2-for-1 and possibly related: Bucking and unstable/nonidle

[Ronin10]

Plug wires were pulled from the spark plugs themselves, but there's no separators. I'll see if I have some around the garage and add them in, but I'm doubtful that this is the issue.

I'm at a loss of what's left to check on the ignition side. Distributor has been rebuilt and gives very little scatter with solid timing at 28* BTDC using mechanical advance only. Coil, cap, rotor, wires, and FI harness have all been swapped with no change. I suppose I could pull the plugs and re-gap them to a smaller value.

Just thinking aloud here...given that the timing light didn't indicate any misses when checking cylinder-by-cylinder, then the spark signal is getting to the cylinders. If an ignition issue exists, it must exist inside the cylinder...that can only be bad or failing plugs or too big of gaps. Otherwise, it's something non-ignition, correct?

[energyturtle]

Hold on there. You changed the harness, issue went away, then returned. Everything diagnostically has been checked and are within the correct tolerances. I still feel you have a ground that is losing contact. Is the ground under the plenum clean and tight? Are the ground spades of the harness clean and attached tightly to it? Is the power wire to the ecu from the coil clean and tight? Have you tried pulling the plug on the cold start valve when running? Maybe place a washer between the temp sensor 2 and the head? It has to be something small. Keep at it, and good luck.

Scottie

[Ronin10]

Getting tight to the big holiday so time is hard to come by, but I did check my dwell. It 47.5 +/- 0.5.

I'm biased towards something electrical as well. My thinking though is the connectors in one of the injectors (theory of the moment). I noticed all my clip on connectors are tighter on the new harness so I'm doubtful on that front. Hadn't considered the cold start valve yet.

I still have to do Colin's nudge test on the AFM at 3000 RPM and check/set plug gaps. Then onward to reinvestigating connectors perhaps.

EDIT: Did a drive with the Cold Start Valve disconnected. No change. Could still be leaking though. Given the intermittent nature of this issue though, leaking as more of a steady state problem, is unlikely.

[Ronin10]

Summary of the last couple of days...

First, my ear is now tuned enough to where I can hear the previously described bass note throughout the rpms range, at least from the driver's seat. Second, on a whim, I checked the timing as close to idle as my son could hold it in the driver's seat, about 1000 rpm. The operation of the engine was quite rough, so it danced around a bit, moving from about 4* BTDC to 0* BTDC. Seemed to me to be way too far off for stable idle. Since my assistant had to skedaddle, I timed the engine statically to 7.5* BTDC. The subsequent test drive showed that my bus seemed to like that. If I manage the decel of the engine, it will hold onto a very rough 600 rpm idle, but it takes a lot of finesse. There seemed to be no ill effects on the rest of the driving range/load.

Today, I pulled my plugs and took a look at them. They seem pretty good to me, but let me know if anything stands out to your more trained eyes.




The gaps were quite large though, all around 0.030", give or take a thousandth. I regapped them to 0.024" and reinstalled them. Test drive showed no real change in behavior. Felt a miss or two on a short 2 mile route.

I did notice after re-installing the plugs - it was a cold, cold start - that the engine idle stably like usual around 600 rpm, nice and even, but as the engine warmed up the idle was more broken up.

I think that's it for now. Off to ponder my next move, maybe checking the spray performance of the injectors.

[Ronin10]

Attempted to inspect the cold start valve a about a week ago, but didn't get far because I couldn't get a wrench or socket onto the lower bolt (yes, it's a bolt, not a screw as it should be). Anyway, after giving up, driving became much more difficult...back to the debilitating bucking fits, and misses almost 80% of the time. Seems like the culprit.

[Amskeptic]

I did notice after re-installing the plugs - it was a cold, cold start - that the engine idle stably like usual around 600 rpm, nice and even, but as the engine warmed up the idle was more broken up.

So it fired up cold and held a decent cold idle.
Good cold start valve, good cold fuel mixture . . .

Broken up idle when warm? I guess that it is leaning out too far, looking at the plugs. Does wiper confirm? Does your diagnostic track keep getting lost in the confusion of potentially overlapping causes?

Attempted to inspect the cold start valve a about a week ago, but didn't get far because I couldn't get a wrench or socket onto the lower bolt (yes, it's a bolt, not a screw as it should be). Anyway, after giving up, driving became much more difficult...back to the debilitating bucking fits, and misses almost 80% of the time. Seems like the culprit.

Why does it seem like the culprit?
Is it a vacuum leak culprit - mounting gasket?
Is it a fuel leak culprit - rich running black plugs horrible fuel economy?
Is it an operational culprit - engine refuses to start when cold?

Your plugs look too lean for a gas leak, and operationally, the cold start valve only works while the starter is engaged.

Debilitating bucking, fits, and misses that go away and come back are not mechanical in origin, they are usually electrical if there is an abrasion in a wire, or a loose ground or etc, or they are some kind of transitory horrendous vacuum leak, like a s-boot pleat crack that only opens up when the engine pulls sideways under acceleration.

At this point in a rapidly demoralizing diagnosis to nowhere, you need to verifiably check off what it isn't.
That means a serious proof of each check. I have seen people point out a new condensor, "well, it can't be the condensor, I replaced it last week", and the problem was the new condensor, mimicking the original symptom of a stuck choke plate. I have seen a painful circle of parts replacements that actually fixed the problem, but the very last operation (plugging in a new double relay) inadvertently knocked the brake booster line off, and it was still a no-start. I have also had to confirm that all four reconditioned "ultrasonically-cleaned" fuel injectors were defective. It is a cruel world.
Colin

[satchmo]

If the bucking is like someone was turning the ignition off and on rapidly and repeatedly, check all the connections on the ignition switch, the wires from the ignition back to the engine, and all the connections in the double relay block connector and the wire harness/AFM. All it takes is one iffy connection to throw your AFM into fits.

Do you have a key ring with lots of keys and/or weight on it? That can cause problems with the ignition switch as well. General rule is one or two keys on a ring or fob, and that is the limit.

I don't think your cold start valve is the problem. Remember that the bucking symptoms improved for a bit after the harness was switched out. There is something going on with your wire connections.

Tim

[Ronin10]

After installing the plugs, the cold-cold start was indeed smooth, albeit low, but enough time has passed with enough cold-cold starts since that one good one to consider it a one-off, thus not lending it much credibility. Same story for the harness change. I've gone through my connectors and grounds and have been unable to find anything that looks like it'll cause an issue.

I acknowledge that the cold-start valve shouldn't be an issue here, but it's hard to ignore the step-function change after I messed with it. The plug photos were taken before I messed with the CSV and I don't have enough miles on the plugs to indicate a change as of yet. Maybe later this week.

Here's the current, typical behavior of the engine, after having fussed with the CSV: during cold start, the engine seems bogged down. I will hold a little throttle to keep the rpm up around 1400 or so, but it's not smooth until a minute or so has passed and then the rpm gets more "crisp." Given that cold start was crisp beforehand, I think my CSV might be leaking and I have exacerbated it. I did notice when the plastic barbs rotate relative to the metal housing on the CSV, not sure if that's allowed or not.

Anyway, with the manually maintained "idle," I'll sit like that for another minute or so. During this time and at any later point without a load on the engine, I cannot feel any misses or bucking, regardless of rpm, but I can hear that bass note I referenced previously. Once I drive away, the bucking resumes consistently, but at a mild level. A few times over the course of an about-town drive, it'll collapse into a major fit and I may have to stop on the side of the road. A few times it died altogether and wouldn't restart until some time had passed, kind of like it wasn't flooded.

I'm trying to focus on the idle issue at this point so bear with me if I table bucking-related issues for the time being. It's a nuisance, but I can manage for the time being.

So far, I've been through my entire ignition system and I'm confident it's up to snuff. I'm also confident there's no vacuum leaks. I've been over the engine numerous times, leak checked the s-boot, etc. The harness and electric connections are all clean and functional. I think we can consider these areas as fully flushed out. Which leaves some aspect of the fuel delivery as being amiss which is why I'm thinking about the injectors and CSV and how they might be affecting things, via internal electrics, leaks, or otherwise. I suppose I could revisit the double relay in there as well. After making a pass through that system, if I still come up empty, I'll gladly make another pass through the other systems. Anything else in the fuel delivery system that should remain on the table?

[Amskeptic]

Here's the current, typical behavior of the engine, after having fussed with the CSV:

during cold start, the engine seems bogged down. I will hold a little throttle to keep the rpm up around 1400 or so, but it's not smooth until a minute or so has passed and then the rpm gets more "crisp."

The above does not follow under the heading "cold start". The above follows under the heading of cold running. Disabuse yourself of any notion that your later symptoms have anything to do with the cold start valve.

The most typical cause of a sluggish cold idle is an AAR that is not passing enough air, and a possibly rich transitory period IF the warm idle mixture is too rich.

Note: back in the day, it was well known that early emissions saddled engines were a little cold-blooded or balky. Do not expect modern car instant driveability. That said, my '78 FI engine is a cooperative and smooth companion from the get-go **so long as I do not punch the throttle for the first five minutes**. When the mixture was adjusted too rich warm, I had choking cold running after an excellent cold start.
When I adjusted the warm idle to a lean 1,000 rpm, then cold running was more acceptable.

Given that cold start was crisp beforehand, I think my CSV might be leaking and I have exacerbated it.

Given that I do not trust our terminology here, what are you talking about?
Cold start - the engine launches off the starter motor to running independently of the starter.
Cold running - the engine's ability to keep running once you have let go of the key upon starting.

I did notice when the plastic barbs rotate relative to the metal housing on the CSV, not sure if that's allowed or not.

If no leak, don't worry about it.

Anyway, with the manually maintained "idle," I'll sit like that for another minute or so. During this time and at any later point without a load on the engine, I cannot feel any misses or bucking, regardless of rpm, but I can hear that bass note I referenced previously.

"That bass note" is extremely difficult to diagnose over the internet. I can guess and guess. YOU need to dissect this bass note and try to zero in on what the mechanical cause is. (The other day, I had to shut down a so-called clackity valve adjustment problem that was actually an exhaust leak )

Once I drive away, the bucking resumes consistently, but at a mild level. A few times over the course of an about-town drive, it'll collapse into a major fit and I may have to stop on the side of the road. A few times it died altogether and wouldn't restart until some time had passed, kind of like it wasn't flooded.

Totally lost here.

I'm trying to focus on the idle issue at this point so bear with me if I table bucking-related issues for the time being. It's a nuisance, but I can manage for the time being.

So far, I've been through my entire ignition system and I'm confident it's up to snuff.
some aspect of the fuel delivery as being amiss

Which is the idle issue? That you have to feather the accelerator when cold for a few minutes?
That the warm idle is more broken up than the cold idle? Look at the photograph of the plugs you posted. The far right plug has a veritable snow storm of lean white ash on the outer ring of the plug, suggesting a lean condition somewhere in its day. Are all injector seals intact? Do you trust your intake manifold phenolic spacer/gaskets? (let engine warm up fully until the idle gets lousy as you wrote above, and hit the intake manifold hard with WD-40 spray. Does the idle pick up?

I do not trust ignition systems, not until I have seen with my own eyes that the spark is consistent under full throttle. Have you actually noted that the ground strap on the breaker plate is good through full advance range? Have you verified that the green wire has not frayed under the insulation at the condensor (which has to be firmly attached for good ground) or at the sharp curve into the side of the distributor (these are speculations that only reflect the several solutions arrived at over the years)
Colin

[Ronin10]

Okay. Let's hit the reset button here so we can all get back on the same page. Internet communication requires so much time to type out all the correct words to ensure clarity, I think we're getting mixed up due to time constraints.

To put a fine point in my idle issue, I would describe it thusly... at no point, hot, warm, or cold, can I get the idle to stabilize anywhere near 1,000 RPM. This is what I'm aiming to fix.

Strategically, my plan of attack has been to validate air, fuel, and spark. On the air front, I have repeatedly been over the engine, looking for any kind of air leak and repeatedly found nothing with exception of the smoke test revealing a small leak seemingly come from around the AFM wiper pivot shaft. I also tuned the AFM, but ran out of wiper adjustment on the rich side of things. On the spark front, I have verified that I have spark going to each cylinder throughout the RPM range with no apparent interruption. On the fuel front, I have verified that I have good fuel pressure going to each injector.

So from here, it seems like the fuel, air, and spark inputs to the cylinders are good. Which leads me to conclude that there is something electrically going on in the injection system to bugger things up. To that end, the harness was verified, including the quality of the grounds. The connections at each component (injectors, CSV, double relay, resistor pack, ECU, coil) are all solid. Which leads me to suspecting the components themselves.

Is there a gap in my logic here? I'm trying to be methodical and not just take stabs in the dark every time someone suggests something, i.e. chase each systems' function down to the business end of things.

One the specifics of your comments/questions:

• I was indeed talking about cold-running, that period after the card has started, but hasn't really warmed up. Typically 3 or so minutes will pass after starting before I go anywhere. I'm tolerant of some non-modern running the first 10 minutes of so of driving. After that, I start considering things as issues.
• Since the idle issue as described above is at all operating times and temperatures, I wasn't considering the AAR beyond if it failed to close. I have had it out in a few months, but it worked correctly prior to installation. I have noticed though that the hose collapses at higher RPM. Is there supposed to be a reinforced hose in place there?
• I meant to say was flooded, not wasn't flood. Must have been rephrasing that sentence and missed making the change. Sorry for the confusion. That was in reference to a massive bucking fit which I had to pull off the road for and then the engine died. Upon trying to restart the vehicle, I couldn't get it to at all. It seemed like it was flooded.
• All of my inspections for air leaks have been done when the engine is warm, including the phenolic spacers, s-boot, etc.. FYI, the spacers were new when the engine was rebuilt two years ago. No changes to report when spraying with carb spray or from the earlier smoke test. I acknowledge that this situation screams "vacuum leak," but I sure as hell can't find it.
• When I rebuilt the distributor, I made sure the condenser wire was in good order throughout the advance range. I even tried swapping out with a couple of spare condensers I have. No change in engine behavior regardless of which condenser is in there.

Here's what I've done since this weekend... I put my LM-2 back on the bus to try and see if I could get an idea of what's changed since I mess with the CSV. Under 1500 RPM was still junk, but above 1500, my AFRs had all slipped to 15 1/2 to 16. So last night, I replaced the gasket under the CSV (had to fab one from gasket material). No change in idle behavior as of the drive in this morning. Cold, warm, or otherwise, it won't idle reliably. Plan on checking tonight after my drive home from work to see what the AFRs are and if the new gasket helped any.

[SlowLane]

[*] Since the idle issue as described above is at all operating times and temperatures, I wasn't considering the AAR beyond if it failed to close. I have had it out in a few months, but it worked correctly prior to installation. I have noticed though that the hose collapses at higher RPM. Is there supposed to be a reinforced hose in place there?

Well, in general, no vacuum hoses should be collapsing anywhere under any conditions. Which hose are you referring to? The one between the AAR and the air plenum (which is subject to manifold vacuum at all times)? Or the one between the AAR elbow (which you have inspected thoroughly, correct?) and the S-boot? The latter hose should never see a significant amount of vacuum, since it is attached upstream of the throttle plate, and the vane in the AFM doesn't introduce nearly the restriction to airflow than the throttle plate does. That hose would see a pretty high airflow velocity during the engine warm-up, however, since it acts as a short-circuit around the throttle plate. Both pieces of hose should be the standard VW rubber hose with the woven cloth covering (10mm? 12mm? I can't remember).

If you truly suspect a leaky CSV, then you can temporarily use a piece of FI-rated fuel hose to take it out of the fuel loop. Just replace the CSV in the loop with the length of hose. It'll make cold starting more difficult, but if you warm up the engine beforehand, that shouldn't present much of an issue. Then you can definitively decide whether it's the CSV or not.

Are you still unsure about the functioning of your decel valve? Again, I suggest doing your troubleshooting with the decel valve disabled (ie. remove the vacuum signal from the can) to eliminate that from the equation.

Finally, getting back to the valves and camshaft question: has this engine ever run smoothly at idle for you with this valve and camshaft combination? That is, do you know for a fact that it can run smoothly at idle, or are you simply assuming that it must be able to?

[Ronin10]

To clarify, I'm talking about the hose between the plenum and the AAR. The one between the AAR and s-boot is fine. The AAR is functioning correctly as of installation in winter/spring of last year. The hose is the standard cloth covered hose. Could be indicator that the engine is simply demanding more air that the AFM accomodate since my wiper is at its max rich adjustment.

I'm willing to table the CSV issue. The LM-2 test indicates more air not more fuel which is why I remade a gasket for it. Also, I have some better accessible bolts in place now so getting it in and out isn't as much of a chore as before. That lower bolt is still a pain though.

In my leak testing, I've removed and plugged all my various hoses with no change in performance and my recent smoke test didn't reveal anything either.

On the valves + camshaft issue...what you say has been lingering in my mind. When I had this on carbs, the idle wasn't great, but not awful either. I hadn't ever set up carbs before and it was mostly a case of "good enough" for the limited time I had back then. And really, I was barely driving my bus then and it was only ever meant to be a stopgap while I found the missing pieces to my FI system. But before I start blaming my cam/valve combo, I want to make sure that I've exhausted any FI system issues. Who knows, Colin may be helping me with this come August. *sigh*

On the decel valve, I think a component by component check is right around the corner so stay tuned, that'll be included as well.

One other point that I just recalled, Colin noted when he was here this summer that he though the oil breather was shot. Might that be in affecting things here?

[SlowLane]

To clarify, I'm talking about the hose between the plenum and the AAR. The one between the AAR and s-boot is fine. The AAR is functioning correctly as of installation in winter/spring of last year. The hose is the standard cloth covered hose. Could be indicator that the engine is simply demanding more air that the AFM accommodate since my wiper is at its max rich adjustment.

I think you ought to replace that hose. There's no way that an engine should be able to generate enough manifold vacuum to collapse a healthy hose of that short length and diameter. If you want to double-down on making sure about your AAR hose attachments, get one of these 1/2" silicone elbows (or similar: Google around). There's enough hose length on the elbow to do both sides of the AAR and banish the failure-prone stock rubber elbow from your life forever (I have one on my Vanagon). At $40, it's a bit pricey, but for me it was worth the peace of mind.

Also highly recommend you get yourself either a vacuum gauge or a MityVac with gauge attached (an incredibly useful tool), just so's you get a better picture of what is happening in your engine.

[Ronin10]

Fair point on replacing the plenum-to-AAR hose. I'm a little mystified by the collapse though. All my vacuum lines are less than 12 months old with maybe 5,000 miles on them, so pretty healthy.

The MityVac has been on my radar for a while, just hadn't been able to prioritize it amongst other expenses.

Did some poking around last night (nothing like an uber-confident statement about the lack of leaks to sow doubt)...used some zip ties to fix the throttle at about 1700 RPM, give or take, and hunted again for leaks. Noticed that my brake booster tee leaks where the hard plastic line to the decel valve inserts, but only on the forward side. It's small, but definite. That's a new one because that is among my first ones I check since it's so accessible. It doesn't seem to be big enough to be the smoking gun with respect to idle, but I'll be refreshing that piece anyway and seeing if anything changes.

I also decided to check the tightness of my intake manifold bolts as Colin commented on the spacers between them and the cylinder heads. Sure enough, they had backed off some. I re-torqued them to 14 ft-lbs per the Bentley (which seems a touch low to me, btw). I'll have to check them again in a few weeks to see if anything has changed. I want to make sure the studs themselves aren't working loose. These loosened previously when I was on carbs.

Lastly, I changed out my s-boot with a spare. It was really hard to say for certain, but there may have been an ever so slight RPM change when spraying around the boot. I couldn't get a good angle with the carb spray so it was hard to tell if it was a general fluctuation of the RPM or a genuine leak. Given this morning's drive in, it likely wasn't a leak as I saw no change in the behavior of the engine from this swap or from re-torquing the intake nuts.

Let's see, other comments I overlooked...90* elbow on AAR is new (< 6 months) and seals well.

[SlowLane]

90* elbow on AAR is new (< 6 months) and seals well.

I've had a cheap-ass aftermarket AAR elbow fail after just two months, so don't be making assumptions on that front. They seem to be made out of rubbery plastic with all the longevity of licorice.

The abysmal quality of available AAR elbows at the time was a primary driver for me to seek the more permanent silicone elbow solution. However, having taken a second look at the silicone elbow in the link I attached in my previous post, the walls are a lot thinner than the one I have, so I'm not sure it would keep its shape under vacuum as well as my thicker-walled one.

If you can, source a genuine VW AAR elbow (or two). They stand up much better than the aftermarket ones. Airhead parts may also have a higher quality one. They seem to take pride in their rubber products. I have one of their decel valve tees and it seems to be as good as OEM.