Pontiac Solstice Forum banner

Car goes nuts, intermittently, under weird circumstances.

4K views 25 replies 7 participants last post by  raygun 
#1 · (Edited)
(I admit that the post title is not the most informative.)

A little background information:

Recently, I installed the A/C delete belt and replaced the oil dipstick tube. Before, there was a horrible clacking sound coming from the engine bay, and I tracked it down to the A/C lines hitting the intake manifold. Looks like the shaft on the compressor is damaged, so I just took it out of the equation. (The compressor is still in the car, just not driven by a belt anymore.)

At the same time, I replaced the engine oil dipstick tube. I had the one from the Malibu (engine donor), but it wasn't the right shape, leaving the dipstick partially covered by the wiring harness. I put a Solstice dipstick tube in, bolted it to the intake manifold and moved on with my life. The dipstick tube is a slightly bigger than the one I removed, though, so I'm waiting for a new dipstick to arrive. (The current one doesn't fit snugly, it's just sitting in the tube.)

Now, the problem: the engine runs *perfectly*, for awhile. The other day, I drove the car around for half an hour without an issue. No CELs, no hesitation or weird sounds. Then, the "reduced engine power" light came on. And, of course, engine power was, in fact, reduced. Still no CEL.

I got the car home, pulled the negative battery cable to reset the computer, started it back up and everything was fine. The next day, I drove it for about half an hour before the problem started again. REALLY bad hesitation, couldn't keep an idle, etc. As long as I stayed out of boost, I was able to drive it, though.

Today, I started it up, and everything ran fine, until I hit the gas (in neutral in the drive way). It revved up fine, but when I lifted it tried to die, basically acting like it was having a seizure. The reduced engine power light came back on. This continued for a few seconds until I shut it off.

Still no CEL.

So, I'm running out of ideas. I suspect there is an issue with the throttle body. When running a compression test recently, I propped open the butterfly valve with a wrench. (In retrospect, this was not perhaps the greatest idea.) The valve seemed like it moved fine at the time, but I didn't spend a lot of time looking at it. If it was the TPS, I would expect a CEL, but those aren't guaranteed by any stretch.

One other thing: every now and then, the air hose connected to the turbo compressor works its way up (and has come off a few times). I have a BOV that's working correctly, though it has a heavier spring than "stock" as I was seeing the valve staying open at idle. Possibly related, but I'm not certain.

Suggestions? The car is an 06 with 55k miles, 8k on the engine, mods in my signature.
 
See less See more
#2 ·
Have the diagnostic codes read. Even without the CEL you will find some historic data.

Also, "Reduced Engine Power" has been related to throttle plate position errors.
 
#3 ·
Okay. I'll dig the HPT out when I get a chance.
 
#6 · (Edited)
I like "obvious" - the obvious things usually don't occur to me.

The battery itself is fine, BUT (and it's funny that you should ask) the negative terminal on the harness is a little messed up. I can't tighten it down securely, and sometimes it rotates a bit on the post and I have to move it to start the car, even though, visually, it seems to have good contact with the post.

Yeah, I know - a little *******, but it hasn't been the highest priority. It's been that way for quite awhile, I just haven't gotten around to picking up a new terminal connector.

Could that be the problem? :/ (Pardon the silly questions - my brain is saturated with geometry and DirectX12 crap at the moment, writing a presentation for a conference at the end of the month. )

Usually, I only have to move it once and the car starts. Yesterday, though, it took five or six tries.
 
#8 ·
I'm with Rob....fix the battery issue first.

On a neat sidenote: A feature of our ECM is that it stores ALL historical data. Everytime you've had a CEL, it's still in there, even if it's been "cleared", you just can't access it with an over-the-counter code reader. When I was first trying to figure out if I was tuned when I bought my car, a buddy that is an engineer at the proving grounds hooked his GM computer up to my ECM to pull my program part numbers (the ECM has 4 "Part numbers" that make up the tune within the ECM). When he hooked up his work computer, he was also able to see that my ECM had been flashed (tune adjusted) 6 times and that I had several CELs in history relating to the O2 sensor.
 
#9 ·
Historically if there is any issue with the DC voltage supply the BCM and ECM have evidenced erratic behavior. You need to make sure that the voltage supply from the battery is clean and consistent to eliminate that as a potential problem.

The problems you are having could be a result of several different causes. It could be the throttle valve is sticking / dirty and causing erratic performance. It could be that one or both of the cam servos are failing and the cams are not responding as required. It is less likely but you could have an issue with the high pressure fuel pump. Check the oil for fuel contamination and verify that it is operating correctly.

Following down the path of fuel delivery you can check the in tank fuel pump to make sure its delivering fuel properly to the HPFP.

There have been instances where fuses have had connectivity issues that have caused strange and hard to diagnose issues. Some have had positive results from removing and reinserting fuses to re-establish good contact.

Good luck
 
#13 ·
Sure, but the system is grounded to the frame. You can completely disconnect the battery with the car running and it will happily keep going as it's running off of the alternator.

Of course, "ground' is a relative term - it's not actually 0V, just ~12V less than the input voltage. I could potentially see an issue with an intermittent negative battery cable connection, shifting ground and causing some high-frequency noise on the line. I could just put a low-pass filter over by the battery. (OR - and I know this sounds crazy - fix it.)
 
#11 ·
I wouldn't rule anything out, and you should certainly fix anything that is obviously wrong, ie: the battery terminal.
I mentioned the throttle plate (AKA butterfly) because it is a known cause of this fault and because you said that you moved it with a wrench.
 
#12 ·
Thanks to everyone for the input. I picked up replacement terminal connectors at lunch. It needed to be fixed, anyway, and isn't a bad place to start. This really feels like an electrical issue.

The battery is less than two years old. It looks fine. I really wish these things had a battery meter. Probably should've picked up one of those, as well - could just install it in the glove box.

While I'm replacing the terminals, I'll pull the LH intake stuff off and clean up the throttle body. I happen to have two more TBs and intake manifolds. They're inconveniently located about ~100 miles away, but, if this doesn't fix the problem, it's a straightforward task to swap them out. (Also, on a slightly related note, the easiest way to get at the starter in my experience.)
 
#14 ·
DDM Works sells a replacement A pillar trim piece that mounts two gauges. Many of us have installed them and like the functionality they provide. I have oil pressure, coolant temperature and voltage gauges.
 
#15 · (Edited)
Yeah, but I just made that fancy single-pod A-pillar, then made a fiberglass version, then made another fiberglass version after I destroyed the first one getting it out of the mold.

I'm tempted to install a PC in place of the head unit, or one of those Pioneers that mirrors apps on your phone to the head unit. Run Torque or DashCommander or somesuch. I assume voltage is available via ODB-II port but, then again, trusting a voltage reading from the computer, which is itself subject to fluctuations in said voltage, seems a little... crazy. And overkill.

Thus, the glove box. :)
 
#17 ·
Thanks for that. I'll update when I have more information.
 
#18 · (Edited)
Replaced battery terminals, no change. I'll pull the throttle body tomorrow and see what I see.

In the mean time, I recorded a short video of what's happening. I'm not sure what that ticking sound is. I've heard it before when driving, and it's not speed- OR RPM-dependent. :|

I have a feeling someone's going to say "Throttle plate," but injectors are a strong possibility. (I can live with the sound. It'd be less noticeable if I had some sort of cover on the engine.)



(I have no idea why the video is rotated)
 
#19 · (Edited)
So, here's a fun one: I went to take the throttle body off, and I found:

(wait for it)

The two TB bolts at 10 o'clock at 4'oclock positions weren't tight. One of them wasn't even threaded in all the way. I vaguely remember using my DeWalt 3/8" impact to snug them up, but must not have tightened them all the way with the ratchet.

So, things are MUCH better, but I appear to have a vacuum leak somewhere. While going for a test drive, the brake pedal got pretty low. Pumping it back up resulted in a good firm pedal, for a few minutes. And, under hard acceleration, it intermittently cuts out.

I love my car... I love my car... I love my car... sigh.

Also, I noticed some play between the throttle plate and the shaft. I'll take it off and see if I can tighten the screws holding it on, maybe with some LocTite.

Update:

After getting some actual work done - the worst part of working from home is that these projects are sitting right there, but I have to do my real (paying) job instead - I plugged up HPT and read the codes:

0x7E8: P0107 - Manifold Absolute Pressure/BARO Sensor Low (Current)
0x7E8: P0122 - Throttle/Pedal Position Sensor A Circuit Low (Current, Old)
0x7E8: P0223 - Throttle/Pedal Position Sensor/Switch B Circuit High (Current, Old)

I had unplugged the throttle body harness connector yesterday, and, of course, the CEL came on. The thing was running very poorly. I cleared the codes, shut it off for awhile, then did some logging. And, everything is working great. Of course, it's just been a few short laps around the neighborhood but, so far, no issues.

Fingers crossed.
 
#20 ·
Glad you seem to have found it. I hope you did.

I have banned impact tools from use in the plant for anything other than disassembly. None of my mechanics are allowed to use them for any kind of assembly for exactly the problem that you just found.
 
#21 ·
That makes sense. I'm very careful about over-torquing (nice analog trigger, battery last forever on that thing), so I use the impact to get the bolts in, then tighten with a (muscle-powered) ratchet. But, this isn't the first time this has happened.

Fortunately, it's just my car. If that happened in a business context, and a customer had an issue because of it, I could see Very Bad Things happening as a result.

Thanks
 
#22 · (Edited)
And, the REAL problem was:

There was an intermittent connection in the injector harness due to someone's poor attempt at soldering. I happened to have another injector harness, just with the wrong MAP connector. I cut the connector off of the current harness and made a very pretty Western Union splice onto the other harness, soldered and heat-shrink-tube'd it. (Thank you Wikipedia and YouTube for explaining good splice techniques.)



And now everything REALLY looks like it's working. I have my old high idle CEL back again but I can live with that for the moment, especially since the car is immobilized due to a brake error.
 
#24 · (Edited)
So the question is how did you discover, or why did you look for an electrical issue? I thought the loose TB bolts was the problem?
Tightening the bolts was a problem, and everything worked for awhile. I drive up to my family's place (~100 miles from door-to-door, mostly highway) for the weekend, and everything worked perfectly until I got to their gate. Then, the CEL came back on, and the car sputtered a bit.

The next morning, I had HPT hooked up in the shop to pull the codes and see if anything else jumped out at me, and I noticed a voltage spike at the same time as some hesitation at idle. This happened four times in ten minutes, all at idle.

Since everything had been working correctly, I went ahead and pulled the TB off again to clean it and make sure I hadn't missed anything, but nothing obvious presented itself. I put it back together, and everything was fine, until I revved the engine. Then, I noticed that the MAP output had no variation, and pulled the codes (no CEL, though there was a code for the MAP). I swapped out the MAP sensor with the old 1-bar I still have, and still nothing. I pulled the injector harness and did a continuity test between the MAP connector pins and the big connector that goes to the main harness, and abnormally-high resistance and intermittent continuity on two of them.

I looked at a previous HPT log and saw that the sensor had been working the previous day.

I had swapped the MAP connector out at one point due to differences between harnesses and the sensors that I had available, and didn't do the best job soldering. So, I cut the connector off, grabbed another injector harness out of the parts bin, did a very nice Western Union splice, soldered and heat-shrink-tube'd it. Tester showed nominal resistance (that splice is awesome), reinstalled and everything ran great. Until the brake issue, which I doubt is electrically-induced. :)

Most likely explanation is that, in the process of moving everything around to get to the throttle body off, I tugged on that harness just enough to finally break what was a crappy connection to start with.

Not a terrible video for the splice technique, which I'll be using for everything until I find something better. It's a little harder to do with stranded wire, but patience, good tiny pliers and a magnifying glass - my diabetic eyes don't work so well at that range - work wonders. (Word to the wise: don't use rosin core plumbing solder for electronics work. )

 
  • Like
Reactions: ChopTop
#25 ·
I absolutely know this sounds dumb, but my wife 07' drove us crazy for months with all sorts of crazy behavior similar to yours. First they said the problem was in the TCM, then they said it was the ignition switch and it could be a dozen other things, lucky it was under warantee. It spent more time on the hook than on the road. Finally, we got stuck an hour away from home and the car got towed to another dealer. The service writer told me that it might be a few days before they could work on it. Less than 30 minutes later, he called and said the car was ready. The tech who worked on the car told me that he had seen this several times before, apparently the factory did not torque the ground strap to the frame. That was seven years ago and the car has been flawless ever since. It might not be your issue, but a simple thing to check, good luck.
 
#26 · (Edited)
@prosoundz Yes, intermittent grounds are a pain, to say the least. The worst one is on the frame rail below the engine bay fuse box. If you don't get that nut tight enough or it wiggles loose for any reason, everything will work - but the car won't start consistently and trying to start it can make the ECU reboot - interrupting the connection with your scan tool, making it pretty much useless for diagnosing the problem.

The nut hanging on the stud is enough of a connection for the electronics to work, but not always enough for the starter. It's awesome.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top