How to test spark coil? (308)

Growing up in TX, we used to go drinking in an odd corner of somebody's large pasture. Occasionally we'd have a newbie along, & after a few we'd go 'visit the bushes'. Those in the know knew where to aim, yes, there was a fence in the bushes...
;^)

 

I'm having trouble following.

I believe you have said that you've established that BOTH coils are cutting out, just not both at the same time. This was done by:

Replacing BACK coil early in this thread.

a) post #23: Pulling the FRONT coil's yellow wire & car kept running on the BACK coil -> FRONT coil isn't firing.

b) post #24: Replaced FRONT coil w/former rear coil. Still ended up running on 1 bank. Pulling the FRONT coil's yellow wire & car died -> BACK coil isn't firing.

IMHO, you've established:
- It's heat sensitive, only occurrs after a run.
- It's probably not a coil problem.
- Problem can affect either bank.
-> Look for things common to both banks such as:
. ECU grounding
. ECU power (fuse or fuse block? wiring?)
. motor speed (Flywheel tooth) sensor?

I suspect that the backfiring & loud exhaust are either due to raw fuel being dumped into the exhaust when a bank starts cutting out, or else you have an unrelated lean mixture problem.

Umm, let me think some more about what the lambda system will do when it detects raw gas from the bank that's cut out. It may try to correct the problem by lowering the K-Jet distributor's output pressure, leaving the bank that's still got ignition running lean.

 

yup! If you want to prove that, disconnect the O2 sensor. It's a single wire plug tucked up on a frame rail underneath the air filter box and to the right of the fuel distributor. The wire is probably green, and the connector is probably white.

 

verell's comment about lambda.

if you are getting unburned fuel into the exhaust, the O2 sensor will think the engine is running rich. The injection ecu will compensate by changing the frequency valve signal to lean out the mixture, which will cause the bank that is still running to go lean.

by unplugging the O2 sensor, the injection ecu will no longer try and fiddle with the mixture, so the running bank will revert to getting whatever the base fuel mixture is as set by the adjustment screw in the fuel distributor.

All that disconnecting the O2 sensor should do is get rid of any lean running symptoms, which are just a byproduct of the misfiring bank. The real problem still appears to be electrical in the ignition system.

'course, this whole rathole is only relevant if you car has the lambda system. You have a frequency valve happily buzzing on top of the fuel distributor?

 

right on the fuel distributor. The frequency valve would be sitting kinda on top of it on the left side. see pic below from a 328 that sorta shows what it looks like. It's the same on the 308.

it's standard for us spec cars. I don't know if your euro has it, since it's primarily emissions related. Second pic is the wiring from the o2 sensor to the car. If you don't have the frequency valve, you won't have any use for an o2 sensor.

The fuel injection ecu watches the o2 sensor output and varies the signal to the frequency valve. The frequency valve adjusts a control fuel pressure in the distribution head, which changes the amount of fuel squirted by the injectors.

The injection ecu can lean or richen the mixture in response to what the o2 sensor is reporting. If you dump unburned fuel into the exhaust, the o2 sensor output will go to about 1V, and the ecu will try and compensate by leaning the mixture as much as it can via the frequency valve. There's a limit to what the frequency valve can do, so while it may be pegged as lean as the fv can make it, the car will probably still run.

You don't want to just unplug the fv, as that would also peg the mixture as lean as the fv can go. If you unplug the o2 sensor, though, the injection ecu will default the fv to the middle of it's operating range
Image Unavailable, Please Login
Image Unavailable, Please Login

 

something doesn't sound right....your car has the little yellow jumper wire between the two coils, right? If so, there is one yellow wire coming from the fuse box to coil A, then the yellow jumper goes from coil A to coil B, and then another yellow wire goes from coil B to the diag connector. If there was a problem on the yellow wire from the fuse box to the coil, both coils would lose the +12V and the car would instantly die. It's also not really possible for the yellow wire to be tied directly to ground....the fuse would blow. If for some reason the wiring was misconnected so there was no fuse, the wire would become a toaster element until something melted/burned up enough to break the connection.

how are you checking continuity? If you just stick an ohmeter between the +12V and gnd, you'll probably see very little resistance. You aren't supposed to measure powered circuits with an ohmeter.

the wire from the digiplex to the coil is grounding the coil terminal through a transistor inside the digiplex. The 12V you see on that wire is not coming from the digiplex, it's coming from the yellow wire through the coil whenever the digiplex is not grounding the wire. In other words, when the digiplex is not grounding the wire to fire the coil, you will measure 12V on both sides of the coil.

in any case, it doesn't sound like the problem is the coils. As verell said, I'd focus on the digiplexes and the plugs between the digiplexes and the coils/sensors.

have you swapped the digiplexes yet? All the have to do is swap the plugs between them.

Have you checked the plugs between the digiplexes and the coils/sensors yet?

if you connect a voltmeter to a non-yellow coil wire, make sure the meter is set for measuring AC voltage, and you should measure around 2.8-3.5V when the car is idling...at least, those are the numbers for a 328. I'm assuming the 308 is about the same, but possibly not since the 328 ecu is talking to an electronic module on the coils instead of grounding the coils directly. In any case, it should be lots less than 12V.

you can also measure the sensors. The tdc sensors should be >0.2V, and the tach sensor should be >2.0V.

if your diag connector diagram is accurate, you can measure all those things via diag connector pins. Hopefully the problem sticks around long enough to make the measurements.

 

I love point ignitions and carburators.

 

WolfTalk- have you or anyone else here made a diagram Labeling the Diag pins with what voltages, wavefoms etc to be expected at each pin ?
(seems like it would be a quick way to narrow down problems at one convienient location)

 

FWIW, my Lucas coil lasted 30 years. Then starting the car got progressively harder until it died. I bought a new Lucas Sport coil which is the gold one, and it's internally ballasted. Believe it or not, it's the only thing I'm aware of that Lucas makes that is superior quality. That thing has super high output and is especially good on old carb cars to help prevent fouling.

Ken

 

if the diagram in the US '84 owners manual is correct, then this is what you should see:

- 12V between pins 7 and 9 when key is in run position
- around 3V AC between pins 1 and 9 at idle
- around 3V AC between pins 2 and 9 at idle
- >0.2V AC between pins 3 and 4 at idle, around 700 ohms between pins when not running
- >0.2V AC between pins 5 and 6 at idle, around 700 ohms between pins when not running

since pin 9 is ground, you can tap gnd off the frame instead. Having gnd in the connector is only useful if you have a plug to stick into the socket.
Also, I lied about the tach sensor. It's not connected to the diag socket on the 308.

If you have a scope, then you'd be looking at pulses on the pins 1, 2, 3/4 and 5/6 above instead of AC voltages.

problem is how the pins are numbered. It appears that ferrari numbered the plug on the diagram in "ignition system components" in the manual, looking into the pins in the plug. On the schematic, the numbering would be for the pins in the socket, so the numbering is mirrored.

since we just care about the socket in the car, it looks like the pins are numbered like in the diagram below. Someone should confirm, though.
Image Unavailable, Please Login

 

My Dino was running well with a slightly bit more backfiring on deceleration so I checked all the high tension/spark plug wires. I'm using the MSD 6200 unit and MSD coil and upon pulling/pushing the coil wire it came out along with the screw in post adapter which was coated in black liquid (oil?). I cleaned and replaced, car still ran well.

What is the failure mode of a coil? (dies, limps or...?)
If all is running well, how do you know when to replace the coil?

Thank you

 

The car got harder to sart: more cranking before it would catch. First it would take a second, then two or three, then even longer. When it got to the point it was hit or miss that it would start (and I was discharging the battery a ton every time), I started replacing the 30 year old stuff: rotor, distributor cap, wires....nothing helped. I finally replaced the coil and wow...cranked instantaneously. Just like a real car! Now one caveat: the old coil was ballasted and the new one is internally ballasted, so I naturally bypassed the ballast resistor. It never occurred to me that the coil might have been fine and the resistor was no good! LOL But It wasn't expensive and the things do die. If your car starts and runs fine, I wouldn't replace it just for fun. It's not like a timing belt that you want to prempt a failure on.

Oh, and the new coil made for less fouling too; much stronger spark.

Ken

 

wolftalk - Thanks ! Excellent and very clear-well labeled. This should speed up troubleshooting. I do have a scope and would love to take readings periodically to look for changes(potential problems). Now I can.
Great job !!!

 

wow!

I can't give you an answer on the main question, which is whether the sensors drop in resistance as they get hotter. I'd expect the opposite, actually. To find out, you've got a couple options.

1] repeat the performance until you get the low resistance, then disconnect and measure sensor resistance at the two pin plug that is near a sensor to see if you get the same value

2] remove a sensor and heat it with a hairdryer or something while measuring the resistance

since both TDC sensors are doing the same thing, and you've got a bouncing tach and the rpms are dropping (even if you lose a bank, the idle shouldn't change that much if the remaining bank was firing ok), I'm thinking something is wrong at the main plug between the ecu's/diag connector and the sensors.

The schematic shows a 6 pin plug someplace, but I don't remember where it is. I'd guess it's somewhere around the top of the driver side rear shock. Should be easy to follow the sensor wire bundle back to it.

If you haven't already, find that connector, unplug it, and make sure nobody sprayed the pins with something nasty. If there's any evidence of stuff on the pins, hose it down with an electrical cleaner or denatured alcohol. Avoid any cleaner that has a lubricant or things like wd-40.

While you've got the plug apart, you can also measure the sensor resistances from there. I dunno what the plug physically looks like, but the US schematic says that sensors are connected to it like:
pins 3&6 - tach sensor
pins 2&5 - tdc sensor for 1-4 bank
pins 1&4 - tdc sensor for 5-8 bank

as a hint, the tdc sensor wires only have a single wire on the pins on both sides of the plug, and it looks like the wire on pin 1 is yellow, the wire on pin 4 is black.

another hint that you'd have conductivity where there shouldn't be any is if you get a non-infinite resistance between the sensors. You'll need to unplug the ecu's for this, and then measure between unrelated pins. For example, between pin 1's on both ecu connectors.

on the 328, the sensor plug is a little round thing that has no "walls" between some of the pins. If someone left a heat sensitive conductive residue inside the connector, some of the pins could be partially shorted together.

The behavior you saw when shorting pins in the diag connector I'd ignore. The sensors are generating pulses on the wires, and pin 2 on the diag connector is also getting pulses from the ecu. Shorting pin 2 and 3 would cause the 5-8 bank coil firing pulse to be fed into the 1-4 bank TDC sensor input. What happens I guess depends on the timing...pretty unsafe thing to happen.

 

if the low resistance in the sensors/sensor wires is not normal, the signals the
sensors are generating could be degraded to the point that the ecu can't detect them or they are so noisy they don't make sense.

in this situation, everything depends on how the ecu interpretes the inputs, e.g. if it's trying to sanity check, average, filter noise, etc. For example, if you start the car when the sensors signals are poor, the ecu may be able to sync to them until some errant spike confuses the ecu, and it either shuts down or is firing the coil at the wrong time.

hmmm....that's not my experience. When I've seen 3x8's with a bank that is out, they still run and idle ok. Losing a cylinder or a bank affects the rpms slightly, and you can hear the difference, but it's not like the you lose 125 rpm for each cylinder that is lost.

if it's not the sensors, I think you only have a couple things left to look at. You can measure the AC voltages as mentioned earlier if you have a better meter (any cheap digital meter should be fine), or use a timing light/spark tester on the plug/coils wires to see if there's an irregular firing pattern.

If still no answer, you'll probably need to find someone with a scope so you can take a look at the signals. Unless they work on fcars, you'll have to tell them what the pins in the diag connector are for, and all they need to do is display the waveforms on them. Maybe we can bribe spiderseeker to hook up his scope and post some pics of the primary and secondary waveforms.

the thing I'm wondering is if you aren't really losing a bank, but you are misfiring randomly across both banks. Disconnecting a coil stalls the car because you've lost too many cylinders. If this is happening, then the engine could stumble to a stop after disconnecting a coil because there's still a couple cylinders in the other bank firing. A timing light or 1+ of those cheap inline spark testers (like attached to the purple plug wire in the pic previously in this thread) can be used to see if the spark is really missing.

also, you'll have to check the spark plugs for fouling eventually.

 

I don't believe any of the Euro QVs had lambda. Steven Rochelin's is an '85 and I'm almost certain it doesn't. JWise's has an '84 & I know it & my '82 don't have lambda.

I think only the US & possibly CH or JA QVs had Lambda.

My experience matches Wolftalk's, I've lost a bank & idle speed didn't drop by half, just a hundred rpm or so. Engine felt sluggish. from the way it felt, I figured I'd lost a cylinder or 2, was surprised when it was a whole bank.

 

did you find the 6 pin connector? look directly under the shelf-like flat area next to and in front of the coolant tank