HELP! my 350i engine misfiring now won't start.. :(

[sprevett]

was intermittently misfiring so I've been fiddling to check for a loose connection....unplugged relays & fuses and plugged them back in....unplugged ECU then back in again etc etc/...now the effing thing wont start at all! clouds of smoke and smells of petrol...any idea?s

[GreenV8S]

Where you've been fiddling - go do some more!

Alternatively: Do you have a spark at the plugs? Are the plugs wet?

[sprevett]

will check the plugs tomorrow - i'm assuming the smell of petrol combined with smoky fumes from exhaust means fuel is present - although I can't hear the pump running ...do they normally whirr/make a noise? so maybe I've done something to the ignition circuit - or maybe i really shouldn't have unplugged the ecu - are they a bit sensitive? i also notice that a short blue wire 'loop' with a diode in it on the end of the realy panel has broken off - is it important?

[convert]

Make sure that the hose from the airflow meter to the engine hasn't come loose / off.

[sprevett]

ok - fiddled around some more, removed ECU from panel to re-connect the big connector FIRMLY, pressed all the relays home again, fiddled with the water temp senders...and it started. still running erraticly at idle and misfiring/stumbling on the road (mainly at low throttle it seems)...so still none the wiser. yes, also checked the air hose to plenum and checked all vacuum hoses too. it's got a replacement Air flow meter i got from a the breakers (off anther 350i) so I probably ought to get it on a rolling road with someone who knows their RV8s....oh and also noticed a slight water leak under the plenum....looks tricky to fix and I assume I would need to remove the plenum so prob easier just to remove the hoses - I assume it's just to heat the plenum for winter use??

[350zwelgje]

- Have you checked the coolant temperature sensor? This can cause misfiring/non-starting especially when making the ECU to flood the engine. You can shortcircuit it for testing purpuses as the ECU 'thinks' the engine is hot....

- And would check the throttle pot setting (idle about 325 miliVolts) if the coolant sensor is not ok.

- Coldstart injector disconnected? When active at normal outside temperatures it can inject massively too much petrol.

The plenum is heated. You can remove it. If you do a search you can find more information on it. Think you need to blank something of which is described as well.

Rob

[sprevett]

thanks Rob - will check the temp sensors again (there are two next to each other in front of the plenum is that right? - what does each one do? and which is the important one for the ECU? i think it's the front one IIRC and if so it doesnt appear to have the 10k resistor 'mod' - is that something i should do too?)

yes i have disconnected the cold start injector.

will check throttle pot voltages tomorrow too.

thanks again
steve

[jon haines]

i have had the same running probs you describe but slightly easier to narrow down due to having alread replaced air flow meter, cold start,coolant temp thing not much left so got a throttle pot for it and cured problem.

[350zwelgje]

sprevett said:

thanks Rob - will check the temp sensors again (there are two next to each other in front of the plenum is that right? - what does each one do? and which is the important one for the ECU? i think it's the front one IIRC and if so it doesnt appear to have the 10k resistor 'mod' - is that something i should do too?)

yes i have disconnected the cold start injector.

will check throttle pot voltages tomorrow too.

thanks again
steve

The ECU is controlled by the sensor closest to the plenum, the other independently of the ECU operates the thermotime thingy in one of the hoses to the plenum (only on startup until a bit warm).

Don't know anything about a 10k resistor mod, tell me more....

Worst case (mine a couple of years ago) was that the soldering within the ECU needed to be redone. Think this is worst on the flapper type, which I have. After resoldering no issues. Check for this could be to plug a know working ECU in and see if it runs fine.

Good luck,

Rob

[honestjohntoo]

sprevett said:

was intermittently misfiring so I've been fiddling to check for a loose connection....unplugged relays & fuses and plugged them back in....unplugged ECU then back in again etc etc/...now the effing thing wont start at all! clouds of smoke and smells of petrol...any idea?s

Regretfully, uninformed fiddling may have caused more complications than existed beforehand.

First, you can read about every Efi component and how they work at this archive:

http://www.vintagemodelairplane.com/pages/Rover_Te...

right down to the fine detail, location, purpose, function, faults, testing and solutions. Nothing is missed out. Open the PDF's where available from each intro page.

Second, with regard to your questions,

Misfiring, clouds of smoke and smell of petrol means its trying to run on a mixture that is too rich. You may just have simply flooded it, in which case it will usually fire up after being left to dry out.

If not, then the probable cause is an open circuit connection on the temp sensor circuit to the ECU or the temps sensor itself. The article on that component explains all.

You were asked if the plugs were wet? and the answer is???

You were asked if the plugs were sparking? nice fat ones are needed, check them out at the king lead first and then the plugs. report back.

You checked for some loose connections and hose air leaks. plain visual inspection has only a minority chance of locating an air leak. Read the plenum article for a full blown air leak testing program.

You seemed to have induced some improvement by fiddling with the temp sensor connector, etc. Your problem could be right just there. Air leaks and connection probs together account for over 80% of all Efi system problems. For connector solutions - read this.

http://www.vintagemodelairplane.com/pages/Download...

Water leaks from under the plenum may indicate that there is also an air leak into the plenum thereabouts because the hotspot securing bolts are drilled right through into the area of the plenum tunnel so check that the fixing holes or the water jacket thereabouts are not open to atmosphere. Making the tunnel heater non-functional is not a critical issue.

The 10k resistor is an "anti-hunting at idle" modification, A very simple additional component for quite complex reasons - again, not a critical issue, see the whole story here:

http://www.vintagemodelairplane.com/pages/Snippets...

I've never had to resort to rolling road and you may find the RV8 injection expert you need. On the other hand you might find yourself paying for lots of new components that were never faulty in the first place. Knowledge is the key, and everything you could need to get started is on the links indicated above.

Good luck

[sprevett]

this is BRILLIANT information - thanks for taking the time - I have printed off some of the pdfs and will try to work logoically thru the various components and will report back. thanks again.

[sprevett]

ok now a mystery has unfolded...the connector for the temp sensor is plugged into the thermotime switch and vice versa...which is possible because the connectors are the same (although one's brown and one's black). Hmm..so I swapped them onto their rightful place, and the engine wouldn't run on more than 3 cylinders! (lots of smoke/petrol smell etc) so I swapped them back again and engine runs OK. (but still with occasional misfire of course)

So what's happened? the only answer I can dream of is that the temp sensor is broken and someone swapped them and for some bizarre reason the thermotime switch gives a better signal than a broken temp sensor and allows the engine to run? and since I don't have the cold start injector plugged in then the thermotime lead doesn't do anything anyway...

anyone got any other ideas? and yes I have double checked the wiring colour codes and they are definitely on the wrong connector (assuming the front sensor is the thermotime and the rear one (nearest the plenum) is the temp sensor, correct?

[adam quantrill]

Yes that's the right way around, the thermotime switch hardly ever goes wrong and is the easiest one to get at, and you usually have to remove it to change the temp sensor ;^)

Next step is to short out the temp sensor and see how it runs.

Then order a new one but DON'T get an Intermotor part or you'll be back here in 6 months time...

[honestjohntoo]

sprevett said:

anyone got any other ideas? and yes I have double checked the wiring colour codes and they are definitely on the wrong connector (assuming the front sensor is the thermotime and the rear one (nearest the plenum) is the temp sensor, correct?

Given that you have all the necessary information about the temp sensor and I already explained that that component was the more likely source of your problem,

honestjohntoo said:

If not, then the probable cause is an open circuit connection on the temp sensor circuit to the ECU or the temps sensor itself. The article on that component explains all.

You therfor must take the next logical step to test the temps sensor and its associated wiring preciisely as directed in the article.

Testing

Conditions as follows: Engine HOT, ECU multiplug disconnected as in A: Ignition OFF: Measure coolant temperature with thermal probe: Connect an ohmmeter as shown in B below and take a single reading. Compare it to the resistance/temperature profile below. Consider connecting the ohmmeter for only short periods, not in fear of damage, but the meter battery may cause the sensor to heat up and give an incorrect reading. Allow the engine to cool and take further readings. If the measurements match the profile within (say) 10-15%, the CTS and its connections are OK.[/colour=red]

TEMP.......OHMS........TEMP........OHMS

-10°C....9100-9300....60°C.......500-700

0°C......5700-5900....80°C.......300-400

20°C.....2400-2600....100°C......150-200

40°C.....1100-1300

If the readings are clearly wrong, remove the sensor for a controlled test using a pan of water and thermometer. Suspend the sensor in the water and bring to the boil. Compare resistance readings to the profile. Variations of (say) 10 -15% are not critical but if they are obviously way off or there is a clear intermittant or distorted result then consider the test failed. Obtain a replacement unit, check it as above to verify OK, fit it into the manifold and reconnect to the Efi circuit. Reconnect the ohmmeter as in B and repeat the measurement between the ECU pin 13 and earth. If this still incorrect there is a wiring problem so check the black/slate wire (BS) and black/white wire (BW) and their connections paying particular attention to the earth stud at the rear of the engine.

The bottom line is that the ECU must see the temp sensor input, hence the check made at the ECU multiplug. If you are unfamiliar with multimeters go here:

http://www.vintagemodelairplane.com/pages/Snippets...

[honestjohntoo]

adam quantrill said:

Next step is to short out the temp sensor and see how it runs.

Er??

If no change is detected that could be interpretted two ways.

1. faulty temp sensor - should be verified by the temp vs resistance test

1. open circuit connector, in which case buying a new component will produce the same symptoms.

[adam quantrill]

I think we would expect a change, because of the behaviour already noted when connected to the thermotime sensor. But we'll see ;^)

[sprevett]

OMG this is sounding suspiciously like a physics lesson I once had in 1978.....are any of you a member of the academic fraternity? own up...and if any of you taught physics at Reigate Grammar in 1978 yes it was me who broke the van der graf generator, sorry.

now, if someone can just tell me which cupboard the thermal probe and ohmeter are in? is it the one with the bunsen burners in? (and surely I should be first locating and wearing eye goggles and ear defenders)

yes, yes I know.....I'm now going to be put in detention for being a smart alec. sorry. yes I will test the temp sensor at first light and plot the results on graph and hand it in by end of school, promise. Or I might just get a new temp sensor and shove it in eh.

[honestjohntoo]

adam quantrill said:

because of the behaviour already noted when connected to the thermotime sensor.

You are surely right if all the wiring is good. Otherwise, that change in behaviour may be because the innards of the TT switch are connected to earth whereas the innards of the TS is not. An O/C on the TS earth line will prevent the ECU seeing any short circuit across the TS contacts.

[Brap_Brap]

One more argument to ditch the stty LUCASfer (Prince of electrical darkness) based EFI IMHO.

[edit]

Can anyone show me a sub 8 second 1/4 mile car running EFI?

I rest my case.

Edited by Brap_Brap on Thursday 17th September 01:00

[350zwelgje]

On the topic of just buying a new coolant temp sensor.
They can be faulty from new!!!!
I bought two new ones based on this knowledge and luckily both are ok, so a known working spare in the spares collection....

Rob