I understand this is a common problem, where one pair of cylinders corresponding to one of the coil pack misfires. I get a fault code of P0352 associated with Ignition Coil B error/fault.

I note that if the ECU believes there is a misfire, then the fuel injectors on that pair of cylinders also stop injecting.

In this case I have a spark on 2 and 3, but seemingly weaker than cylinders 1 and 4.

I have swapped connector and plug leads between the two coils and still a misfire on cylinders 2 and 3.

I have swapped the plugs around, swapping 1&2, and 3&4 around and the problem still exists on cylinders 2 and 3.

I have an oscilloscope on the ignition drives and they look rather different. Are there any waveforms I can compare these with?

All this does suggests that the ECU ignition driving cylinders 2 and 3 is faulty.

Any thoughts? Anything to avoid paying out for a new ECU!!

A minor addition to this which is probably the most confounding!

After a while of being driven, the centre 2 cylinders do kick in, and you can drive 100 miles or more to a destination without problem. After cooling down the problem may or not recur.

This is a common problem on both Cento and Punto MPI ecu's.

I have heard people claiming you can pop them open and are likely to find a burnt track (all though it sounds to me more like a transistor issue). I wouldn't recommend this though unless you're a dab hand with electronics.

Plenty of places repair them. I've always used ECU Doctor in the past, I've never sent them a Fiat unit though so it would be best to phone them first and check. I'm sure some other chaps on here will be able to recommend similar places they have used.

http://the-ecu-doctor.co.uk/

Also worth noting a bad coil pack is known to damage the ecu so I'd probably change these at the same time .

Edit:

In your case it sounds like an issue with a dry joint if it's intermittent.

I have swapped connector and plug leads between the two coils and still a misfire on cylinders 2 and 3.

I have swapped the plugs around, swapping 1&2, and 3&4 around and the problem still exists on cylinders 2 and 3.

Click to expand...

Unless you can induce misfiring on 1 and 4 by swapping the coils around, it's not, repeat not, electrical.

If you can't get the misfire to jump cylinders, do a compression test.

There is a long standing issue with the "micro" ECUs (which have 2 connectors). The later, large 1 pin ECUs on the later MPI cars are exceptionally reliable (if a little fussy).

If I'm reading right he sais he's swapped the coil plug then moved the leads to the same two cylinders? Then tried swapping the spark plugs.

Also the ECU IAW 59F.M7 is the micro ecu.

The fact that a scope shows different outputs is almost a complete confirmation.

I have to admit though it is slightly confusing to read. I'm fairly sure I've read it right though.

Many thanks for your ideas, however after further examination I can't agree with your diagnosis of "Unless you can induce misfiring on 1 and 4 by swapping the coils around, it's not, repeat not, electrical." I'm wondering why you say that?

I put an oscilloscope on the coil drive pin of the coilpack driving pots 2 and 3, and after the ECU pulls down the drive pin to ground, it goes much of the way back up to 12V before opening the switch fully to create a substantial, but weaker spark than on pots 1 and 4.

This and the previous tests and swaps leads me to suggest that the ECU drive to the coilpack is very weak, and where the ECU knows this and turns off pots 2 and 3 injectors. The fact there is no petrol smell at the exhaust confirms this.

Can you any alternative to the ECU being at fault?

fingers99 said:

Unless you can induce misfiring on 1 and 4 by swapping the coils around, it's not, repeat not, electrical.

If you can't get the misfire to jump cylinders, do a compression test.

There is a long standing issue with the "micro" ECUs (which have 2 connectors). The later, large 1 pin ECUs on the later MPI cars are exceptionally reliable (if a little fussy).

Click to expand...

I apologise for the series of confusing posts. This is indeed a "micro" ECU which has 2 connectors.

Are they easy to take apart? I am an electronic engineer so if possible it is something I can do!!

Mikexx said:

Can you any alternative to the ECU being at fault?

Click to expand...

Absolutely -- HG failure between 2 cylinders.

Note -- really we need Woj on here to confirm -- neither spark or injector are individually switched.

If your hypothesis is correct, it should be easy to get the misfire to switch cylinders. If mine, then it will be completely impossible.

The ECU can be dismantled -- thaty's what the folk offering exchange ones do.

In the Punto section, guides or FAQs (although the Punto section is a dog's dinner at the moment, it should be in the Mk2 FAQs/guides) there's a pretty comprehensive guide to the common micro ECU failures.

Since they can be repaired I assume they can't be potted so it should just be a case of undoing the screws and opening it up/ maybe a seal around the edge. However I am speculating a bit but others claim to have done it.

I haven't had the opportunity to open one up yet.

I have seen a few but they were both exchanged with Fiat due to some ongoing part exchange thing with the newer Puntos (the problem is that common...).

The biggest challenge is simulating a running the engine to chase the fault. Places like ECU doctor have rigs for this very purpose which is why I have sent stuff to them in the past for other cars. Especially nasty Can bus junk!

You could pulse/test If you can find the amplifier stages etc. But most likely you'll have to look for anything obvious.

Anyway if your an Electronic engineer there's no point in me trying to teach you how to suck eggs.

fingers99 said:

Absolutely -- HG failure between 2 cylinders.

Note -- really we need Woj on here to confirm -- neither spark or injector are individually switched.

If your hypothesis is correct, it should be easy to get the misfire to switch cylinders. If mine, then it will be completely impossible.

The ECU can be dismantled -- thaty's what the folk offering exchange ones do.

Click to expand...

Many thanks for your input.

This is a MPI Seicento. Not sure what you mean by "neither spark or injector are individually switched." I understand that where a coil is deemed faulty by the ECU, the corresponding injectors are turned off to save the cat.

I don't believe this can be head gasket failure because the centre 2 cylinders do kick in suddenly, quite literally doubling the available power - or more!!

Most of the ECU exchange prices are quite high, implying ECU dismantling and reassembly isn't trivial. As you say it must be doable.

Injection is semi sequential -- 2 injectors are fired together. Ignition is wasted spark (plugs fire on compression and exhaust strokes).

I'm not convinced that the ECU knows which individual cylinder is firing (maybe from the cam sensor?). And I'm pretty damn sure that ECUs of this age don't cut out individual injectors (or pairs of injectors) to save the cat.

BWTH. 1 and 4 are fired by one coil, 2 and 3 by another. Ergo, if swapping the coils swaps the misfire
the fault is most likely with the coil. (I'd not completely rule out the cam sensor) Or elsewhere with the ECU.

Otherwise, it will be a mechanical issue. The fault most likely to cause a low speed misfire which clears at higher speed is an HG failure between 2 cylinders. It's to do with gas speed, IIRC.

It doesn't clear at higher speed. It does clear over a journey, and presume it is temperature related. No water/oil problems plus the instantaneous cure of the problem suggest its not a head gasket.

Similarly the voltage waveform on one coil is very different to the drive to the second coil.

I have already swapped the coils and the misfire didn't swap with the coils, but remained on the centre 2 cylinders.

Regarding the injector being turned off corresponding to the faulty coil, I was going by:
http://www.permoveo.ltd.uk/tabid/122/OBD-Error-Code/P0352/Diagnostics/Fiat/Default.aspx The lack of petrol smell from the exhaust does suggest the injectors are turning off.

I am surprised that the injectors are fired in pairs. There are 4 separate wires to the ECU, and I can't help thinking that Fiat wouldn't any more wires than they had to! One day, if I have time, I'll scope the injector drives and report back.

I have already swapped the coils and the misfire didn't swap with the coils, but remained on the centre 2 cylinders.

If it didn't swap with the coils, all logic suggests a mechanical cause.

I would do the swapping routine. Again, in the morning.

Remember that the ECU does not know which set of coils (it does not drive individual cylinders!) it is driving. Switch the LT connector, you switch the drivers. Switch the coil, you switch the coil. And so on with the spark plug leads (but in pairs, please). In this fashion you can eliminate every possible issue with the coils/drivers/ecu, at least as far as ther ignition is concerned.

On a car of this age, I would do a visual inspection of the wiring to the injectors -- it is pretty exposed -- and to the cam sensor (evil invention).

But sure, a head gasket failure between cylinders will often not result in water or oil problem -- there is no water or oil gallery there.

This seems appropriate, but I've not read it thoroughly. The one on the permoveo site seems to presume Coil On Plug (generally fully sequential) which in the Sei's case, you certainly do not have!

Now, again, if the misfire doesn't switch cylinders, how can it be electrical?

Come back Woj, all is forgiven!

fingers99 said:

I have already swapped the coils and the misfire didn't swap with the coils, but remained on the centre 2 cylinders.

Click to expand...

Because he swapped the coil connectors and the plug leads at the same time. He hasn't tried just swapping the coil conectors without moving the plug leads. This shouldn't even work on bank fire if the wrong set of injectors are firing and plugs sparking, I'm not sure what the timing is like it may run to some extent.

This is what I think he's suggesting he is going to do next and will prove it's not the head gasket if the misfire moves. This is what he hasn't done and you seem to beleive he has...

Cam sensor lets the engine no exactly where it is at all times, and yes the engines are bank fire, i.e. they fire the injectors in pairs.

Cam sensor is definetly worth a look but I really don't beleive this is a head gasket.

Easy answer, pull a plug from the duff pair and earth it to the block with a jump lead. Compare it to one from the good pair by doing the same thing.

Compare the quality of the sparks.

My money is on the spark being weak on one pair. I've seen this myself, the only thing I've never known before is one come back to life during a drive, but my money is still on the ECU.

I guess it could be one set of injectors not firing but the spark test will prove this.

If it wasn't an electrical problem why would the ecu be showing a coil fault code also??

Last point is a good one.

I dunno. I'd suggest coil pack, bwth.

Those ECUs are evil. Wonder how easy it is to swap to a late SPI type?

Swapping to an 18f ecu would be do able,but would take a bit of rewiring and a new rad from a cinq due to ecu fan control cam sensor can be ignored if the crank sensor is still attached to the engine and you would need the code box and key locks won't need changing if you can swap the key chip.

As for coil issue it sounds like an ecu fault with you're diagnostic research.

As for injection suspend why not? Coils and injectors are based on the same bank fire so makes perfect sense.

slugworth1987 said:

Swapping to an 18f ecu would be do able,but would take a bit of rewiring and a new rad from a cinq due to ecu fan control cam sensor can be ignored if the crank sensor is still attached to the engine and you would need the code box and key locks won't need changing if you can swap the key chip.

As for coil issue it sounds like an ecu fault with you're diagnostic research.

As for injection suspend why not? Coils and injectors are based on the same bank fire so makes perfect sense.

Click to expand...

But the car would almost certainly fail emissions.

One signal it takes, though, from the cam sensor, t'other from the crank sensor. My guess, they're seperate systems.

I was suggesting a swap to the later MPI ECU, which replaced the accursed Micro-crap one in late 2001/early 2002 on Seis (but it looks like the Punto retained the Micro ECU 'till the Mk2 went ourt of production.

Mikexx said:

I apologise for the series of confusing posts. This is indeed a "micro" ECU which has 2 connectors.

Are they easy to take apart? I am an electronic engineer so if possible it is something I can do!!

Click to expand...

This is the MF59 ECU inside.

Damn full sub minature SMD that will be about as much fun as licking a cactus!

cc1 said:

This is the MF59 ECU inside.

Click to expand...

Many thanks for the link to the photo.

Unfortunately the photo is a little blurred, and it's not obvious where the EPROM is.

There seem to be a number of software packages which can re-write the EPROM to virginise it, but wary of buying any without knowing I can get to the EPROM and remove it!