Damn car dying again - Help UPDATE FROM EMERALD, NOW FIXED

[Will ing]

Damn car is dying again after the problem from the thread below had gone away.

http://forums.seloc.org/viewthread.php? ... 820&page=2

Summary
When the car dies the Emerald looses power (led goes off) temporarily - anything from 2 seconds to 2 days.

So I've started to trace the wires and I'm confused about the +12v feed into the emerald. In the emerald manual it sates Pin 11 (from ignition switch) and 28 (from the MFRU) are the two +12v supplies.

So I've disconnected the main ECU plug and put a test light into Pin11 - when the ignition is switched on the wire is live on a test light - so in theory no problem there.

Pin 28 comes from the MFRU and splits off to the Injectors, IACV and ECU. When a the test light is put into Pin 28 there is no live. HOWEVER if you reconnect the ECU connector and stab the wire going to pin 28 it is live?????

So simple question - is the live actually going to the ECU or coming from it?

Sorry this may be a really stupid question

Thanks Will

[robin]

The ECU has another signal - the main relay control output - when you turn the ignition on, the ECU turns on the main relay; the main relay then connects pin 28 to the battery.

So if you've ascertained that in your failure cases the pin at 28 is not live whilst the pin at 11 is live AND the ECU is connected to the loom there are the following possibilities for failure:

(1) The ECU is skunnered and just isn't turning on the main relay. This would be a very unusual fault.
(2) The wiring on the earth path to the MFRU is intermittent. That's ECU pin 4, white&pink wire.
(3) The live feed to the MFRU is intermittent - that's fuse 3, 40A solid brown wires to MFRU. Note these wires also travel through the engine harness connector at the ECU panel (all solid brown wires). If that isn't making a good connection then the junction will heat up and may then present too high a resistance to provide a reliable supply downstream.
(4) The MFRU itself is knackered. This isn't unheard of.
(5) The wiring from the MFRU back to the ECU is intermittent - that's the brown&pink wire to pin 28 of the ECU, as you've discovered.

Reading through your thread on SELOC I see that it still cranks whilst it won't start, but the fuel pump does not prime.

If that's correct you can eliminate fuse 3, 40A as it feeds the starter solenoid too. However, the engine harness connector is still worthy of inspection - look for signs of corrosion or overheating on the terminals.

Also, you've changed the MFRU so that eliminates that. Again, it's worth checking the MFRU connector for signs of corrosion or overheating.

How did you get on splicing in the spare wires and trying to hotwire on failure?

I would look very carefully at the white&pink and brown&pink wires for possible interior wiring breaks - I've seen it before on one elise - you may need to bite the bullet and replace the white&pink wire (you can buy replacement terminals for the ECU connector - I can post you one once I'm back from holidays) - at the other end you'll need to splice the new wire into the stub of the old wire, or work out where to source the appropriate terminal from.

The brown&pink wire is a lot harder to deal with. What I suggest is this: when it next fails, disconnect the ECU and measure continuity from pin 23 to pin 28. You expect to see somewhere around 5-10 ohms. What you're doing is measuring the path backwards through the injector pair 2&3, through the brown&pink wires, through the injector harness connector, through the bridging link in the MFRU (this is a permanent link - no power required) and back to the ECU pin 28.

You might want to practice that whilst all is well to see what resistance to expect and to make sure you've identified the correct pins, and the correct multi-meter settings. The key here is that the brown&pink circuit is in two parts in the loom - one part feeds the injectors, the other part feeds the ECU, IACV, canister purge valve and ignition coil. If there's a split somewhere inside the loom that effects the power to the ECU, this test will identify it's presence - you'll still need to then split the loom to locate the problem (or just bypass the wiring).

Cheers,
Robin

P.S. I'm away on holiday but if you want to chat about this feel free to give me a call on 07973 391 393 - especially if you end up stranded by the side of the road somewhere - more than happy to help.

[Will ing]

Thanks robin excellent info as always,

The ECU has another signal - the main relay control output - when you turn the ignition on, the ECU turns on the main relay; the main relay then connects pin 28 to the battery.

Ah that would explain it.

So if you've ascertained that in your failure cases the pin at 28 is not live whilst the pin at 11 is live AND the ECU is connected to the loom there are the following possibilities for failure:

I'm pretty sure this is the case, now I understand the first point above it should be a lot clearer - my mind was going into overdrive trying to explain what was going on (adding 2 + 2 and making 5). I will check this again just to make absoutely sure

(1) The ECU is skunnered and just isn't turning on the main relay. This would be a very unusual fault.

I just have a feeling it might be this - the car was fine before the engine replacement and the only new electrical item is the Emerald - I was going to see if I could borrow one from them to eliminate this (they will probably say no so I might just buy another one and if it isn't that sell it.

(2) The wiring on the earth path to the MFRU is intermittent. That's ECU pin 4, white&pink wire.

Not checked this but i've gone through all of the earths on the car so will do this next - looking at the wiring diagram on the Elise service notes this does look like a possible

the engine harness connector is still worthy of inspection - look for signs of corrosion or overheating on the terminals.

Gone through all of the connections they all look fine

Also, you've changed the MFRU so that eliminates that. Again, it's worth checking the MFRU connector for signs of corrosion or overheating.

Again all looks fine

How did you get on splicing in the spare wires and trying to hotwire on failure?

I didn't need to on Pin11 as it has a good live, I got cold feet on pin 28 due to the first point above not really understanding when it should be live but I suppose now it will not hurt to have a go

I would look very carefully at the white&pink and brown&pink wires for possible interior wiring breaks - I've seen it before on one elise - you may need to bite the bullet and replace the white&pink wire (you can buy replacement terminals for the ECU connector - I can post you one once I'm back from holidays) - at the other end you'll need to splice the new wire into the stub of the old wire, or work out where to source the appropriate terminal from.

I might do this one last as Ive wiggled all of the wire when it has failed and if it was an intermittent connection you would have thought it would be be found by doing this (not always of course)

The brown&pink wire is a lot harder to deal with. What I suggest is this: when it next fails, disconnect the ECU and measure continuity from pin 23 to pin 28. You expect to see somewhere around 5-10 ohms. What you're doing is measuring the path backwards through the injector pair 2&3, through the brown&pink wires, through the injector harness connector, through the bridging link in the MFRU (this is a permanent link - no power required) and back to the ECU pin 28.

You might want to practice that whilst all is well to see what resistance to expect and to make sure you've identified the correct pins, and the correct multi-meter settings. The key here is that the brown&pink circuit is in two parts in the loom - one part feeds the injectors, the other part feeds the ECU, IACV, canister purge valve and ignition coil. If there's a split somewhere inside the loom that effects the power to the ECU, this test will identify it's presence - you'll still need to then split the loom to locate the problem (or just bypass the wiring).

I will have a go at this

robin thanks again for your advise and help

[tut]

Don't forget to take your laptop with you Robin, there are bound to be wireless points along the canals or rivers, and we may just need to tap into the Mekon brain.

Holidays are no excuse for getting away from it all.



tut

[Mikie711]

Will,
if you need ECU connectors I've got spares, also if you want to come round I can unload your map from the emerald and load it to mine, stick it in and see if the same happens. Might eliminate/confirm your emeralds health.


Mike

[Will ing]

Mike,

Thats very kind I will take you up on that offer - you have U2U.

Thanks Will

[robin]

Just back from my holidays ... did you make any progress with this? I have a spare emerald, though it sounds like you have one closer to home

Cheers,
Robin

[meatball]

Just back from my holidays ... did you make any progress with this? I have a spare emerald, though it sounds like you have one closer to home

Cheers,
Robin

Welcome back sir!

[Will ing]

Hi Robin,

Its pretty disgaceful..... I've only been able to get on to it today, with work and easter etc

So i'm just charging the battery now and will have a go this afternoon - actually a thought on this, I'm runing with a small Varley Red Top battery (pretty new) i'm just wondering if this could be a potential cause - surely not.

Anyway so far I have ascertained ive got 8.8 ohms between pin 23 and 28 as per the instructions above and now i'm ready for the next break down

Also ive wired in two 'Hot wires', one to power pin 28 and the other for earth.

For the live I've gone through the live connection point from pin 28 (in the loom) where it splits off to the IACV, injectors, canister and it all looks excellent so I've spliced a flying lead into there and im now ready to touch it to the starter as soon as it breaks down.

The second lead is spliced into the main earth connector point again in the loom (about 6 inches down from the ECU plug where they all come together with solder/crimp connections) again when it dies i can simply touch this to a good earth and see what happens.

I will post up agian with what I find

Will

[Will ing]

OK had a couple of days at it now but with limited success as it only died once for about 20 seconds. I managed to ascertain the two following points...

1, I can confirm when it fails I have a reliable live at Pin 11 from the ignition but pin 28 from the MFRU is dead.

2, next I hot wired a live feed into Pin 28 from the starter motor (I tested it was live with the ECU connected by stabbing the loom) with this live feed into the ECU the ECU LED did not come nor did the fuel pump prime and thus I was unable to start it.

NEXT - When it next dies I need to...

1, check the resistance between pin 23 to 28 as advised earlier by robin
2, earth out the flying lead I have fitted to see if the earth is dropping out
3, lastly try another Emerald.

[robin]

Hmm. Have you checked that your pin 28 flying lead has +12 on it under normal circumstances - i.e. it is actually spliced into the right lead.

AFAIK as soon as pin 11 goes live and the main ECU earth is connected, the ECU will attempt to turn on the MFRU main relay and thus make pin 28 live. If pin 28 is not live when pin 11 is live AND forcing pin 28 live does not turn the ECU light on THEN you would have to conclude that either the ECU is gubbed or the ECU earth is not connected very well.

Cheers,
Robin

[Will ing]

Hmm. Have you checked that your pin 28 flying lead has +12 on it under normal circumstances - i.e. it is actually spliced into the right lead.

Yes definately has, just rechecked to make sure

If pin 28 is not live when pin 11 is live AND forcing pin 28 live does not turn the ECU light on THEN you would have to conclude that either the ECU is gubbed or the ECU earth is not connected very well.

Yes agree i just need to rule out the earth but typically it has not died again all weekend

Cheers

[Will ing]

Well after a week or two with no sign of dying it died twice on the way to work this morning

I attached the flying earth lead and it did nothing (the only exception here being I did not attached it direct to the battery earth - but I have gone through all of the chassis earth connectors right back to the battery already).

Only test I still haven't managed to do is check the resistance between pin 23 to 28.

SO, bar this last test, I can now say confidently that as there are three main connectors for the ECU, 2 live and 1 earth, and as I have connected good sources to all of these then the ECU must be temporarily cutting out.

Is it time to call emerald to discuss this with them......or is there a simple way I could test the ECU (without knackering it of course) i.e. when it dies disconnect and remove the emerald and connect a good earth to the Pin 29 and two lives pin 11 and 28 via a 12v battery - if the LED comes on the ECU is OK if it doesn't it is f@cked?

what do you think?

[Will ing]

after the car dying on the way to Knockhill the other saturday I was able to finally prove it was the Emerald and have just got the following email

Hi Will,
Your ECU has now been checked and repaired, free of charge. There seemed to be a fault with the PCB, but Karl has now repaired this and all is working as it should.
I have sent your package out today with Fed-Ex, so you should receive it in the next few days.
Kind Regards
Kim
Emerald Office

Thank you all for your help and input especially robin for helping me to prove it was the ECU

[robin]

Unusual fault!

Glad it's working again ...

Cheers,
Robin