Club G40 Forum
Technical => Engine and Transmission => Topic started by: supercharged spaniel on November 20, 2008, 08:58:12 am
-
following on from the previous thread...'rough start up'
me and dan knight have had the same dodgy miss fire on start up. ive had it for about 18 month now and dan knight around a year. we have been comunicating for weeks in a bid to solve it and i have prviously changed so much stuff in a bid to find out what it was, including...dizzy cap, rotor arm,lambda, plugs, fuel filter, timing checked manifold off, isv cleaned, throttle body cleaned, the list rerally does go on....
anyways i had previously changed the blue temp sensor but it would appear that this is the problem and we have duff sensors? checking the resistance whilst on the phone to dk last night before starting the car up after work (been left outside cold for approx 9 hours) the resistance was intermittant and would not hold a reading...start the car and its rough as..
only way i could think of checking mine was to do the following:
in the morning when it would do it every time, take the blue temp sensor off, place in a cup of boiling water leave for 2 mins and refit. if the miss is gone then the sensor is obviously not reading right or it at least shows it only gives a reading when warm. The problem only lasts for about 30 secs to a minute so then car runs fine. so its obviously when it gets warm it starts giving a reading.
anyway..refit it and boom! starts no problems! cant believe its something so simple. fingers crossed when i get a new sensor it actually play along and fix it...but so far its a break through!
-
you seen the guide for testing it in the FAQ? resistances at certain temps.
-
yep..but dan just text saying with another new sensor in...his still did it this morning? what else controls this sensor? what would influence the sensor? surely they cant all be duff ha ha
-
its got to be along the lines of this sensor because why would the car run fine with it warmed up and refitted?
-
well it is 100% the blue temp sender that is causing the problem.
But now the knock on problem is, the sensor is getting a wrong reading on start up, its as if the car is too cold from start?
if i fire car up with cold sensor, it idles like shit.
turn it off.
stick a different sensor in a cup of hot water, fire up, runs fine.
so what is causing the sensor to get too low a reading? or high, however it works...
im going to check what reading im getting off a cold sensor, any ideas?
-
the readings are in the faq section i think.
It could be overfueling for a number of reasons i suppose... the hot blue temp just cuts the fuel (thinking its hot) and the problem appears cured.
-
its a good point pete. so what else would make it over fuel? what are these 'number' of things? ???
-
i feel we are close to a solution but need more input as im lost for what to check next..
-
a number of things.... er
fuell pressure regulator
injectors
chip
lambda (although not till it gets warm as doesn't work till hot)
timing
worn valves/seats (blow through)
off top of me head
-
fuell pressure regulator - tested it all fine
injectors - tested all fine
chip - tested all fine
lambda (although not till it gets warm as doesn't work till hot) - tested all fine
timing - timing could be wrong but if the timing is out it only exagerates the existing problem
worn valves/seats (blow through) - mine are new
im sure its the blue temp sender still
-
well mine works fine so you can swap with mine if you like mate....
as for fpr - you could do with a pressure gauge hooked up
injectors - mine are refurbed so could try them.
valves - well check compression to make sure that's still tippety top. Andy strange's head wore after 1500 miles or something similar
-
another option is a dodgy wiring loom. checking that involves a long session with a multimeter checking the pins on various sensor connectors back to the ecu connector, there is at least one thread I've written in the FAQ sections about doing this. Plus checking the earth crimp inside the loom. Although to me, it doesn't make much sense that it would only happen when cold and everytime it's cold. Unless it was a break in a wire which fixed when some heat went into it to soften the wire insulation, it seems pretty unlikely though!
Has either of you had a VAGCOM on it yet to watch sensor values when the problem happens?
-
hmmm..but heat getting into the wire is the wire and therfore it wouldnt matter what temperature the sensor is if you know what i mean? but with a warm sensor and starting the car from cold there is no issue?
-
is it possible that the plug end is at fault? not on the sensor but on the loom? would it be easy enough to resoulder a new one on? if this could not hold a good reading for resistance? im just guessing now?
-
yes but warming the blue temp sensor could just be a red herring. Yes doing that will cause a reduction in amount of fuel the ECU puts in but it could be something else that's causing the ECU to overfuel which you are then compensating for by reducing fuelling manually by warming the blue temp up.
I think the best solution might be to "borrow" a working G40 for a day and keep swapping things over till you find the issue. Or use VAGCOM as I already said to watch for weird sensor values when the problem happens.
-
i plugged mine into a vag com thingy a while back and nothing noticable came up but that was just a fault code reader....what other sensors could make the ecu add more fuel?
-
I don't want to divert this thread [so just go round me and my annoying shit] but I have to ask - What is the blue temp sensor and what does it do?
-
fault code reading on a G40 is useless, you need to look at the measuring blocks section in vagcom and actually watch the readings the ECU is seeing from various sensors when the problem is occurring. It's not going to be a five minute job to do though. Plus there are some oddities to look out for such as the fact that vagcom reports the air temp sensor reading at something like -20deg all the time if it's working correctly!
blue temp sensor is in the thermostat housing below the distributor, it is blue and has a blue connector. It tells the ECU what the coolant temperature is and the ECU adjusts fuelling accordingly. Colder = more fuel enrichment adjustment to the map. The ECU also contains several different fuel maps for different conditions, when it first starts up from cold, the cold temp reading makes it use a "warm-up" map.
-
Look back to my post on 13th on this . I reckon the ECU is holding a memory of the old senders output and needs some sort of zeroing. You can re-boot the whole ECU to some extent so why not one individual component of it ?
-
another option is a dodgy wiring loom. checking that involves a long session with a multimeter checking the pins on various sensor connectors back to the ecu connector, there is at least one thread I've written in the FAQ sections about doing this. Plus checking the earth crimp inside the loom. Although to me, it doesn't make much sense that it would only happen when cold and everytime it's cold. Unless it was a break in a wire which fixed when some heat went into it to soften the wire insulation, it seems pretty unlikely though!
Has either of you had a VAGCOM on it yet to watch sensor values when the problem happens?
if polo g40`s have vag com were do you plug in to?
-
they have a diagnostic connector yeah. It's two, two pin connectors clipped behind the dash where the indicator relay is. Under the glovebox. You'll need an ODB to 2+2 adaptor (available on fleabay)
-
Look back to my post on 13th on this . I reckon the ECU is holding a memory of the old senders output and needs some sort of zeroing. You can re-boot the whole ECU to some extent so why not one individual component of it ?
g40s ecu isn't that technologically advanced... if the sensor is faulty it will log it and not use it. if you replace the said sensor, it'll start working properly again whether you clear the code or not
-
yeah it doesn't really store codes like more modern ECUs do. You just need to fix the fault and then after about 10 mins of driving the ECU realises that signals are back within tolerances. So you don't need to clear fault codes like on more modern stuff.
-
well i examined my connector last night and the plastic wire outer crumbled and snapped, and i sheered the plug connector off
so i botched some wires on and hooked a multimeter on this morning and it wouldnt start, tried again same
so i unplugged the multimeter and it fired, ran rough for about 10seconds, then it was fine...
i feel a bit stevie wonder with this tho, im still sure the problem lies with the blue temp sender, i think it maybe the connector, which spaniel thinks the same, im going to get a replacement one today and see how it fairs.
-
its got to be that plug/loom/sensor...if the wire/plug is corroded even slightly it will alter resistance ever so slightly which could be giving duff readings when very cold..
-
fingers crossed for you2.
I have about a billion blue temp senders i can test and give you to try if you like.
Pete
-
which connector are you talking about where the insulation crumbled off? the ecu connector or blue temp sensor?
-
the blue plug connecting to the sender
-
well i swapped my plug for a new one, got on machine and checked the emissions and it was running very rich
so tweaked the co2 pot acordingly and it was running nicely, moved timing to suit aswell, my mechanic didnt want to do the timing from the crank or the cam as he said it wouldnt be accurate.
and we couldnt get in on the spy hole on the gearbox
i mentioned about the co2 pot setting that it needs to be, and he just said, thats a base figure and with all the extra bits changed etc it needs adjusting accordingly
so he did and im running it now, and it feels nicer, gota wait till morning to see how it starts :D
-
i mentioned about the co2 pot setting that it needs to be, and he just said, thats a base figure and with all the extra bits changed etc it needs adjusting accordingly
yeah absolutely, I keep saying this too. Only proper way to set it is on a gas analyser. Fingers crossed for tomorrow morning then. ;D
-
gas analyser is the answer fo shooooo
-
ok fingers crossed for sure....how do you set the co pot to match the gas analyser? is it simply a case of screwing it until the emissions lower or something? mine too is set at around 550 ohms and i noticed the problem after the decat was fitted so it hasnt been adjusted to accomodate modifications at all. Im happy with the timing on mine but like mentioned the co pot is just at its recommended base setting...
god i hope its this!!!
-
i have the mutimeter at the ready... :-X
-
you need to take it to a garage with a gas analyser and they'll know what it should be set to.
-
im just gonna play with it and see what happens when i adjust it. then get it set up properly
-
adjusted it to various readings and kept trying it but it made no difference whatsoever to the behaviour of the car..the blue emperature sensor being warmed is the only way to stop it doing it..if you plug a spare in which has been warmed the stuttering/missfire stops.
-
Car's not overfueling/misfiring. When thermostat opens , ECU is driving revs down and knock sensor is rescuing a stall by backing off the timing. Regular"chug" effect. I've got nothing to prove here, only telling you what helped me although as stated previously I'm not exactly sure why. Fire up from cold and before thermostat opens disconnect blue sender for abot 30 secs then re-connect. Seems to make the sender compatible with the ECU. What you got to lose?
-
you're wasting your time trying to do it by ear Dan, you need to get it on a gas analyser to monitor the CO levels, seriously, there's no way of properly DIY'ing it. But 550-ish ohms usually works. Setting it to that shouldn't be causing any problems.
-
changing the CO will not be fixing the problem, more masking the consequences.
Gas analyser is needed. Most places will do it for a drink
-
you're wasting your time trying to do it by ear Dan, you need to get it on a gas analyser to monitor the CO levels, seriously, there's no way of properly DIY'ing it. But 550-ish ohms usually works. Setting it to that shouldn't be causing any problems.
yes but adjusting it to various different readings should give some indication of improvements which it didnt..the idea was to see if it would improve and then get it on an analyzer and then set it up exactly..i dont have access to an analyser right at this very moment...
anyway it made no difference whatsoever what ever setting i put it at really..not noticably so i set it back to 600 Ohms or 550 i cant remember now but around the level it was originally.
its defo not got anything to do with the co pot..
i tried it low around 200, 250, 400, 500, 600, 700, 800, 850, 900, 1000, 1050, 1100, 1150 and so on upto around 1800. you would expect some change at some point if it was this sensor...it didnt and neither did dk's as far as i know which was set up with an analyser.
-
Car's not overfueling/misfiring. When thermostat opens , ECU is driving revs down and knock sensor is rescuing a stall by backing off the timing. Regular"chug" effect. I've got nothing to prove here, only telling you what helped me although as stated previously I'm not exactly sure why. Fire up from cold and before thermostat opens disconnect blue sender for abot 30 secs then re-connect. Seems to make the sender compatible with the ECU. What you got to lose?
hmmm i dont really understand what this will do? even if you take another sensor in your hand that has been cooled in water and plug that in to the loom it starts to miss fire even if you have driven the car all day long so was well warmed..but what it should really do is just raise the revs as it would on cold start. Unplugging the sensor as its missfiring/overfuelling completly though does nothing it just continues to do it.. im starting to loose heart in this now and think stuff it!
-
i'd start checking the loom now!!!
-
ive checked mate visualy and i cant see owt on first looking..ive looked right along the loom of the blue sensor and nothing apparent is visible...i think i need to leave it for now and come back to it another time when i have a little faith...and inspiration of course ha ha. how would you go about checking the loom exactly properly with a multimeter?
-
You need to get a multimeter in ohm mode and check the pins from the sensor to the corresponding pins at the ecu connnector. Resistance should be very little (but not zero), if it's noticably higher between then there maybe a break in the wire. Some multimeters have a continuity test function with a beep or buzz.
Did you check that earth point that people have mentioned ?
-
yep, visual check not good enough. Check out the ECU connector pin out thingy in the FAQ section.
-
I have to admit to being a right twat on this; I've been looking at it the wrong way round. My rough running came in when the thermostat opened and the blue sender heated up. The ECU was driving the revs down too far. DK and Spaniels is the reverse; the hot sender gives smooth running, lifting the engine out of an unstable mode.That being the case I'd say the sender and loom are ok as the ECU's getting a signal through. The sender wouldn't appear to be that critical,not sending a variable message to the ECU, just a basic on/off action. Someone else has posed the question, what else governs start up, what have we got ?
-
right i will do cheers lads....gonna leave it for now im sick of the sight of it ha ha..havent checked this earth thing you speak of..is this the one on the near side chassis leg hidden in the thick of the loom? defo next to do is to check the resistance of the wire on the ecu loom plug back to this sensor. thank yoooo
-
I have to admit to being a right twat on this; I've been looking at it the wrong way round. My rough running came in when the thermostat opened and the blue sender heated up. The ECU was driving the revs down too far. DK and Spaniels is the reverse; the hot sender gives smooth running, lifting the engine out of an unstable mode.That being the case I'd say the sender and loom are ok as the ECU's getting a signal through. The sender wouldn't appear to be that critical,not sending a variable message to the ECU, just a basic on/off action. Someone else has posed the question, what else governs start up, what have we got ?
well ive thought about this too...i spoke to a vw specialist and the bloke said 'well your sensor is obviously sending a signal and so are the wires because the engine or ecu is reponding to the info given by it' but can the wires just be faulty? or like you say...is there something else controlling start up?
either way ill check the wires but i need a little info to get started as its not something i have experience of doing...
say i take the big connector off the ecu and start checking the loom side of the plug for resistance? looking at the faq thing it says: ignition on pin 6? (for the coolant temp sensor remember) but when ignition on at 20deg and 80 i think it says pin 10? why does it change? hows it not always pin 10? another thing..we should be testing ohms here? but the table says it should be 0 volts? and 1.5 volts etc? hows that? it could be plain obvious to someone whos clued up on this side of things but im delving into the unknown here..some help is greatly appreciated.
also (quote) 'Make sure the earth pins on the ECU harness have almost zero resistance to the negative on the battery' . a little more detail here would be great...earth pins are 13 and 19 so what do you do exactly with the multi meter?
'Then just go round each sensor that connects to the ECU making sure there is almost zero resistance between the sensor pin and the pin in the ECU connector' so is this a case of one wire (of the multimeter) on the loom pin and one on the ecu and checking resistance?
-
Took the bull by the horns and fired up without blue sender connected. Engine ran at smooth fast idle(1100rpm) till thermostat opened. That made no differance, still 1100 adjustable with idle screw up or down.Plugged in sender and normal service was resumed ok. Does the idle screw setting on it's own provide a base line that the ECU modifys courtesy of the blue sender to suit engine temp.
-
fired up without blue temp sender in???
it shouldnt start?
-
yeah it'll start without the blue temp sender in it'll just be in limp-home mode though and will run shit.
The idle screw just lets a certain amount of air past the throttle plate to allow the engine to idle, more air = higher idle. The blue temp sensor is used to adjust fuelling according to engine temp, colder engine = richer mixture.
-
well today im gonna start by stripping the loom mon the chassis leg and check these earths....see what we got.
-
right well today i stripped the loom down and checked the earth crimp and found nothing...then i started stripping it around the lambda/starter/knock sensor plugs and found a horrendous bodge on the starter plug??? some gimp had cut the plug (loom side) on the live red wire and soldered two single core wires into the gap? looked awful!!!! the two wires were also stripped half way along and twisted togther? obvioulsy didnt have a thick enough section of wire so made do with 2?
anyway cut i out and soldered the loom section back to the plug as they should have done! re wrapped and all good. going to start checking the voltage on the loom for the blue sender in an hour or so...
ordered some genuine vag loom tape to re wrap the loom tape ive stripped too which is around 4 quid a roll and arrives wednesday so hopefully i cn find the problem before then.
-
just a quick note....i will NOT be defeated!!!
-
well i fired mine up to show pinches the problem and it did it rough for about 20seconds and that was all, so i have altered it, but not cured it...
ive lost intrest in it so im going to see what else i can think of...
-
Solve this please haha I dont have the problem but I read this thread like 3 or 4 times a day to see what the cure is to this frustrating problem for you guys.
-
ha ha...well ive not checked the values on the loom as yet but will do this week.
I did notice this morning which im not sure ive mentioned but is worth a good mention, is that the car does not idle high as it should on cold start? it starts and idles at the normal speed it would when warm? so around 950rpm not 1100 ish that it should? obviously miss firing at the same time. it only idled properly as a cold start when i put a warmed temp sensor on the plug, the other day...does this indicate anything or spark any new ideas?
what about throttle body?
isv? these are two things i havent changed but only cleaned..
without checking the loom with a multi meter i havent found any apparent breaks or problems and the earth crimp was perfectly clean under the wrap...cleaned the earth straps on the inlet manifold too.
-
looking through the bay again today ive noticed that there is quite a lot of oil vapour build up around the isv...now i cleaned it a while back just before summer and i put new jubillee clips on it on both ends and tightened them on tight and i know my boost pipes are as they should be so im not thinking dodgy charger seals but a possible leak on the case of the isv? either its leaking and when the sensor says cold....open more its got a leak which makes it open but loose pressure??? i dunno im stabbing in the dark here but its electronically controlled so its possible its just not working properly either...dan knight reckons the issue came soon after fitting another 'cleaner' isv. some one explain some good reasons why this could be at fault please! the oild vapour is defo from the isv and not anywhere else everything else is bone dry....
-
oooo ooo i also suffer from a slight hic up on ilde every so often too which im convinced is linked to this ongoing problem...idle control valve...idle miss...see the link? wink wink...ha ha..
-
well as it is freezing outside i thought i would go have a play
my car blue temp sender was reading 7.00 i think
and the 4 spare senders i had in garage, ranged from 4.23, 5.24, 4.27, 4.20
so it is this figure that is effecting it im so sure, its like the sensor is getting too high a reading, causing it to borderline not run...
now how to fix it? not sure, its like the waterhousing needs flushing, or i dunno, but the blue sensor is the problem, for sure...
an answer would be for anyone to check their blue temp sensor from cold before starting...
-
im not too sure cos even if you plug a spare sensor in thats cold in your hand not in the housing it still reads wrong? so cant be adiry housing surely?
-
erm....i think i have fixed it....i think...ill update tomorrow but i am very very optimistic but not holding my breath yet...
-
been 2 days nows spanky the suspense is killing me ::)
-
couldnt agree more hurry up already
-
npe sorry boys..changed the isv..still there. sick now...what next?!