1972 erratic / no spark

[rattlsnak]

Hey all, I have a '72 911 with all stock ignition parts. Car has been sitting for 20+ years. I have cleaned the fuel system (new lines, filter, went through MFI pump, etc..) and got it running but it only runs for 2-3 minutes and then loses spark.

So what it does is this:.. It will crank up and run and then all of sudden shut off after about 2-3 minutes, if that. If I wait around 10-15 minutes it will start back up and then shut off again. I can pull the coil wire off and see no spark when cranking or sometimes just one or two sparks erratically. I can pull the dist cap off and use a plastic stick and move the points and it sparks perfectly. I hook everything back up, it cranks right up and the cycle repeats. I had another coil from another car and so just swapped that out and no change at all. Then I checked and readjusted the point gap and then it ran for a solid 10-15 minutes or so and then shut off and now it is back to where it was before. The CDI box connections and all the wiring checks good and the box does whine like it should. I have a new set of points coming just to put that out of my mind but kind of thinking it is CDI related.

Any thoughts? I have the test procedures but by time i go to check it, it has 'cured itself' so everything checks OK.

Now, that being said... this car has MFI and has a problem with the shutoff solenoid staying energized at all times. That signal comes from a RPM sensor relay box mounted on the relay panel. That same RPM signal comes from the tach wire (which is hooked to the points) and is also hooked into the CDI box. That solenoid is only supposed to be energized at closed throttle and above 1500 RPMs so maybe that RPM signal is wrong and feeding back into the CDI box in some way?? That could be far fetched and not at all even related but it is something I noticed that might be worthy of mentioning.

[kltarga72]

A few years ago I had the same exact problem, car sat for seven years. I could drive the car around the block and most of the time it would make it and it ran great. Then it would sputter and quit, let it set for two minutes and I could drive it back to the garage.

The problem was the filter in the gas tank, it was so packed with debris gas would not pass through it. The fuel pump will suck gas from the tank to the filter inside the engine compartment and to the MFI. When the car sits some of the debris will flow away from the fuel tank filter. Your car will start up and run fine but the fuel pump will pull the debris back to the filter and will clog.

There are two fuel hoses at the bottom of the tank (rear) and one will have a brass nut connection, this is the filter. You simply disconnect the hose remove brass nut with wrench and the filter will come out. When you see the filter you will understand why it would run good then quit.

The filter is a screen with a very fine paper material covering the screen. I think a new filter was $50.00

Hope this helps

When testing only put a gallon or two of gas back in the tank.

Lash

[rattlsnak]

The fuel system has been totally cleaned and checked, including the MFI pump. (gas tank was sent out and boiled, filters replaced, etc..) When this happens, its clearly a no spark situation.

[kltarga72]

Looked at early 911 Registry to see if anyone had a similar problem, no luck.
If you are loosing spark, then with everything that you have checked it does look like the CDI.
I had to have my CDI rebuilt.

If you haven't already make continuity checks on the ignition wires from CDI to distributor etc. and check ALL grounds .

When the engine is running are you able to increase rpms?
If the engine will run (idle) for a few minutes, try to wiggle or pull slightly on the tach wire etc. just to see if the problem is a loose connection. I would do this with an insulated glove.

Dont forget that loose battery grounds can cause some very erratic head scratching problems.

Lash

[Porboynz]

I would disconnect the RPM relay and the tacho in case they are involved. The car will run fine without the relay albeit with a few backfires on deceleration. Maybe even consider running a temporary wire from the points to the CDI trigger terminal. Check the distributor earth wire under the points plate while you are at it. Also check the earth stud behind the fuel filter and make sure all connections are good, including the fuel console earth. It is interesting that you can promote a spark by manually opening the points, it does not completely clear the CDI but it does sound promising. (The CDI could have some sort of heat related failure that stops it producing enough multiple sparks to run the engine, but works well enough to produce a few manually generated sparks.) At least the fault is repeatable, keep eliminating components and you will solve it.

[Jonny H]

^. Very good advice.

If the RPM fuel cut off circuit was damaged it could potentially pull the distrbutor signal to ground or at least interfere with it in some way.

I have also seen a situation where the black signal wire had lost its insulation and was occasioanally shorting on the oil line behind the distributor.

Having said that, We see lots of CDI boxes that will run a car, but they are not functioning properly. Reduced output voltage and intermittent triggering are fairly common.

We are working on a firmware upgrade to our CDI+ boxes which will allow them to control the fuel cut off, bypassing the original temperamental circuit.

[rwest]

I would be suspicious of the CDI box as well and as you have found out the intermittent ones can be tough to trouble shoot- I started to carry around a timing light and an oscilloscope so I could finally verify it was bad.

Mine had the same symptom: it would work fine then after some running and me shutting it off, it would start right up again and then a few minutes later die. I figured some component would heat up and then open or short, once it cooled down it would work again and so on.

[al lkosmal]

The fact that you reset your points and the problem went away for 15 minutes and spark is there when you actuate the points with the plastic stick, would lead me to focus on the points 1st. it may indicate that your points gap is changing/closing after your adjustment somehow....have you checked the gap both before (running good) and after (not running) to verify that the gap/points are still adjusted ok and/or that the points are clean and can make good contact? It's a simple check, might as well eliminate the inexpensive, yet critical, point contact closure/gap 1st.

regards,
al

[rattlsnak]

Thanks all for the input and advice. I picked up a new set of points and will put them in Monday morning. It was a cheap $8.00 and will put that out of the equation. If it still does it, I will pull the tach signal or bypass the RPM sensing relay, after that, it only leaves the CDI box.. I will report back soon..

[rattlsnak]

Put the new points in today and far as i can tell, it is fixed. It ran for a solid 20 minutes today a few different times with no issues.. Keeping fingers crossed!

[wwest]

Originally Posted by rattlsnak

Put the new points in today and far as i can tell, it is fixed. It ran for a solid 20 minutes today a few different times with no issues.. Keeping fingers crossed!

My guess would be that ignition point "chatter" (common) was causing sparks close enough together to trigger the rev limit module, compounded by the battery overcharge.

Fly in the ointment is that with battery overcharge the rev limiter latches up rather than just a one time event.

Turn the key OFF, count to 10 slowly, long enough to fully depower the rev limiter module's capacitors, restart.

[Porboynz]

My guess is oxidised points, what is this rev limiter thing you speak of? On my 72 there is a rather simple spring loaded over rev cut-out in the rotor (about 7000rpm). The only rev related electronics is the MFI over run cut-out that is electrically disconnected at idle by a micro switch. I love it when its a simple fix, new points and voila!

[mysocal911]

Originally Posted by Porboynz

My guess is oxidised points, what is this rev limiter thing you speak of? On my 72 there is a rather simple spring loaded over rev cut-out in the rotor (about 7000rpm). The only rev related electronics is the MFI over run cut-out that is electrically disconnected at idle by a micro switch. I love it when its a simple fix, new points and voila!

Right, no electronic rev limiter!

[rattlsnak]

yes, and I put a 'normal' rotor on it long ago!