1998 Silhouette No Start Condition - Odd Symptoms...

Mark R.
06-28-2017, 12:13 PM
1998 Oldsmobile Silhouette - We have owned it since brand new.
Kids are grown. All seats are out. Used mainly for hauling and moving.

Running great, then died unexpectedly at a stoplight, and will not restart.

Strong battery cranks well. Engine sputters and tries, but won't start.

Died pretty suddenly. Checked for DTC codes, and there are NONE set.

No blown fuses found in either location (side of dash and under hood).

Excellent fuel pressure (300+ kPa). Builds fast at key on and checks well.

Strong spark signal present on all six cylinders while cranking.

It just seems like the injectors are not being commanded to fire...

Any ideas what to check next..? Thanks very much in advance!


06-29-2017, 08:14 AM
Which engine does your Silhouette have? Do you have a noid light set, or access to an oscilloscope to check for injector pulse width? Have you tried spraying some starting fluid or throttle body cleaner in to the intake to see if that will get the engine to start, supporting the lack of injector pulse?


Mark R.
06-29-2017, 10:27 AM
Hi Rod, Thanks for the reply.

It has the 3400 V6 engine (3.4 Liter).

I do not have a noid light or an oscilloscope.

I did try spraying starting fluid into the intake.
The engine started and ran for a few seconds.

So I'm pretty sure it's the injectors not firing.
I just don't know why... Thanks again!


Mark R.
06-29-2017, 10:51 AM
-> Could a failed Camshaft Sensor or a failed Crankshaft Sensor keep the engine from starting?

In one of my search results, I read that with either failure the car should still start, but will run poorly.

Both sensors have dedicated DTC codes that should set when they fail, and I had no codes set.

Just wondering if I can rule out these two sensors...


06-29-2017, 01:41 PM
Usually if those sensors are bad, you won't get spark. Since it will start with starting fluid you have a fuel delivery problem

Mark R.
06-29-2017, 02:29 PM
Yes. And as I mentioned, there is very good fuel pressure present.

So the question is: What would keep the fuel injectors from firing?

What signal or input is needed to fire them, that might be missing..?


06-29-2017, 03:54 PM
Crank sensor, wiring to it, ign module another possibility, see if you have cranking rpm read on a scanner. Bad injector can pull down voltage from the remaining injectors, not an easy diagnosis. Disconnect the maf, see what happens.

Mark R.
06-29-2017, 06:42 PM
I had already tried cranking with the MAF sensor disconnected and that gives no change.

I do see an RPM reading on my scan tool. It bounces around 180-200 RPM while cranking.

When cranking, I get what appear to be valid readings for TPS, RPM, MAF, MAP, ECT, and IAT.

What can I check next...?


06-29-2017, 07:40 PM
See if you have voltage to the injectors, key on.

Mark R.
06-29-2017, 08:59 PM
OK. I will have to check that tomorrow.

I'm not sure how accessible they are. The fuel rail is buried underneath the intake manifold on that engine.

So am I looking for full time voltage with the key on, or just a momentary voltage pulse, or..?


06-29-2017, 10:09 PM
Unfortunately I no longer have access to the GM service manuals to confirm, and most of my experience is with Fords. From what I've seen on Fords, and suspect GM is similar, the injectors receive constant voltage with the key in the Run position (probably battery voltage, but could be 5V) and the ground is switched via driver transistors in the PCM. So you should expect to have constant power and switched ground.


Tech II
06-29-2017, 11:26 PM
It could be his Theft/Security system if the injectors are not firing.....fuel pressure is ok, and he does have spark.....

Does your owner's manual state PassLock, or PassKey I, II, or III?

You don't have to remove the plenum to check for voltage to the injectors, check at the injector connector outside the plenum....terminal D, pink wire, should be hot with key in run position....

06-30-2017, 08:11 AM
^^^ Good point.


Mark R.
06-30-2017, 12:19 PM
The owner's manual is very vague about the security system, with no mention of which type (Passlock, Passkey, etc.).

It simply refers to it as the "theft deterrent alarm system" option, which we do have on this vehicle.

There is NO security indicator light or symbol anywhere on the instrument cluster gauges.

There is only a red LED that sits up on the center top of the dash, and blinks slowly when it is armed.

The manual does state: "The system will also cut off the fuel supply, preventing the vehicle from being driven."

That is literally the only line that even mentions this. There are no other explanations or details...

In the past, when we've set the alarm off, the horn pulses and the lights flash. None of that is happening now.

Both of the keys we have are original from the factory - the ones that came with the vehicle when brand new.

I did go and get the other key to see if it made any difference, but it yields the exact same symptoms...


06-30-2017, 01:04 PM
Pretty sure you have a PK3. If original key it should be stamped on the key. But with no security illuminated at all then I'm not sure what that means.

Tech II
06-30-2017, 02:25 PM
You need to have this vehicle scanned for "B" codes...unfortunately, only high end scanners can do this....the BCM turns on the theft/security light....

Mark R.
06-30-2017, 04:49 PM
With key on, we have 12.5 Volts at the pink wire in terminal D of the injector harness connector.

Had to remove the front struts and rotate the engine forward just to get access to that connector...


Mark R.
06-30-2017, 06:18 PM
OK, a bit of a new development here (possibly a setback).

While I was out today, I picked up a can of starting fluid to confirm what my son had told me.

He and his buddy had tried a few things, before paying $120 to send it home on a flatbed.

They had said that they were able to get it to fire using a can of carb cleaner.

I'm not sure if something has now changed, BUT, I can NOT get it to start using starting fluid.

And that really surprises me, because it appears to be getting a good spark signal. And I doused it pretty good.

My apologies for any confusion this adds to the situation. All other parameters are as previously described.


Mark R.
06-30-2017, 06:32 PM
I've been using a Timing Light with an inductive pickup to check for spark on each of the cylinders.

I get a nice bright pulsing light from each of the plug wires when the engine is cranking.

Is this a reliable enough method to conclude I have good spark..?


Mark R.
06-30-2017, 06:40 PM
Also, I checked the vehicle alarm system and everything appears to be functioning like it always has.

I set it off intentionally (horn pulses and lights flash), and then turned it off with the key like normal.

The red LED blinks slow when it's armed, faster when unlock the door, and goes dark once the key is inserted in the ignition.


06-30-2017, 07:39 PM
To review, this happened suddenly, no early symptoms, how many miles? Sound like it has good compression?

Mark R.
06-30-2017, 09:16 PM
Yes. It happened quite suddenly with no previous indication or symptoms.

Just coasting up to a stop light and it stalled and hasn't started since.

Mileage is ~170k. New head gaskets were installed around ~120k.

I believe compression is generally good. Have not measured it recently.


06-30-2017, 09:39 PM
Can you get a noid lite on one of the injectors? Confirm pulse? I know this can be tough. Also, may have jumped time, compression test the front cylinders to confirm its still in time. See if there is raw fuel on the plugs when doing the compression test, if none the injectors are not pulsing, you probably have determined this already.

Mark R.
07-01-2017, 01:23 AM
Tomorrow I will pull the front spark plugs and check their condition and the compression.

And I will see if I can locate a noid light, although I don't see how I could possibly get it installed.

I can't even see any of the injectors, let alone get my fingers onto any of them.

From my experience, if the injectors were firing during all these attempted starts, I would smell fuel by now.

I have cranked this engine so many times, that I've had to keep a battery charger on it most of the time.


Blue Bowtie
07-01-2017, 11:48 AM
Unfortunately I no longer have access to the GM service manuals to confirm, and most of my experience is with Fords. From what I've seen on Fords, and suspect GM is similar, the injectors receive constant voltage with the key in the Run position (probably battery voltage, but could be 5V) and the ground is switched via driver transistors in the PCM. So you should expect to have constant power and switched ground.


Your supposition is correct - The B+ is constant and grounds are sequentially switched by the PCM, if all required conditions are satisfied.

Mark R.
07-01-2017, 01:49 PM
The B+ is constant and grounds are sequentially switched by the PCM, if all required conditions are satisfied.

And as previously discovered, there is good voltage at the injector input connector.

So the question is: What would cause the PCM to loose the ability to ground the signal?

What input (sensor, signal, etc.) that the PCM requires, might currently be missing?

The sudden way the car died makes me think electrical, rather than anything mechanical...


Blue Bowtie
07-01-2017, 03:51 PM
The PCM needs to see crankshaft reference pulses from the ignition system. On the 3.4L V-6 that's usually a function of the 24x CKP. Since you are (were?) getting spark, that's already a "given." The system also requires a signal from the CMP to properly sequence/phase the injector pulses. If that is missing the injectors may not fire. On the older 3.8L Buick V-6 the system would eventually batch-fire the injectors in the absence of the CMP signal after protracted cranking. I'm not sure the 3.1/3.4L PCM setup does that, however, and thus may NEED a cam signal from the CMP. Of course, all of this has to route to the PCM.

Another item which will prevent injector pulses is the security system. If the "SECURITY" warning is flashing or on, it likely will not start even if it cranks.

A third item is the wire harness itself, including wires and connectors.

Most of that has already been covered here, too.

As for other, less likely causes in this instance, one item which is required for the injectors to operate is sufficient fuel pressure. The mere presence of pressure is not enough. It must be above 25-27 PSI to open the pintles and flow properly. Since you've already measured that at around 43 PSIG it should be more than sufficient.

Max is correct that a slipped timing chain could be indicated, but unless the chain is broken and the cam is not turning, there would be injector pulses, albeit out of sequence. The resultant fuel should either show up on the spark plugs and/or cause backfire in the intake or afterfire in the exhaust. Pulling a couple of easily accessible plugs might be revealing.

Instead of dismantling the intake, you might try back-probing the C102 connector to check the injector signals:


Mark R.
07-01-2017, 07:00 PM
Well, finally enough of a break in the rain to go out and remove the front spark plugs.

The good news is that there's a strong spark on all three plugs with the engine cranking.

The bad news is there is very little compression on all three cylinders...!!!

During multiple readings on each, the values were inconsistent, and never got over 50 psi.

This seems like some very major bad news...


Blue Bowtie
07-01-2017, 08:05 PM
It will still run on low compression. I've seen some lower than that driving into the shop as smoothly as can be. It's not optimal, but it might not be the worst thing. It also might indicate slipped timing, but there should still be fuel. If you can remove the front rocker cover without a major effort you might see if the valves are moving.

The fact that you have spark means the CKP is reporting to the PCM. You're halfway there.

Mark R.
07-01-2017, 08:45 PM
Now that I'm cranking with the front three plugs out, I'm actually smelling more fuel.

So the injectors may indeed be firing. I also hear an occasional light backfire pop out the exhaust.

When I remove the oil filler neck, I can clearly see both of the valve springs for the middle cylinder.

And no matter how long we crank the engine, those two valve springs never move.

I think there's a significant mechanical disconnect (like maybe the camshaft isn't turning)..?


07-01-2017, 10:16 PM
Broken camshaft is my thought-- if valve springs are not moving- seen it on 3.1L acting the same way as yours...

Blue Bowtie
07-01-2017, 10:57 PM
Beat me to it - Snapped (hollow) cam or lost chain.

It' s also possible the rocker studs have pulled on those valves, but that would be highly coincidental. Pulling the front rocker cover would reveal that quickly.

Mark R.
07-02-2017, 12:06 AM
I was thinking ----> broken timing chain.

Wouldn't a broken camshaft make a bunch of racket (clatter noise) when cranking?

I was thinking what's left of the timing chain is probably just laying down in the oil pan.

Is there any easy way to confirm either of these (without tearing into the engine)..?


Blue Bowtie
07-02-2017, 08:18 AM
Yes, you can remove the rocker cover and observe the valves. If only the first few move, the chain is connected and the cam may be snapped. If none of them move, I'd expect the chain drive.

The tubular cam can break at any point, and not necessarily make any noise. The cam will simply be retained in its journals and not move.


Add your comment to this topic!