Dan Shea

Hello All,

If it's not one thing it's another, but I guess this should be expected with a 28yr old car…

So I just finished changing the O2 sensor and spark plugs because I was getting ~15 mpg calculated between fill ups. This was completed Wednesday, and since then I've put about 70 miles on the car and it was running fine and MPG appear to be back into the low 20's. (MPG gauge works intermittently, may or may not be related).

Friday I went to start the car and it turned over very slowly and wouldn't start, quickly assumed battery was dead. I Hooked up a battery charger, waited for it to complete, and tried to fire the motor... nothing. Then I tried using jumper cables, and it will spin fast enough but not fire.

I'm getting 50psi of fuel pressure at the rail, (within spec for the 16v) but no spark and I'm not getting tach bounce or spark (pulled a plug and grounded it to see if it would spark). I tested the resistance of the speed sensor and I get 1.01k Ohms between 8 and 27, and infinite (not conductive) between 8 and 23.

I’m hesitant to jump and spend the money to replace the speed sensor if it could possibly be something else <cough >ICM <cough>.

Testing done so far:
Fuel Gauge at the Rail: 50-55psi
Speed Sensor Resistance: ~1k Ohms between 8 & 27, ∞ Ohms between 8 & 23
No Spark
No Tach Bounce

Are there any other tests I could complete without an osciliscope?

mytrplseven

DME relay?

Dan Shea

Yeah I tried it with it jumpered to no avail. I figured it was ok because the fuel pump is kicking on just no spark.

MB968

If you are getting the correct resistance numbers for the speed sensor ( I assume you mean the one that reads the position of the crank 60-2 tooth wheel on the flywheel) then I'd look somewhere else like connectors/wiring, etc. to make sure the correct signals are getting to/from the sensor. The sensor is a very simple electrical device, and usually OK if the resistence #'s are OK.

Rogue_Ant

The S&R sensors and wiring are the single most important sensor in the car. Without proper operation the DME will not know correct engine position, and can fail to initially 'synch-up' resulting in a no-start, no-run, or intermittent-stalling issue situation...


While the sensors (variable reluctance) are a simple device, resistance #'s won't tell you much other than an explicit short or open-circuit. The signal from the sensors is an alternating-current signal. The DME looks for the zero-crossing caused by a tooth passing by the sensor. If the sensor is bad, or the wiring is suspect, there can be noise on this signal which might prevent the DME from properly identifying the zero-point crossing - resulting in an improper synch or complete failure to identify crank-position.


The 951 suffers from this more commonly than the 944 due to the added engine-bay heat. But, at this point in the 951/944 life, if the S&R sensors and wiring are original, then it is time to replace both. Wiring corrodes underneath the sheath, and might look good from the outside, but can still cause issues.

mhariush

I had the same issue with my 16V, it was the ICU. It's worth trying it... I have a used one that works if you're interested. Already heat sinked to a plate so you can just bolt it in...

I also had an issue where the DME (computer) died...

harveyf

I'm sure it has occurred to you that you may have bumped the sensor wiring when you were in there for your O2 sensor and initiated a problem.

When you check your sensor resistance, do so at the plug for the DME. That way you are checking the entire wiring path.

I don't have a scope either but I do have a Fluke automotive meter. I have had some luck using the function that captures the peak voltage to see if the sensor is working. So I crank the car while measuring the AC voltage. You won't see any significant voltage in the normal AC voltage mode but if you set the meter on the highest sample rate, I can usually catch a peak of around 2 volts, for a properly working sensor. Rogue Ant can probably fill in upon my missing knowledge here. Of course this assumes you can borrow a meter with this capability.

I snagged 4 of these sensors from Rock Auto for about $35 each, just to have them around. Do a search on the sensors and find the thread where a picture of the box is in the post. Rock Auto says these are a factory close-out item so no idea how long this screaming deal will last.

I am fortunate to have a spare known good DME that I plug in as a test for no starts. That was actually the answer once, but usually it isn't.

My wife hates to look at them but buying a parts car of the same model has been invaluable for me.

MB968

-I know that the resistance check is not an absolute check. But if you don't get the right resistance for both readings, you can assume the unit is bad. But even if you get the right numbers the sensor still could be an issue.

-Also, I agree that you should test the resistance/signal at the DME to include everything in between. If it doesn't look good there, you can walk back the line to see if it is at least good at the sensor.

-I've not done this with the crank sensor, but I've successfully tested ABS sensors (which are the same basis sensor) with an old style needle type (analog) volt/ohm meter. When rotating the wheel slowly I'd get a 0-5 volt type sign (or maybe it was a square) wave. Haven't had need to test the crank sensor, but thought if you just pumped the starter so the starter never got up to speed you might be able to see the characteristic change in voltage associated with this type of sensor. I know at idle speed, the frequency of the wave would be too high for an analog meter and you'd need a scope. But, jogging it may work.

Rogue_Ant

Many ABS sensors (and pickup sensors in general) are Hall-effect, which are powered and produce a logic-level square wave. This type of sensor can be tested using the method you describe.
However, the VR crank sensors used for our cars are not powered, and produce a non-consistent voltage amplitude 'sine wave'. The output voltage amplitude depends on both the air gap distance and the speed of the passing tooth.
For example, at cranking speeds the sensor might output 2 volts peak-to-peak (or less if the gap is too large). At 3000rpm, they could be producing 50 volts peak-to-peak.
The slower the crank spins, the less voltage produced.
An oscilloscope is really needed to test it... The good thing is that with new wiring & sensors, it should be quite reliable - after all the stock stuff most likely lasted well over 20 years without problem.

white924s

If the tach isn't bouncing and the DME relay is jumpered, then the problem is most likely in either the speed/ref sensors or the wiring. Both can go bad over 20+ years and lead to a no start condition much like what you described. In particular, the speed sensor feeds the signal that the tach uses to know where to point, so if the tach isn't bouncing, that means it is not getting the signal. I am fairly certain that the sensor is wired directly to the tach too - if the DME dies, the tach will bounce but the car won't start.

One other test that you can do is confirm that the injectors are firing. The easiest way to do this is to pull the fuel rail and point the injectors to a large bucket while cranking the car. If nothing comes out, I think that would support the conclusion that the DME isn't getting a good signal from the sensors. If you have a known-good DME, you can also try plugging that in to see if it makes a difference. I think it is more common for the speed/ref sensors on these cars to fail than it is for the DME to fail, but both are known to happen.

Rogue_Ant

Exactly. That, or the DME itself.

No - the tach is driven from the DME ignition circuitry. If the tach isn't bouncing, then the DME is not trying to fire the car.

william_b_noble

let me point out that the 16V has a hall effect sensor in the distributor and an ignition control module on the left side of the car behind the headlight bucket - I would suspect that control module, based on prior experience with an S

Rogue_Ant

Yes, the 16valve cars are different...

Dan Shea - is your car an 8v or 16v swapped setup? You stated the car is 28years old, so it should be an '85 - which are 8 valve engines. Also your testing DME pins 8 & 27, and 8 & 23 suggest you are testing for an 8v setup.

white924s

My mistake. When the DME in my car failed, I still got a tach bounce when I tried cranking it, so I assumed the tach pulled the signal directly from the sensor. Swapping in a known good DME fixed my problem, so I know that it was a DME failure

mhariush

Like William said, and what I tired to say in the post before. You should try and swap the ICU/Final Stage Module, Bosch, located on the front of the wheel well behind the headlight.

It was a culprit on my S once.