Alfa Pages

[Ben_nz]

My grey 33 has an interesting problem. While idling, a high voltage spark from the ignition system jumps to earth on a regular periodic basis, approximately twice per second.
First it was escaping from the top of the ignition coil to one of the coil's connectors (approximately a 1cm jump). Then I reversed the plug lead from the coil to the centre of the distributor so the larger rubber insulator was on the coil, and now I can hear the spark jumping inside the distributor instead. When the spark is doing its jumping thing, the rev counter needle wavers up around 1500rpm. If the revs are raised slightly off idle, the spark stops jumping and the rev counter drops down and starts working accurately.

I replaced all the spark plugs recently and the problem still exists. Rotor and cap look to be in good condition. I disconnected plug leads one at a time and the frequency of the spark jumping doubles with a lead off. I borrowed a lead from my other car and used it in place of each lead in turn, but this made no difference. One of the leads is a different brand than the other four, and I don't know whether there are multiple types of lead and which I should be using.

Anyone got any ideas why the spark is doing this? Some weird side effect of a worn distributor or loose wire?
_________________
"Now, all that is left to do ... is wait that something else brokens, to fix up again!"

[ZeNiTh-PbArM]

Hi, would your 33 be fuel-injected? if yes chances are the dizzy is out of alignment and at idle under min advance the rotor arm is not facing the cap's contacts. Try rotating the dizzy a few degrees and see if the situation improves.
Other possibilities are faulty plug or plug wire, or faulty/non matched rotor & cap assembly. Don't hesitate to replace the whole lot as it is a regular maintenance item anyway.
regards,
zp

[Ben_nz]

The car is carburetted, a series 2 1.5ti with Weber carbs and the distributor where the cap is held down with two screws.
I notice sometimes when changing up gears that the engine pinks for half a second in the new gear, and this happens without the revs being too low. Perhaps the timing is off under certain circumstances.
At the moment my brother is borrowing the car (ha ha, he thinks I'm doing him a favour but it's the other way around..), but when I get it back I'll try rotating the distributor. I presume non-injected cars would be affected/fixed in the same way as injected ones.
_________________
"Now, all that is left to do ... is wait that something else brokens, to fix up again!"

[ZeNiTh-PbArM]

hi, since this is a carb model don't bother tinkering with dizzy timing. Since the pulse sender and the rotor are on the same dizzy shaft they cannot get out of adjustment unless something's broken inside the dizzy. Injected cars are a different story since TDC sensor is on the crank and dizzy only bears the rotor and cap.
in your case i'd replace every bit of the high voltage system (spark plugs, wires, rotor and cap) as the most likely culprit.
regards,
zp

[Ben_nz]

So I borrowed an oscilloscope from work and here's what I found:

The picture above is the ideal situation. I get this when I raise the revs above idle.
There are two waveforms on the screen - the top one was recorded on the.. um.. negative / switched terminal of the primary / LT side of the ignition coil, and the bottom one was recorded at the ignition module (?) near the coil and shows the position sensor signal from the distributor.
My understanding is a pulse from the distributor position sensor (lower waveform) causes the ignition module to switch the coil (upper waveform) and create a spark.
But now look at this:

This is at idle. Noise on the lower waveform fools the ignition module and it switches the coil too much, creating extra sparks that try to escape. Currently they escape inside the distributor - you can hear them jumping. Also, the extra switching fools the rev counter that the engine is running up to twice as fast as it actually is.

Any new ideas in light of the pictures?
Unfortunately my two 33s have different ignition systems (Magneti Marelli coil in this 1.5ti, Bosch coil in the 1.7QV) so I can't swap parts.
The car with this problem doesn't cruise smoothly and only runs well when thrashed, also it often pinks briefly when a new higher gear is selected at engine speeds where this should be OK.
_________________
"Now, all that is left to do ... is wait that something else brokens, to fix up again!"

[ZeNiTh-PbArM]

Hi,

you are right, the VR sensor voltage getting positive arms a trigger then the zero-crossing detector fires the coil.
So the spike you see is indeed causing an extra spark, now the question is where does it come from? This is hard to tell without further investigation but i see two likely culprits around here.
First option is, the sensor is actually sending the parasitic pulses to the amplifier, due to mechanical play or wobble on the shaft.
Second option would be an EMC issue, the sensor is not sending the pulse but the sensor & wire act as an antenna and pick noise up. I suppose every connexion in your ignition system has already been cleaned up and tightened. Can you make sure the sensor wire is not routed between spark plug wires, and that shielding of the wiring is not broken? If in doubt, use aluminum paper from the kitchen (don't tell the wife!) to make a wrap-on shield and ground the shield at both ends and see if the problem goes away.

regards,
zp

[Ben_nz]

[ZeNiTh-PbArM]

Hi,

ok so this is very likely to be an emc issue, if the alternator generates noise and ignition module does not have sufficient rejection ratio, the alternator spikes go through the internal supply and are picked from the sensor interface and cause false triggering.

On the alternator there is a spike killer, it's the little square black unit (basically a condensator) on the picture below (bosch alternator) :


You might want to get the regulator out to check for brush wear ; if the brush are badly worn, the contact surface between the brush and the rotor is degraded and it generates a lot of RF noise.

regards,
zp

[Ben_nz]

I tried swapping the square black spike killers between my cars and that wasn't the problem.
I checked the alternator brushes and they seemed long enough to be making contact.
I had to get the alternator out to remove the regulator because the screws were tight. While I had the drive belt removed I thought I'd run the engine briefly to check how loud it was, because I believe the water pump in this car is often noisy. And when I did that, I disproved the alternator theory! Unlike the other day, the spark problem was present even with the alternator sitting on the driveway.

So I put the alternator back in and started the car again. As is often the case with this problem, it reduced as the engine got warmer. To try to work out whether this was due to idle speed increasing as the engine warmed, I was amusing myself by running every electrical accessory on max to load up the alternator and slow the idle. This had no effect, infact the rev counter had almost permanently settled down.
So I turned all the accessories off and sat there watching the temperature gauge cool down as the engine fan ran.... and before my eyes, the rev counter started wavering around right up to 3~4000rpm
Then the engine stopped
And now it won't run at all
The rev counter doesn't move while the engine cranks over on the starter motor.

Tomorrow I will take my other Alfa to work to borrow the oscilloscope again and work out what has gone wrong.
_________________
"Now, all that is left to do ... is wait that something else brokens, to fix up again!"

[Ben_nz]

It appears the problem was a broken wire to the sensor inside the distributor. The wires are too long between the sensor and the exit point and can stick out the edge and get pinched by the distributor cap.
I tried to fix it even though the wire is made up of very fine strands, possibly to carry high frequency - my joint may have become a rev limiter? Unfortunately the wire then broke right next to the sensor. I'll have to get an intact sensor and cable from a wrecker.
_________________
"Now, all that is left to do ... is wait that something else brokens, to fix up again!"

[BigAl]

glad you got there Ben, well done
_________________
x2 33 16v
pictures http://www.alfa-pages.co.uk/viewtopic.php?t=5203

[Ben_nz]

I might not actually be there. When I replace the distributor sensor the car will run, but will probably still do the double sparking at idle thing. We shall see.
And with all this testing, I have stripped one of the threads in the distributor body that the cap screws go into. Maybe I can just use cable ties..
_________________
"Now, all that is left to do ... is wait that something else brokens, to fix up again!"

[Ben_nz]

Finally got a new pickup for the existing Magneti Marelli distributor - for a while I thought I might have to replace the ignition system with a Bosch one.
I didn't realise you were supposed to set the pickup gap with feeler gauges - couldn't work that out anyways so I just moved it closer until the car started.
Car runs just fine now, no extra sparking or false rev counter readings. =)
_________________
"Now, all that is left to do ... is wait that something else brokens, to fix up again!"

[BigAl]

wahooo!!! you must be feeling great, i love it when you get to the bottom of a problem, just in time for christmas as well.
_________________
x2 33 16v
pictures http://www.alfa-pages.co.uk/viewtopic.php?t=5203