Hi there, I’m new to posting questions, please be gentle.

Our Mk1 Fabia sdi 05 plate has a cold start issue that has become intolerable for the rest of family (and the neighbours!!!)

Symptoms: key in, glow plug light only illuminates for only a second or two, has to crank for 10 seconds or more before firing with much white smoke. Runs beautifully after that, 50+ mpg or runs.

The engine will start easily when warm.

I’ve changed the fuel filter, glow plugs and had the injectors refurbished. I’ve also replaced all of the rubber and plastic piping/O rings in the fuel lines and I’m pretty confident there is no air leak.

After refurbishing the injectors and replacing yesterday, I have exactly the same problem.

I did some digging on this forum and found some posts about the coolant temperature sensor (CTS) /wiring being a possible cause.

So, I disconnected the sensor and now the glow plug light stays on for ~ 5 seconds and it starts immediately.

I connected a new sensor and I have the exact same problem as originally.

I’ve line checked the two wire circuit back to the ECU multi-connector. (Brown/yellow and Blue/red).  I have good continuity/low resistance in both wires. The brown/yellow core appears to be connected to a common ground within the ECU as it shows virtually zero resistance to the chassis earth when the ECU connector is made.

A previous post on a similar topic ended in apparent success when the solution was claimed to be fixing a wiring fault in the CTS to ECU loom. My circuit seems to be ok assuming the the brown/yellow core is supposed to be a ground connection in the ECU.

It looks like the ECU (and hence the glow plug circuit ) thinks the engine is hot when it is actually cold. Clearly the glow plug circuit does work in the cold start condition but only when I have the CTS disconnected.

I’d welcome any suggestions or experience of solving this problem.

thanks, Dave Mac

Do a bodge. Run a wire from the sensor to a switch inside the cabin and back. Switch it off to start from cold, then put it back on when running. I presume you are intending to keep the vehicle for a while and when you get shot it's not going to be worth a lot so the bodge won't matter.

Any sign of moisture/coolant inside the loom connector housing at sensor or ECU end?

The CTS is simply a thermistor whose resistance should decrease as it gets hotter, therefore anything which increases resistance will make the problem better not worse, this is why disconnecting the sensor so the ECU reads infinite resistance improves starting.

Therefore you should be looking for a partial short circuit rather than continuity testing, it is possible that the new sensor is junk, you should check resistance readings on the new and old sensors to confirm.

Wino - thanks for that, I’m a newbie to asking questions!! , yes engine code is ASY, it’s the non-turbo diesel 1.9. It’s been a good car for the kids in the high insurance cost years. We’ve kept it well maintained so keen to run for a bit longer, only 109k miles.

Have you seen or heard of this problem before? Could it be the ECU itself? I’d thought about Peter’s switch ‘bodge’ and may end up doing that, but I’d prefer a family proof fix if you get my drift.

 

That bit below the  + Quote line in my post is known as a 'signature'. It's a message to anyone reading my posts, not specifically you, but it is indeed useful to have the engine code in order to look at wiring diagrams, which are sectioned by engine code.  In your case, I think there is only ASY for the 1.9 SDI, so you gave enough info in your first post for me to work that out.

 

The question above those words, could be relevant though, in establishing possible causes for decreased resistance readings.

 

Edited 18 March, 20197 yr by Wino

Sepulchral, thanks, Existing CTS reads 3.6 kOhm, new one, (HAAS 223 44 0211 equivalent to OE no. 1100736) reads 4.1 kOhm. 

With the sensor and ECU connectors unmade, I get a good open circuit between the cores .

 

Wino, I found you question. No sign of moisture at either end, I’ve cleaned thoroughly with contact cleaner too.

 

 

 

It's gonna be a bit tedious, but another avenue to explore, I guess, is wire-to-wire shorts within the loom. So disconnect ECU plug and measure Ohms between each of the CTS pins and all of the others, watching out for any that aren't open-circuit.

Tests to other cores looked ok, but to be sure (in case of shorts to other cores in the loom), I  made up a new circuit between the ECU connector and  the sensor connector.  Still have the problem. Glow plug light goes out almost immediately if CTS connected and mega long cranks before smokey start. Disconnect sensor then nice long glow plug time and starts straight away. ECU? Has anybody seen this before??

Does the coolant temperature gauge in the cluster behave as expected?

It would be good to get VCDS on it and see what coolant  temperatures the engine ECU is reporting, compared with reality, cold and at operating temp.

 

Also, as Sepulchrave suggests, measuring cold resistance of your new CTS is wise, in case what you've bought is duff. At 10C it should be in the region of about 3.3-4.4k Ohms. Easy to do this from ECU connector end, I guess.

 

I would suggest swapping in another glow plug relay, just 'cos it's in the circuit; but it's beyond me how it could possibly be causing this scenario.

Wino

thanks,

The temperature gauge has consistently reached around 90 deg C once warmed up. 

I did check the old and new CTS resistances, they are both in spec.

I have yet, checked/cleaned the upper multi-connector on the ECU so I’ll do that before changing the relay and having a look around that circuit too for joint corrosion etc. 

I'm wondering if there's been some moisture ingress into the ECU casing,

If you didn't do it this way, it would be best to measure the cold resistance of your sensor from the ECU loom connector, rather than across the sensor pins; as that would prove that the loom was behaving itself too. Does that make sense?

Unplugging the CTS gives maximum glow plug time, have you considered that this may be covering up other issues? How have you refurbished the injectors? My 170K ALH starts on the button without waiting for any glow plugs and yours should be no different unless its -10

 

On this engine i'm almost 100% you can alter the glow plug time and make them come on for longer at warmer temperatures if you wanted an easy way out

Edited 20 March, 20197 yr by SuperbTWM

I have to consider a fault/moisture within the ECU. I have checked both old and new CTSs back at the ECU connector, with a newly made up loom that bypasses the buried loom that runs under the battery casing. The resistance values that I measured on Sunday  (see earlier post) were little changed whether probes were attached directly to the sensor or via the existing or my new loom.

I have not yet checked/cleaned the upper ECU connector for moisture ingress/corrosion, (The CTS cores, blue/red and yellow/brown are connected via the lower connector). That will be my next step when I get back to the car at the weekend.

Yes I realise that disconnecting the CTS simulates maximum resistance and hence longest glow plug time.  The tests imply that the total effective sensor resistance is lower than that of the CTS alone, but only when the ignition circuit is live (implying a parallel low resitance that becomes effective when starting). So the possibility of an ECU or glow plug circuit/relay fault would seem credible. Any advice on the glow plug relay location? My manual does not carry that information.

thanks for all your advice so far. Glad to know that others out there enjoy a challenge !!

Relay panel is approximately in front of your right knee when sitting in the driver's seat, behind lower dash panel.  Glowplug relay is in position 3 in the panel, third from left, top row, for your car.

The guide for brake light switch replacement covers removal of the lower dash panel nicely, here: https://www.briskoda.net/forums/topic/211083-diy-fabia-i-brake-light-switch-replacement/

 

I feel duty-bound to reiterate that I really can't imagine how the relay or its wiring can be the problem, it seems to be working when commanded to, in a consistent way, from what you describe.

It's probably easier to get into the ECU case to check that, than remove the dash panel to get at the relay.

 

If it's useful: glowplug relay coil is energised by pin 42 of an 81-way connector at the engine ECU being pulled low. Wire is red/blue at ECU end, changing to yellow/blue at the firewall connector. High side of coil is supplied by the output of the 'Diesel direct injection system relay' in position 15 of the fuseboard.

 

Edited 21 March, 20197 yr by Wino

Did you get anywhere with this?