Hi all,

First time here so please be nice!

My girlfriend's N reg 1.3 Felicia (GLXi - i think!) is stuttering.

It seems to sound a bit uneven when idling, and judders noticeably when accellerating throughout all the gears, once she reachers a decent (60+mph) speed and it cruising along it seems to stop, though I don't know if it's just no longer noticeable.

It also appears to be a little intermittent, the problem will just disappear for ten minutes at a time.

I've changed the fuel filter (seemed like a logical idea at the time anyway!) and used injector cleaner - seemed to have no effect.

Any ideas where we could start? Lamda sensor, Coil Pack, HT Leads, or something else entirely?

any ideas how I can diagnose the problem?

If it makes any difference it's one of the ones with the 4 spark plugs in the front of the engine, and the leads going to a small round box to the left.. (i've read some posts referring to spark plugs having no leads - just a box).

Thanks in advance for your help!

Andrew

The Felicia is a mans car tell the girlfriend to get a Micra :P LOL

When you say "judders" do you mean the steering wheel?

My sisters Micra started developing steering wheel wobble which anished when the tyres got changed.

Other guesses are prob the timing chain or engne mounts

This is an odd one; I'm not used to misfires clearing up at higher engine speeds.

Usual causes of a misfire on these are bad HT lead(s), bad spark plug(s), bad coil pack (the box thing, which contains 2 double-ended ignition coils, firing both the cylinders which are at TDC, on the "lost spark" method, and the TDC sensor, which is located on the top of the bellhousing on these cars.

If by juddering you mean it feels like your riding a kangaroo rather than a car, my problem was a faulty lead to the lamda sensor. It would clear up at high speed and sometimes I'd have days when there was virtually no trouble at all, worst days were when it rained oddly enough.

I think it clears up at high speed due to the ecu ignoring the lamda sensor as I think it makes the mixture intentionally rich ..... but I'm no mechanic so don't take my word for it.

Rob, the reason your problem was worse on wet days would be that water was getting into the cable, and degrading the signal.

I think you might be onto something, but not so much because the mixture varies with engine speed (below full throttle, when it's allowed to go rich for extra power) as because the extra gas flow over the sensor stabilises the reading.

Either way, the OP should get under the car, find a wire attached to a plug on the exhaust pipe, and inspect it for damaged insulation.

I'm just reporting what she's been telling me to be honest, so i'll take it for a run and write back, see if the juddering really does disappear at higher engine speeds - but i'll check the Lambda sensor, how would i tell if it's okay or not? it's supposed to output one of two or three voltages - am i right? If i was to connect a multimeter - would this show if it worked, or if not - how can i find out if its failed?

Just to clarify it is kangaroo stylee, not a vibration or anything.

Thnaks Andrew

From what Rob said, you're looking for damaged wiring attached to the lambda sensor, not testing the sensor itself.

I had mine done by a garage as I couldn't get hold of a vagcom and there were too many possibilities. Firstly I would use your multimeter to continuity check the leads, in the continuity mode of your multimeter it should beep if the two probes touch together to indicate continuity, test each wire to the lambda probe (disconnected if possible) also check for cross continuity (ie two wires touching that shouldn't be). I imagine if your multimeter is digital it will not give you a very good way of measuring the actual signal as it flicks between 0.2 and 0.8V ish faster than your multimeter will pick up I would expect you to see 0.5 constantly or something. If you could get hold of an oscilloscope you'd be able to use that. Alternatively if the continuity of the leads is ok try a vagcom ?

EDIT : Damaged wiring is not always visible, often fractures occur from within.

Okay, i'll do that this weekend.

Once i've checked the leads and lamda sensor (lead) as best i can, would i be then best to look into the HT leads?

Does anyone know of a decent breakers in norfolk / suffolk that i could get leads from to try, surely i'd only be looking at a few quid? Are there any other models / makes of cars that I could harvest the leads from?

Would my girlfriends car have vagcom ports - I didn't think it would for some reason? How is best to read the info from it, can it just be done by reading the raw serial data connected to a PC and intrepeted by software, or do you need a specific device to read the info?

If it has separate HT leads, then as long as you change one at a time, a new set for about £10 (I think; I've had diesels the last 10 years) makes sense.

I'd also think this car will be too old for the USD standard so no VAG-Com.

Sorry, I'm not sure when VAGCOM actually kicked in just assumed it was standard in all Felicia's.

Asking a genuine question, if one of the HT leads was dodgy and then only firing on 3 cylinders or intermittently a 4th would that not make it still a very noticeable problem at 60mph ?

I'd usually expect an HT misfire to get worse with increasing engine speed; I just figure that if HTs are more than a couple of years old it's a cheap and easy fix that usually improves running even if it's not the actual problem.

These engines can be funny old things.

I had the same problems quite a while ago now. It was juddering and hesitating a lot but not at higher engine revs..

After lots of fiddling around lining up the timing marks on the distributor and cleaning air filters etc and the problem not being as bad but still present I finally changed the spark plugs and HT leads, air filter, oil, oil filter and fuel filter and the problem went away.

Also I think things like valve clearances are often over looked on these cars. Get your valve clearances checked and done. Just done mine this weekend and what a difference it makes to the running of the engine.

BTW my 1996 'N' 1.3 SPi has a VAGCOM interface under the steering collumn. I think all 96 fellys shoud have it.

Phil

The lambda sensor wire seems to be in a decent state, there was continuity between two of the wires, but this was on both sides (lambda sensor and connector going to ECU i think) - so i guess this is not a problem.

Is it wise for me to focus on the lamda sensor now, or something else? or is this a garage jobby?

Air filter is okay, Fuel filter changed, Sparks look fine.

HT leads i doubted as it seemed to get better at higher speeds (though i know these things are more like witchcraft).

Valve clearances - is this a garage jobby, how much should i be looking at to get those done - I take it valve clearances on their own couldn't cause severe kangaroo-ing?

And I found teh OBD-II - I guess vagcom? connector. 16 pin connector to the right under the steering wheel. Is this worth spending time looking into?

Valve clearances are olde-worlde locknut tappets, so eminently DIYable. They shouldn't cause kangarooing, but it's another cheap elimination.

If you have a vagcom port it should show up a problem with the lambda sensor or leads though it won't tell you which is the problem. Vagcom should be your next port of call in my opinion or you could end up doing lots of unnecessary work. I think you can try running the engine without the lamda sensor plugged in to see if the issue still occurs or if you get new issues etc. Based on the fact that at high speed the problem disappears I'm inclined to keep focus on the lambda until its checked by a vagcom or tested by an oscilloscope at the ECU end of the cable.

Other things I could envisage it being is timing slightly out, or a very minor pressure leak somewhere as if you saturate pressure leakage at high speed the car might not noticeably under perform and timing again might not be noticeably out at high speed if its very minor. Another thing to throw out there may be clutch related issues but all of these are a lot less likely than the lambda in my opinion.