut into the wires. This model does not seem to have the bigger diagnostic port in the vehicle bay. Just the one that's under the dash.

Cheers
J.

[sergei]

Thnx for the feedback, I might give that a quick check.

Fair enough about the heater element only being used on a cold start, hence I'm not convinced it's my problem. In any case would be good to check the operation as you say.

I've had a hard time finding easy access points to the terminals, without having to disassemble or cut into the wires. This model does not seem to have the bigger diagnostic port in the vehicle bay. Just the one that's under the dash.

Cheers
J.

Here something that might be useful to you:
http://gt4.mwp.id.au/Electrical/Caldina ... trical.pdf

It looks like the ST215 caldina is too new, so there is no easy access to Ox output (apart paper clip into the connector), VF output is also missing..

[Crucible]

Has it improved since replacing the o2? Has the code gone?

You may well have an issue elsewhere but it would need to be addressed before going any further. If its still an intermittent no start I would be checking the usual when faulting. fuel press, inj pulse, GOOD spark across a decent gap, vacuum, exhaust restriction etc

[inno]

Hi All,

Unfortunately, the 02 sensor did not fix it. Still no 02 error code, but still the same problem. Cold starts always fine, warm starts often won't do anything.

I tried the trick of disconnecting the coolant temp sensor, and interestingly, it starts almost immediately then. Connecting the lead back causes the engine to significantly increase idle speed (most of the time). When I turn off the car at that point, I can't start it until I disconnect the temp sensor again.

Note that there are no temp sensor fault codes until I had it disconnected on puprose.

So I went out and got a replacement one from the wreckers. The sensor measures around 1.7kohm at room temp and 700ohms around 80/90C. Which I think is ok. The sensor out of the car was fine at cold temp, and seemed ok at 80/90C but I got the idea that the readings were some times off, way lower then expected. So maybe the sensor is unreliable at higher temps.

Also, just to note, when the car does not start, it smells terrible of unburnt fuel and when it finally starts there is a puff of smoke coming out of the exhaust for a couple of seconds. So seems like it's way to rich during that non-starting-time.

Anyhoe.. will have to drive around for a while again.. at least I have a "I'll-make-you-start-now" method if it happens

As a side note, does anyone have a working OBDII or similar reader that works with a caldina? I have not been able to find anything that works..

Cheers
J

[RS13]

Unplugging can generally make them harder to start if anything.

I tried the trick of disconnecting the coolant temp sensor, and interestingly, it starts almost immediately then.

Heh.

Anyway, so you've replaced the temp sensor with a replacement and its' still hard to start? It still sounds like its' overfuelling when warm which would explain the cloud of black smoke once it does get running, as was mentioned you won't get a code unless the sensor has an open circuit. Does the check engine light come on with it disconnected, or is it a passive code? Re the reader.. you might just have to lube up and go see Toyota or a garage capable of reading it properly.

[inno]

Anyway, so you've replaced the temp sensor with a replacement and its' still hard to start?

I have not used the car long enough yet for the it to properly heat up. The problem does not always happen, so will give it a few days. I'll report back though.

Does the check engine light come on with it disconnected, or is it a passive code?

Not sure what you mean with passive code, the engine light does not come on until connecting the 2 pins on the diagnostic connector (forgot the pin names at the moment).It flashes "22", which means "Temp sensor" afaik. It only happens when I disconnect the sensor.

[RS13]

Sweet as, if you were getting odd readings from the temp sensor you removed then that may well have been the problem.. let us know how you get on.

[inno]

A quick update.. so far call still running fine. Have not even had a hint of the problem anymore so at the moment I'm considering it as "fixed"

Thnx for the help all, was really helpfull..

J

[RS13]

Just to add to this, I came across this exact same issue in a workmates' 3SFE Caldina a couple of days ago.. nil idle, rough running, hard to start and no codes. Unplugged the coolant temp sensor, fired up right away and ran smoothly with the CEL on.. replaced the coolant temp sensor with a good one, problem gone. It seems to be quite a common fault with these engines (I've read of maybe half a dozen others on forums, 3SFE and GE as they have the same part number sensor apparently) so definitely something worth checking out if you come across similar issues.