[SteveB]

One of the most frustrating problems . To cut a long story short, the car (‘95 214SEi) had been running fine until a couple of weeks ago. As I accelerated gently it missed a couple of beats - a few more revs and it was fine.

The problem quickly got worse, with the odd misfire on tickover and small throttle openings. Higher revs were fine, so it didn’t sound like any kind of fuel starvation.

I plugged in RoverMEMS – no fault codes, the TPS seemed to work fine but output from the lambda looked erratic. Probably caused by the erratic running as unplugging the lambda sensor made no difference.

I swapped the rotor arm and distributor cap – no better.

The car started and ran pretty much ok from cold – the problem seemed to get worse as it warmed up. I swapped the crank position sensor in case that was on its way out – no better.

I replaced the king lead then cleaned up the earth point on the slam panel. One of these did the trick – tickover was smooth again and the car ran really well.

For a few days - then the same thing happened again. I swapped out the coil, plug leads and spark plugs (I’d actually renewed the whole lot last year but wanted to rule them out). Also unplugged the coolant sensor. No difference,

Next morning it was running fine again(!). I took a chance and headed out to get some petrol – no problem at all, running smoothly with no misfire. After filling up it started no problem but after about twenty yards it started misfiring and the engine died. It was cranking over fine but not firing at all. I removed a plug lead and checked for a spark – nothing. For the sake of trying something I refitted the old rotor arm and king lead. It fired up straight away and got us home.

So – I think the breakdown confirms it’s spark related. I’ve replaced everything from the coil through to the plugs, and the crank sensor. I’m guessing it’s either a low tension supply problem to the coil, or the ECU isn’t firing the coil correctly. I've tried moving the coil connectors with the engine running (in the hope it would cause a misfire) but nothing happened.

Does anybody have experience of ECU failure? Is it likely to fail gradually/erratically like this, or would it just stop working completely once and for all?

Any comments appreciated, thanks.

[Mr Teddy Bear]

Sounds like ECU failure to me! my experience was of a flat spot at the bottom of the rev' range that got worse. I then stupidly left the bonnet up all day with the sun shining directly onto the ECU case that finished it off!
The heat must have destroyed various soldered joints that were going dry anyway.

I couldn't find what was wrong with in on my drive although it smelt like fuel was present but no spark. The car was low loaded to Wilcox's over in Wooten Under Edge as they were still Mg-Rover dealers.


The car was a 1995 registered 414Si with a plastic inlet manifold, the old ECU was sent away for test & repair and who ever looked at it reported back that it was fine [it wasn't] they checked everything else as you have done. Eventually they swapped in a new one, they must have had one in stock [they buy up old Rover dealer stock when it comes onto the market] because a new un fixed it!

Swopping around the ignition lead set is a red herring imo.

95 model year builds started in November/December 1994 I know this as owner of one; they still had the alloy inlet manifold & different throttle body ect' the ECU differs from my M registered vehicle [that had the fault] that I have now scrapped. The build order changed again in 95 to the plastic inlet manifold etc, although I've been told it had a wet liner engine because of the casting strengthening ribs on the block.

The ECU differs between the alloy & plastic inlet manifold engines probably due to the throttle body and TPS changing?


I still have the ECU kept in case I ever need it!


Edited

[g259fsg]

Sounds like you've been very thorough and replaced everything potentially a problem apart from the ECU. If you can get it to stall again and there is no spark at the plugs, check to see if there is a spark at the coil. Also check the resistance of the HT lead from the coil to the distributor. If you've got access to an oscilloscope, you could see if the ECU is sending pulses to the LT side of the coil.

Years ago, I had a lot of problems with my old Rootes cars stalling when hot and not restarting until they had cooled down. Spent a lot of time replacing bits, and often found they restarted if I swapped ignition components. In the end I found it was the rotor arms which break down when hot giving weak or no HT at the plugs. I'd fitted new ones without curing the problem which was very confusing. Eventually found the after-market components are often substandard even when new. The ones used on the Rover engine did have a reputation for this kind of failure, though I've not had that problem recently with good quality parts.

It's always worth doing some tests and not assume that a new component is fault-free and therefore eliminates the cause of the problem.

[rjessett]

ECU's changed M/N as the later systems went to "robust" i.e. they have to get a code from the 5AS alarm to run, and thus, these mems1.6 ecus need to be matched. You can't put an earlier ECU in as the Flywheel configuration was changed (from 2 gap to 4 gap)

Im guessing your ecu part number is MKC103020 or it could be a later one (there was a supersession to this)

MEMS1.6 doesnt really tell you when its going wrong (even on T4) . In some cases, they will either work or they wont, in others, if components/joints degrade, I can easily see how it could effect the signals the ecu is getting

Id certainly get another ecu and swap it to rule it out.

Its reletively easy to convert these later cars to MEMS1.9 from a Bubble 214 if needs must.

[SteveB]

Many thanks for the quick replies – much appreciated.

I’ve been trying to track down a spare ECU for the past couple of years, planning to send it to Technozen to have the immobiliser overriden – just in case! It’s actually MKC103580 which seem to be very thin on the ground.

Would it need to be exactly the same part number? Any idea of any alternatives? Conversion to a later unit sounds very interesting, as these seem more plentiful - more chance of finding replacements. Is there much involved in the conversion?

Thanks again for the help.

[rjessett]

MKC103580 is indeed the supercession to MKC103020 for the R8 214 MPI Both need to be matched to 5AS.

Both of these also quite hard to get hold of, as they were not used for every long. Im not sure if there is actually any practical difference between the two. "In thery" they should both be the same, Ive just not tried it yet.

I do have a couple of MKC103580's , but one of them is a "never used remanuafactured version" wasnt cheap - there is a company that did it - https://www.atcdrivetrain.co.uk/c/rover-214-petrol-ecu (MKC103580 is listed on there as being for the 8v, but im assuing yours isnt an being an SEi! ? and a listing error?)

Im working on how to remove the immobilisation function from MEMS1.6 - I think I can do it, just not got a car to hand to 100% prove it works at the moment (but I will do shortly)


To migrate to MEMS1.9 - it basically involves getting an ecu from an R3 214 - matching it with the 5AS, and modifying the injector wiring. mems1.6 is sequencial firing, 1.9 is paired. I also swapped the injectors over so they were the later cream coloured ones, as opposed to the blue, just so they were the same spec as 1.9 expects. Most of the other sensors are the same (apart from the crank, but that doesnt really matter)

[SteveB]

Thanks for the replies, suggestions and information.

There’s been a re-manufactured item on ebay for a while now which I’ve been a bit wary of. Turns out it’s the same company in your link, based in Cannock so I’ve ordered one. And you're right, the car’s 16v so presumably an error in their listing. Just hope it fixes the problem! If not, at least I’ll have a spare ECU

As for the cost, yeah, compared to a lot of Rover parts it’s not cheap, especially as the same company are selling a re-manufactured MNE101110 for less than half the price! However, to put it in context, a friend of my wife’s just been quoted £2,700 by a main dealer to replace the ‘infotainment’ screen in her Honda SUV! £2,700!!! For a touch screen thingy! Could buy a nice Rover for that

Incidentally, there are loads of MKC103020s on ebay – the vast majority from Spain! I hadn’t realised Rovers were so popular over there – ex-pats flying the flag?

Updating to MEMS 1.9 sounds like an interesting project for the future – more widely available ECUs and more widely available injectors.

Interesting point about different ECUs for the different throttle bodies and TPS. When you look at the listings it’s amazing how many different part numbers there are, all to run basically the same engine. There must be some very subtle differences, relating, presumably to the rest of the car spec.

When it comes to dodgy rotor arms, my first Rover was a ‘91 214 Si which I ran for 14 years. At one point it developed a very intermittent misfire (long before forums like this existed) and after a lot of head scratching and trial and error I eventually replaced the rotor arm and sorted it.

I should know by the end of the week if the replacement ECU solves the problem; I’ll post the outcome then. Thanks again everyone.

[SteveB]

The replacement ECU arrived on Thursday. Since then, I’ve done about 40 miles – the car’s started every time and run quite smoothly (so far!), so it does look as if the ECU was the problem.

To be fair, it’s done well to last 29 years, when I think of the number of PCs I’ve had that have fallen over in that time!

[rjessett]

good news