alan_131

I guessed it was something like that - they're great for families, dogs, towing, mobile shed. You'll easily do double the s4's mpg, on my old mk1 one I was getting over 50 over course of a week away, 4 up plus dog and lots of toot. Even towing two tons+ (I mean, exactly within the regs officer) of boat back from Scotland I got close to 50mpg and though there was a lot of motorway also a lot of hills - no problem at all, hardly knew it was there. Apart from the 3hours in Grantham trying to get a trailer tyre retubed - long story). Mk2 better than mk1+ quieter/better safety features - that's why I was looking for one when I got the ghia. Compared to your S4 - very,very boring.... Agree about the fixer-upper - so's mine - impulse/accident on ebay, and not quite as described. Came with 11months mot & limp mode, not sure what the next test will bring. Wanted one to try the oven-cleaner fix on, but it came with so many other interesting features I haven't got round to that yet. Nice to have the aircon when the sun shines though. I think the voltages should have some similarilty on the kline, even though the connections/influence of other devices may be very different. I'm going to do this on my spare car when I'm doing some other fiddling about, it lives in a container where I keep my boat so probably not till tomorrow at the earliest. All this stuff is useful knowledge for one day/others so not wasted even if no help this time. Keep us posted...

When I can, I'll try to pull the ecu and track the vcds init voltages as you have. As I've got a mk1, may not be entirely valid. I don't know how easy it would be for you to follow the init when talking to say the immobiliser? At this point I'm getting a bit worried that the best response to your original query might turn out to have been "yes - get a new ecu". Still hopeful that not so, and the nice thing about here is you never know when somebody may turn up with the answer...

Am I right thinking you have the clicking from the ecu all the time you're driving with the instrument cluster connected? Do the panel lights go on and off to a regular pattern? With your turbo problem, is there any change in what's happening on the instrument panel? Also, on the flat even with wide open throttle it's still fine, it's only on a hill this "power on, power off, power on" happens. If you run the engine, then stall it (horrible I know) does the flickering stop & things go back to as if you'd just turned the ignition on? If you turn the ignition on, then shake the engine, does anything happen at the panel? If you connect to the engine with vcds, then start, does vcds stay connected / let you read measuring blocks (ditto instrument cluster) I'm wondering if something is re-initialising the instrument panel, either a power/earth fault or the ecu. My favourite would be, as biturbo.s4 says, power/earth. Not sure how much the haynes manual for pre-98 is to be relied on for wiring diagrams, and that doesn't help. Do you still have easy access to the fuel sender connector?

All good fun... Not come across anything quite like this before myself, but I bet someone has. your profile says 1996 tdi, might be worth expanding on this a bit for this problem. Engine power, manual/auto, trim level..any vcds codes....

Progress if not neccessarily what you were hoping for... With the ecu disconnected, do you still get the same changes in continuity from obd pin 7? Also wondered what voltage you see on that pin now? Would be nice to be able to connect directly to ecu pin 16 without any other kline connected - suppose if nothing was showing on the line when it was disconnected from the immob and ecu that would be a clue....

Sorry I was being incredibly thick there, thought you were referring to the supply to the cluster. doh.. I see now why you were expecting the supply voltage to be more-or-less constant. I'm guessing from your last post that for a while the problem went away (long enough for you to drive it till the gauge dropped below half). Is that right, and was it immediately after you'd been looking at something (he said hopefully). Is it now back solidly? If it is,does it happen as soon as you crank, or as soon as it runs (appreciate might not be much time between the two). Is the ecu still clicking with the instruments back?

Sorry missed your post re voltage/panel lights. Can I ask where exactly you were measuring the voltages (between where live and where earth). Assuming this is an unregulated feed to the panel and I'd say from the voltages you're reading that and my haynes wiring diagram it is, it's normal for the voltages to vary between engine running and not, though your not running 11.06v sounds rather low. Is this the same as battery voltage measured at the battery? I need to change some of the bulbs in mine (gave some to someone the other day and not replaced them yet) but what I have left only light up when I turn the lights on and I think this is a clue. However just to complicate things I found this old post http://www.fordgalaxy.org.uk/ford/index.php?/topic/15010-dash-lights/ which says different. Edit : ah - see I was wrong - thank you seatkid. Do you have door open/bulb failure on your car? Edit: think you don't. Must read more carefully.

Funny - after digging around this morning I'd come to the conclusion it wasn't canbus and was about to post humble pie, based on the supported i/f info on ross-tech and some other sites. Yet clearly it is. If it was me I'd be connecting directly too. Might also look for voltage at the kline odb pin (to see if there's anything there with those other controllers disconnected if they still are) and at the corresponding ecu pin before rigging a connection. Don't have proper pinout for your ecu - is pin 16 kline - I seem to recall it is, but my memory is rubbish... Good if the diags are available on the canbus - nice and speedy i would think... Do you have autobox?

Thanks for that - think I'll go for the elm327 - sounds worthwhile addition/flash. I have a mk1, which I didn't think had canbus. The haynes mk1 manual shows the mk1 diesel as having odbc canbus pins 6 and 14 connected to engine ecu - which they aren't - they're unpopulated. My interface is a cheapie klline only one and works fine. All the unmatched pins on my i/f are unconnected apart from 15. Which interface do you have - does it support canbus? Am I right you're using a genuine vcds/um "evaluation" clone with the vagcom version designed for a ross-tech i/f? fwiw looking today the current vcds is 11.11.3 and I think there have been a few driver issues in vagcom's past, some vehicle specific. Which extra connections do you have on your odbc port? Wondering if your interface may be physically connected to more than one data pin, and if you have extra connections it doesn't really understand this may be related to the problem. I expect you've been here http://download.ross-tech.com/VCDS/download/VCDSCompatibilityCharts.pdf but in any case I have a feeling (others may know better!) that canbus was in there somewhere on mk2.

This is the topic I meant. ... If you look towards the end you'll see that Chrispb123456 put his finger on the problem first suggestion and biturbo.s4 has explored and posted successful outcome. It fits your problem symptoms quite closely (apart from the fuel gauge's role) but I think it's definately worth a look. Does this sound like it could be what's happening with your cluster - when the gauge is high, resistance is high, the current drawn by it from the cluster is lower and the cluster has enough to work properly. When it gets low, it pulls enough current to drop the cluster voltage because the supply to cluster is restricted by an iffy connection. Same applies sort of if earth is duff and current flows are misbehaving. Anyway, lots of waffle but see what you think.

Remember sleep? (3.31 am) can't be sure but I think candidates are Instrument cluster, abs, and if you have them autobox & climatronic. Sure this sort of thing must be documented somewhere- - vag issue briefings for new engine types etc, you'd think there'd be one for diagnostics but it's the old "minimum information" thing. You'd assume they'd be on the wiring diagram but I'm sure I've read somewhere of undocumented links. Wondered why you use 10.3 (apart from the obvious "I proved it works on my laptop") and also what you used with your elm327 - I was thinking about getting one + torque probably just for otr issues.

You might want to start your own thread, but fwiw I think the click from the starter shows the basic control circuitry works, so you're probably looking at the old-fashioned causes. Bad battery, dud starter, dud connections/poor earth/feed. If hitting the starter with a hammer sometimes works then I'd be thinking starter or its connections first, but your other comments point at the battery. If it were me, I'd make want to make sure the battery was good and then move to the starter. Maybe take it out and give it a clean, though I'd suspect it might need more than this. Might also be worth trying (after its done its slow cranking thing) just feeling around connections on battery, starter, engine earth, to see if they're hot/warm. (engine off and beware they might be very hot.) Few questions - Is it just a single click for each time you turn the key? Is the click definately coming from the starter motor? If you turn on headlights, do lights go dim when key is turned? Any other odd happenings/funny noises? Just after it starts, if you have headlights&heated rear window on,does revving the engine make the headlights lights brighter? Good luck

Just a thought but if the system is overpressure on the low side - assume it's the same both sides as your compressor not run for a while - could it be that at some point in the past (maybe when garage sent it back without ac working) they'd overfilled it without telling you? Given it the dual evap charge perhaps? Now in the very hot weather the system reads overpressure and won't start (I'd assume that the switch is simple pressure and not temp. compensated) If it was this, then might be worth trying at the coolest part of the day. I've also read that it's not hard to inadvertently overfill with the refill aerosol, but don't know if this is right. I think the guage on these cans attaches to the low pressure port , is that right? Do you know what the tri-state pressure switch is putting out? Are you getting anything from vcds? I used kwik fit to refill mine (because it was empty and I assumed a leak) and they were quite helpful - vac'd it for 20 mins, filled it correctly and it works a month later. Always a bit reluctant to use them (scratched paint from an airline dragged over car with no rubber buffers, couldn't start my automatic (because it was in "drive") , nearly dropped a car I'd only had a day off the four-poster when "the jack slipped out" leaving the car slewed across the lift resting on the front chassis member) but watch them very carefully now and don't let them do anything critical apart from tyres...

You've probably seen it already but if not I wonder if biturbo.s4's result might apply to your car too?

Result on the instruments etc, Chrispb123456 patiently sorted the first problem I posted (re aux heater controller reset) and really helped me too. Has this made any difference to the voltage you see on pin 7 now?

Good news about the MAF One down.... Haynes shows the mk1 instrument cluster functional earth on pin1, but interestingly has the earth for the fuel sensor (and coolant level and temp sensor) on the same earth point (E7). it says this is lh engine bulkhead(pre-99) and lh rear engine (post 99). May be completely irrelevant , but fwiw...

File not found... should have that tatooed on my head.... I suspect you probably know a lot more about this than me, but the ac pressure switch flags over as well as under pressure i think - but would have thought that should have been a "flash the panel" error too.. Not noticed the lights dimming on my mk1 when I turn auto on - is this normal? Been a couple of topics last few days re low-temp lockout Agree about the control alogrithm - this sort of information seems a) vital and B) unavailable....

Gray/white stripe as per your post is generally vw diagnostics. Your instrument cluster has a controller, autobox if you have one, aux. heater, etc. I don't know what functionality that used to be separate has migrated to the main ecu - maybe none, but immob? c/l? Presumably (I haven't seen one) the mk2 manual has its usual wiring diagrams - ie not as good as manufacturer but a good starting point. Mirez has mk2 fuse info on his site http://www.mirez.co.uk/Galaxy/f&rinfo.html Do you have anything else that doesn't work? Realise how frustrating this must be for you. It really bugs me that manufacturer's don't supply full information when you buy a car.

Duff MAF certainly stuffs things up and it probably is the MAF but being extremely stingy/impoverished if it was me now you've got a working lead I'd clear the codes and then see if they come back, maybe try logging maf measuring blocks and check the wiring first. I think any time it's been run with the cable off (for the old "is it better unplugged" test maybe ) - would cause something like this code to be logged.

The Mk1 haynes manual shows a "typical" air-con having thermo-switch connected between pressure switch and ecu. For anybody with climatronic reading this topic later On my (late 99 built climatronic) the sensor provides info to the display. At 8 this morning display showed 25C, my IR temp gun agreed (nice day). Sensor resistance was about 2.5Kohm, socket had 5v across it. I've only fumbled around the receiver/drier and don't know if my mk1 also has the same cutout as above, (but for a protective device simple go/nogo has a lot going for it so it) or if it relies on the bumper sensor.

I don't think your ecu can be dead if the car runs, because the PD is electronically timed unlike the mk1s which are more mechanical with electronically adjustment. Not to say that the diagnostic interface isn't stuffed though. As I understand it, the diagnostics from the different controllers are multiplexed somehow, maybe by each controller monitoring all traffic and sharing a common cable or a single (the ecu?) having connections from each controller and doing the multiplexing - don't know which applies here. You can see why this could stuff things up when either wiring or dud device is doing things to the outputs. I think you might find following pin 16 back to discover why it has no power might help (easy to say). Presumably the new haynes mk2 will have wiring diagrams or maybe TIS or somewhere on here perhaps? (Powering 16 (battery +) from the fusebox will give your interface power, but does this normally come direct from the fusebox or is it via the ecu/some other stepping point that might be related to/help diagnose the problem?) On my mk1, the diag socket has only 4 connections populated - 4 (chassis GND), 5 (signal GND), 7 (K-line) and 16 (Batt +). 16 > 4 shows approx batt voltage at all times 7 > 5 shows about 10.5v ignition off and 11.5 ignition on 4 > 5 shows resistance of about 3 ohms BUT all of these were measured using makeshift probes into the socket and this could be due simply to poor contact/surface coatings on the probes. Experimenting today I found that, with the cable connected to the car and laptop, it would pass the (vcds-lite 1.0,Vista) connectivity test on com3 but could not access any controllers. (been messing about with virtual com ports for other reasons). It showed a very poor latency (> 26ms), this might have been the reason, maybe not.... Changing to com4 (around 11ms latency, also I clicked apply this time before saving) worked fine. Another oddity to think about....

You're right, there is a temp sensor which locks out the a/c if outside temp falls too low, as you'e worked out. It lives somewhere (haynes manual says clipped to) the front lhs bumper. On mine the sensor is dangling below the bumper - I suspect the bumper was swapped before the car was sold - but in any case I think the manual is probably wrong and it clips near the bumper. SilverBeast was kind enough to post some pics in this topic http://www.fordgalax...r-conditioning/ showing where it lives on a mk2 - I haven't had time to explore yet.

Thank you for the kind words but I'm largely just parroting good advice given by others in the past. I'd be thinking just pull the cluster and reseat the connectors if it was just the dash, and also check any visible earth points. Given your wider ranging issues I'd probably still do that just to get it out of the way... I've seen posts where people have had strange happenings in the past due to blown fuses, dud relays, shorts - tailgate wiring for example, and so on. Is the radio an aftermarket unit (maybe not if it works in the Lupo?) This topic on another site might be interesting http://forums.vwvort...d-some-guidance Proving the cable works in your Audi is a big step forward, and the change with the abs controller disconnected is interesting (see the topic on vwvortex above). Depends a bit on if your cable just links all the lines together then looks at them, or has some smarter arrangement. Never simple! I think others will learn from your work... If you can't find a cause for everything, then might be worth working backwards from a known fault. Seem to recall somebody having problems due to the loom near the obd socket itself. There's many others on here far more knowledgeable than me, and you'll probably find somebody makes a better-informed suggestion very soon. Good luck. Sorry - started reply a while ago so a bit out of synch with your posts

Only guessing but do you have to put your foot on the brake pedal normally before the button will go in? The brake switch (about £10 ish if you need a new one may need resetting (easy when you know how). Quick search thru the forum will point you in the right direction. try under cruise control if ou can't find it directly. Good luck

Sorry that didn't help. I'm wondering if your reader is really pointing in the right direction. I know it's a bit of a pain, but I'd really trust the output that vcds gives over a standalone reader, just because the software is specific to vw and used enormous numbers of people. And being software easily updated if a problem is spotted. . .