musso2010

What colour should a working Lambda Sensor tip be?

I removed mine today, and it is almost white, with only a slight bit of sot on it. Is that normal or is it knackared?

Also, if the lambda was in good working order, would disconnecting it cause the car to run rough?

Splitpin

White with a slight bit of soot on it doesn't sound at all unusual or bad. Why are you asking - because you suspect the sensor has failed? If so, it's not normally possible to tell visually. If you'd like a second opinion on the way it looks though you can always post a picture of it.

And no if the lambda's in good working order and you disconnect it, the car would be expected to drive more or less normally. It won't perform any closed loop fuel correction but will still use any learned corrections it has built up - assuming of course you haven't just reset the ECU. Also bear in mind that the O2 sensor only affects the fuelling in low load cruise modes - so it'll make no difference under hard acceleration, one way or the other.

musso2010

Thanks splitpin. Im having some problems under hard acceleration and the car holding back. I was told to check the lambda sensor, and disconnect it and see if it changed anything. The car ran totally the same with it disconnected.

Also, when i disconnect the battery to reset the ecu, the car idles really low as if it is about to stall, and then after a few days it idles fine (almost as if its learnt to ajust itself).

Ive recently replaced the MAF sensor so i can rule that out i guess. Any ideas?

Splitpin

Who told you that? As above closed loop fuel control is switched off when running above a certain load threshold, so the chances of the lambda sensor (whether broken or otherwise) affecting engine performance under hard acceleration is pretty much non-existent.

Given the above, not surprising!

That's not the ideal way to perform an ECU reset. What happens if you do it the "proper" way with the black and green plugs, and then driving the car slowly? Do you get an all clear, or have you not tried that method?

When did you first notice this idle instability problem? Did you reset the ECU after fitting the replacement MAF sensor and notice it then, or, do you remember it happening from some time before you started to deal with this current problem? This could be something like a badly calibrated throttle position sensor or idle switch. However, if the MAF unit you've fitted isn't a fresh out of the packet new one, one thing to rule out would be the possibility that you've swapped one faulty unit for another.

This new MAF unit - is it a brand new, Subaru Genuine part, or a secondhand/non-genuine one?

Also we'll need to know what type and model year car you have, plus any relevant mods. Are you running the standard airbox/filter, or an aftermarket cone filter (etc)? Have you also checked the condition and gapping of the plugs?

musso2010

--I was told this by someone who thinks they 'know' about cars, which obviously they dont.





--I have never tried to do the ecu reset with the plugs, however, i will give it a try. I was always under the impression that leaving the battery off was another way.



--When i first bought the car i was having the same hesitation problems i am having now. I checked the fault codes, which were CTS and MAP sensor (these codes were stored before i had the car, there is recipts for the replacement parts). I left the battery off over night and it cleared the fault codes. But it was then that i noticed the idle problem, which sorts itself out after a few days. And it only goes ****ty if i disconnect the battery.



--Second hand MAF from a car i know to be working spot on... however is may be a possibility that the MAF may be faulty.


--Its an MY95, UK with a Polar performance chip, TD05 turbo, Full decat, 3 port boost solenoid, K&N, dump valve, hks super dragger exhaust and a FMIC.

I was advised to remove the FMIC, and fit the top mount again. I am waiting for the top mount to arrive. I will also run the std dump valve and see if that makes any difference.

Glowplug

iTrader: (8)

Exellent time for me to verfiy something.

When you do the ECU reset in the 'proper' way, with the black and greens, it says to hold the throttle at half way. Does this/is this to calibrate the TPS??

Splitpin

It is another way, yes. However using the plugs is better as it performs an active error check on all of the systems covered by the error detection system and will only complete and give the okay light flash sequence if they all pass. It doesn't entirely rule out a failed component but it's more likely to show one up than simply pulling the battery.

It certainly seems to be learning around something, only question is what. As above, try to reset the ECU using the plugs. If it completes successfully, switch the engine off, disconnect the plugs, try the following and tell us what happens.

Firstly start up and run the engine for long enough to confirm that you've got this low idle problem back. Then switch off, disconnect the MAF sensor and restart. The CEL will be on. Drive the car around very slowly (no faster than parking speeds). Is the idle still bad, or is it noticeably better without the airflow meter plugged in?

Either way, switch off and reconnect the airflow meter. You can either reset the ECU again to get rid of the stored #23 error code from the historic memory or not worry about it.

Yes, it's a possibility you need to consider until you can either rule it out or find the cause.

The "K&N", this is a cone filter, yes? Furthermore, is the Polar Performance chip mapped for the FMIC + cone?

Unless the aftermarket dumpvalve is broken, it shouldn't cause trouble under acceleration.

Eh? What "says to hold the throttle at halfway"? Where did you read that? When are you supposed to do this?

And no, wherever you may have read this, the throttle angle calibration is not done during execution of the reset. It is done afterwards, when the ECU returns to normal operation, and is constantly updated thereafter. The ECU keeps a record of the minimum and maximum throttle voltages in its RAM and - if the "current" voltage is below the stored min or max values, then the stored values are updated. So, in practice it's done the first time you let the throttle idle, and the first time you go to WOT after a reset.

willsy369

SIDC web site has this, could be what he means?

Classic (JECS) ECU Reset

Ensure engine is at normal operating temperature, turn engine off
Locate the two ECU connectors, located under the steering column and consist of a black plastic male and female connector, and a green male female connector
With the ignition OFF connect black to black and green to green.
Turn on ignition, do not start the engine, depress the accelerator pedal to full throttle, then return it to the half-throttle position and hold for a few seconds, and then release
Engine Check lamp turns on
Start engine and then drive for at least one minute, keeping road speed above 11 Km/h.
ECU is now re-set.
At this point the check engine light should start to flash the all clear signal (steady 1/2 second interval flashes). If the check engine light does not flash, or indicates some other sequence, there is a fault present in the system, and should be professionally checked for necessary repairs.
Stop the car and turn off the engine.
Disconnect the black and green connectors.
Job complete

Glowplug

iTrader: (8)

Eh?

I have never been able to figure out 'why'. It was just your mention of TPS calibration and 'proper' reset procedure flicked a switch in ma heed.


Yeah, it is on Soobypedia.

Splitpin

Ahhh, yes, that would make sense. If so, the technical term for this half throttle thing is "b*ll*cks". Although it's not the only thing in this summary of the reset sequence that isn't quite accurate:

There is no "need" to ensure that the engine is up to temp prior to performing the reset sequence, and no inherent advantage in doing so. It is perfectly fine to begin it with the engine cold if it suits you. The only difference is that it takes a little longer to complete on a cold engine than a hot one.

The reason for that is because the ECU checks the coolant temperature sensor and O2 sensor for stability - and when the engine's stone cold it needs to wait a little longer for them to stabilise. The reset sequence has a 180 second timeout - so as long as everything checks out within three minutes, that's all that matters.

There is certainly no reason why you'd need/want to do that on the 99-00 era ECUs. I can't think why the earlier ones would require you to do it either. This sounds like another one of those groundless Scooby myths that has come to be treated as fact just because some well-meaning but uninformed individual wrote it on a website somewhere.

As I've already said, the ECU doesn't calibrate the throttle position during the reset sequence, it does this afterwards, in normal execution mode. You also don't need to hold the pedal in any specific way to test the sensor for functionality. Driving normally will do that.

Slightly splitting hairs now but seeing as you quoted it, the order above isn't quite correct either. The ECU only actually completes the reset sequence (inc the bits that erase the RAM) when you turn the ignition off after the all clear flashes come up.

musso2010

Thanks Splitpin
I will perform the ECU reset tommorrow, and let you know the results.

musso2010

Splitpin -

I did the reset today with the plugs, and it came up code 43 - Throttle switch. Am i correct in saying this is the throttle position sensor?

The car has also gone into what i think if safe mode. Wont rev over 4k, boost gauge is boosting but nothing happening.

Ive since replaced the sensor with a new one, but the car is still throwing a code 43 and in safe mode. HELP

Glowplug

iTrader: (8)

This post may be of use to you. Especially as you have a PP chip fitted in a UK car, Scott.Ts' input, particulaly in post #38 may be relivent to you.

I think the 'trottle switch' may be different to TPS. (Splitpin will confirm.).

Yes, it sounds like limp/safe mode, you could try another full reset to see if it clears it. Is it throwing a CEL day to day or just when you fault check?

ps Hows the boost since replacing the FMIC?

musso2010

Hi mate, i havnt fitted the TMIC yet, it still hasnt arrived... (guess the guy doesnt really want to send it considering i got it for 99p lol).

I just randomly did a 'proper' fault check using the plugs, and it came up with 43 and put the car into safe mode! That post from ScottT has just confused me even more now...

Have done 3 resets now and still no change. Do you think that maybe the PP chip has gone **** up???

Glowplug

iTrader: (8)

Bargin!

Yeah i know what you mean. As i said Sp will know more re the switch.

So it's defo a fault then. I have never heard of a PP chip going '**** up' but any thing is possible i suppose.

musso2010

The impreza sure is a learning experience. I just hope its not to expensive, havnt got much cash this side of christmas, especially with the wifes spending habbits.

Glowplug

iTrader: (8)

Well you certainly learnt that quick enough.

All i can find for 'throttle switch' is for later cars.

Splitpin

No he won't, because Splitpin doesn't know enough about the old ECUs and throttle body setups to be able to do so.

Before Musso's most recent posts, I was, ironically (or not) wondering about a misaligned or faulty throttle position sensor as being a potential cause of his problem, or at least certain aspects of it. However, Scott's post in the other thread makes sense - fits in with my own knowledge.

Bat, old buddy, tell me, do you know whether the phase 1 cars have a separate idle switch in addition to the main TPS pot?

Is the check engine light on all the time now, when you're driving normally, or not?

Also, could you try the suggestion I made earlier about unplugging the airflow meter, starting up and checking the idle stability? After doing so, connect the black plugs again (just the black ones) and tell us whether you have code 23 stored in addition to (or instead of) your 43.

Then, repeat the test but this time unplug the throttle position sensor and then start the engine. See if there's any change to the idle. Then reconnect it, do the black plugs, and tell us whether the ECU is now reporting code 31, in addition to the other two. Will hopefully give us a bit more of an understanding of what's going on.

The diagnostic system won't allow the reset sequence to complete if it thinks there's a current error. However, if Scott is right (and he probably is), there's probably another reason for the limp.

It's important for us to know whether the check engine light now stays on all the time when the engine's running, and whether, as above, unplugging the MAF and throttle position sensors will give you the expected errors.

Nothing's impossible, although it's difficult to see a set of circumstances where that could happen. There are other more likely explanations.

As above one possible cause of that idle issue you were reporting after an ECU reset is that the throttle position sensor pot isn't calibrated properly. Whereabouts are you, btw? Potentially quick and easy way of getting to the bottom of this might be to have someone who knows have a quick look at the car.

Difficulty with trying to help you sort something like this via the interwebs is that, as you've seen with Scott's post, the diagnostic system may or may not be giving you reliable information here. As such, and as we can't see/hear/drive/look at the car, the one thing we might otherwise rely on to help you sort it out is possibly less than trustworthy (although we'll know more about that when you've tried the procedures suggested above).

musso2010

Me too. I have searched quite hard for some information but cannot find anything.

The plot has now thickened - Just dropped the misses off out for the night, and i tried to drive through the rev range. I can exceed 4k, but its kangarooing, jolting and backfiring like mad, definatly not something i would like to try again. The boost gauge is showing 16psi... so now im not even sure if i am in safe mode

I cant have a burst pipe cause the car was fine untill i reset the ecu...

Someone pass me a can of petrol and some matches lol

musso2010

Splitpin - There is no check engine light on the dash display during normal driving, which makes this whole thing even more confusing to me

I will try the other tests you mentioned above first thing in the morning.

I am from Wolverhampton, so if anyone is local and fancies having a look, feel free to pm me and we can arrange something.

Just a quick point to note, reseting the ECU with the plugs today gave no dodgy idle issues unlike the times when i just left the battery off.

musso2010

Splitpin -

I have just done the checks you mentioned.

If i pull the maf connection and start the car, i get a rough idle, up and down, like someone is reving the car slightly. The check engine light is constantly on. Reconnect and check black connectors and ive got code 23.

Pulled the throttle position sensor, and i got a smooth constant idle, at 1.5k. Check engine is on. Reconnect and check, i got a code 31.

There is no mention of the code 43 that is haunting me during the reset procedure...

What are your thoughts...???

Splitpin

This adds credence to Scott's post in the other thread. The error you are getting is more than likely a red herring caused by the way the Polar chip allows the use of "non-proper" firmware

One possible reason for that is because, as above, the reset sequence will normally only complete successfully if there are no errors. As such, if it thinks (even mistakenly) you've got 43, or whatever, it won't complete - and therefore your idle control hasn't gone **** up.

Let us know what happens after you try the suggestions above. With luck things will be clearer.

musso2010

Not sure if you missed my post, but ive done the checks you mentioned. The results are above your last post.

Splitpin

Yep, sorry bloke - had hit reply to your first post, gone off to answer the phone and then wrote the reply, during which time you'd updated. Anyway...

Mmmmhmmm. So we know the error reporting system is working there.

Excellent.

Right. Well, first thing I'll say is to repeat what I jokingly said to B13 earlier. I'm slightly out of my comfort zone trying to help you with your phase 1 car as I have almost no direct experience of the older Imprezas and thus I will have to make a few educated assumptions. However, as nobody has (so far) chipped in with any better info, let's go with this and see where it gets you.

Given what you've recently done, I'm now even more confident than earlier that the "43" error you are seeing when you try to execute the reset procedure is, as Scott said, a red herring. However, if the ECU thinks it is real error, it will prevent the reset procedure (when done via the plugs) from executing successfully.

You came onto the thread trying to fix that hesitation you were getting under acceleration, and then along the way detailed that weird behaviour at idle after you forcibly reset the ECU by pulling the battery. As such there are potentially two separate issues here, or it may be that whatever's causing the idle problem is influencing the hesitation under acceleration.

Have a feeling in my water that the weird idle behaviour after the forced ECU reset is rooted in incorrect calibration of the throttle position sensor. If you look closely at it you'll see that it's mounted to the throttle body by two screws through slotted holes, idea being that you can rotate the sensor in relation to the body to give a reading between the margins the ECU is programmed to expect.

In a further attempt to rule this in (or out), reset the ECU by disconnecting the battery, and then, the first time you start it up afterwards, do so with the TPS disconnected. Do you still get the weird low idle as you have come to expect, or is it noticeably different/better?

If it is better like this, do you have a digital multimeter? If you do you can check the voltage as it enters the ECU and tweak the sensor position if necessary.

If the idle is still low/stall-prone immediately after the forced reset with the sensor unplugged, then this likely rules out the TPS. In which case you would want to look at (and potentially clean out) the idle speed control valve. Instructions on how to do this will be elsewhere on the forum.

As far as the hesitation under acceleration is concerned, compromised MAF sensor is still a potential cause you need to rule out. Again if you have access to a DVM you can investigate this yourself.

Other obvious one to look at would be the condition and gapping of the plugs. Worth taking them out and checking both that they're in good nick and gapped to 0.65mm.

Also worth having a look in the regional forum for your neck of the woods and seeing if there's a reputable tuner you could possibly call in on and get a first-hand opinion from.

musso2010

How to i check the Throttle Position Sensor and MAF with a multimeter?

Splitpin

Pull up the front passenger footwell carpet, cover plate and unbolt the ECU so you can get at the back side of the connectors. Then you'll need to check Slipstream_UK's manuals thread, or possibly Jolly Green Monster's website, to get the pinout configuration.

With the throttle position sensor, you will need to double check (again you may be able to get it from the sources above) what the correct idle voltage should be. On the later classics it's 0.48-0.5 volts, but I think it's different on the phase 1 cars. Once you know what to aim for, shove your meter probes into the back of the ECU connector on the sensor ground and TPS signal wires.

Then start the engine, and check whether the TPS voltage is in the expected range. If it isn't, loosen off the two screws holding the sensor to the throttle body, and rotate it until you get the right voltage. Then tighten it back up.

Then reset the ECU again by disconnecting the battery (looks like the "proper" reset procedure is a no-go for you with the "phantom" error) and see if the initial idle behaviour has improved.

If you did end up changing the position of the sensor, try going for a brief drive and seeing if the behaviour under acceleration has altered at all.

With the mass airflow sensor, again, check the pinout, shove your test probes into the back of the sensor ground and the MAF sensor signal pins and make sure they're pressing firmly enough against the pin shells to get a good contact. Take someone with you to hold/watch the meter and go for a drive. You (or your passenger) should see the voltage fluctuating between around 1 and 5 volts. The more air being taken in, the higher the voltage.

If you accelerate hard, you should see the meter reading rising steadily and consistently. If at any point you see it dropping significantly out of the trend while you are accelerating (so, for example, you see a progression like 3.20/3.30/3.35/2.00/3.50/3.60), and you know you didn't lift off the throttle, this is a warning sign.

If you try and accelerate into that "dodgy" 2800+rpm area, watch the airflow voltage closely and again see if it drops out of the rising trend either momentarily or more consistently.

Again this would be an indicator that the airflow meter is potentially duff but it is not a 100% cast iron proof. The reason for this is that if there is something else causing the hesitation, you would naturally expect the amount of air taken in to drop - so a reduction in airflow voltage could be a legitimate side effect of something else being wrong.

Under ideal circumstances you'd datalog the behaviour of a number of the car's sensors in parallel and at high speed - which would allow you to determine whether any downward spike in MAF voltage was the cause of an engine hesitation or the result of one.

As above check the condition of the plugs (and examine the coilpacks closely while you have them out), and, if you can, find someone local who can give the car a look over.

Glowplug

iTrader: (8)

Morning.

I didn't get any 'reply notifications' for some reason.

Musso, are you running the chip in a 7D or V7 as Scott mentions?

If so then you will need to remove the chip and rejoin the J1 and run OE map while you aquire a suitable ECU to run it in.

That however will not cure your running problems. It will only allow you to have a properly operating engine check system. Which, as you can see, causes all sorts of problems when inop.

Once that is done then i would go with me old chum Splitpins exellent prognosis, of realignment of the TPS.

musso2010

Right guys...

I left the battery off for a few hours this morning. Started the car and it idles spot on (unlike the previous times... so sounds like the throttle position sensor has cured that idle problem).

The car is also out of safe mode, and revs freely with boost...

Sooo, am i right in thinking that using the propper reset procedure with the plugs, is causing the ecu to go into safe mode, due to the fact that it is looking for a sensor which isnt there (as per Scott.T's post, number 38 in the link b13bat gave me?)

This would make a lot of sense, and i suppose disconnecting the battery the car is reseting all fault codes, but not activly checking them as it goes unlike using the plugs? Tell me if im barking up the wrong tree here...

b13bat - the PP chip is fitted to a V7 ecu... Is this ecu not compatable with this chip? Or does it just cause problems with the checking system?

Glowplug

iTrader: (8)

It would seem so.

Usually, if in safe mode the CEL would be on continually and not just when you checked. And a wrongly interpreted/set TPS could give simular issues to those that you had. It is most likely that the ECU was still running on the learnt TPS settings and was over compensating in some way.


I think the TPS could of been at fault in some way all the time. As Scott says:


As Splitpin said earlier, the reset will not complete if there is a fault or a no 'all clear' flash. And the completion includes the 'flash' part of the reset.
By disconnecting, you are forcing the RAM to erase and therefor clear any stored faults.


Again, just going by what Scott says in the post.


Doesn't sound like a big problem. But as i said earlier, a inop engine check function can cause all sorts of problems. Though not so much now as you know about it.

Personally, i would get me another ECU, one that will not throw red herrings about.

I've used chips in 3 different ECUs with no issues, the Z4, U8 and the Z5.

musso2010

[quote=b13bat;9733202] It would seem so.



Usually, if in safe mode the CEL would be on continually and not just when you checked. And a wrongly interpreted/set TPS could give simular issues to those that you had. It is most likely that the ECU was still running on the learnt TPS settings and was over compensating in some way.




I think the TPS could of been at fault in some way all the time. As Scott says:

Quote:

Originally Posted by Scott.T

I ran a UK car with initially a Prodrive chip years back and always got that code. It doesn't cause a CEL or any difference in performance

Thanks mate, you and splitpin are an asset to this board
Im glad its revving properly again, its still hesitating but i am sure this is down to the FMIC like you previously said. Im also going to stick a new set of plugs in aswell.