Read my spark plug...

[add|ct]

This is all on the same tank, weather, time of day/night? All of these crazy operating differences.

[theholycow]

Same tank, different time of day and different weather.

1. First drive on this configuration: 5:00 pm, rainy, 50s I think, drove a few miles before switching from previous configuration. Idle @ 1000, no knock, no buck.

2. Second drive: 7:30 am, 40s I think. Idle @ 1000, no knock, bucking at high RPM + heavy throttle in 4th

3. Third drive: 5:00 pm, don't remember weather. Idle @ 1000, no knock, bucking at any RPM with heavy throttle

4. Fourth drive: (today): 10:00 am, 60s. Idle @ 1000 initially then 400-500 once warmed up. No knock, no remarkable bucking but didn't have situations that would normally cause it.

[theholycow]

http://www.beforeblack.net/knock.htm

The 3.8 had electronic knock control (called Electronic Spark Control/ESC) in 1978 and the 4.1 got it in 1981. If I'm reading that article right, the 4.1's ESC was tied in with the ECM (which also showed up in 1981) but the 3.8's ESC from 1978-1980 was standalone.

I wonder if the 3.8's ESC system would work right on my 4.1. While the engines are nearly identical (bore, siamesed cylinders, and possibly weight of reinforced crankshaft differ), the differences could affect acoustic conduction. OTOH I imagine that the primitive system isn't sensitive enough for that to matter.

It's not likely I'll do it, probably too expensive.

[theholycow]

Today it idled 900-1000.

Originally Posted by Erik

I agree with your reasoning, but what about when the idle is extra high when it is cold (fast idle cam is engaged)- is there ported vacuum then? If so, that would support the idea that the butterfly is completely closed now. Maybe you could also investigate the secondary butterfly linkage- to be sure the secondaries are closing nice and tight.

That supports the idea that the timing needs to be more advanced at idle- which manifold vacuum would do for you.

Ok. Switched lines.

Distributor advance now gets straight manifold vacuum from the nipple that went to the automatic transmission modulator before. We'll see if that runs better.

Vacuum gauge now gets ported vacuum. We'll see if it still reads 0 while idling high.

Warm starting it after that switch-up, it idled ~1100. Gauge read 0.

[theholycow]

Ok, idle is 1500 again. Vacuum gauge reads at the edge of zero instead of right on it as when the engine is off, so there's got to be 1/4 in/hg at least. Next, I'll have to look at it idling low (disconnect vacuum advance or tee in for ported) and see if it reads the same...now I'm not sure if my zero reading yesterday was true zero or not.

[GasSavers_Erik]

There have to be air and fuel for you to have such a high idle. Assuming your fast idle cam system is working OK, I can think of only 4 possibilities to increase the idle that much and/or explain your high idle speed /low idle speed issue:

1. The primary or secondary carb butterflys are not fully closed (binding linkage, throttle cable too tight). Weak throttle close spring?


2. You have an intermittent vacuum leak (brake booster, etc.) and a rich condition at idle- together the extra air and extra fuel give you about the right mixture to increase the idle speed. (To test this- If you disconnect a small manifold vacuum hose does the idle rpm go up or down?)

Do you still have vacuum assist brakes when the engine is off and has been off for 30 minutes?

3. Your distributor bearing or advance or gears or timing chain is/are worn so that there is a lot of slop in the ignition timing. Wiggling/lightly twisting the rotor (check for bearing play and make sure the advance mechanism springs back when you twist the rotor lightly) and using a timing light would tell if this is the case.

4. You have an intermittent misfire. What all cylinders are firing, you get a high idle. When some are not firing, you get a low idle. Still, there would need to be extra fuel air leaking in to give such a high idle.

[theholycow]

Thank you, that all makes sense. Those are things I need to investigate.

I did try pushing the throttle closed by hand but that was during 1000 idle. I could hear a slight difference in RPM when I pushed it further than it naturally can go (and I think I could bend a bracket to make it do that on its own). I need to get a chance to try it at 1500.

An intermittent vacuum leak would explain the intermittent behavior, that's for sure. If the intermittent vacuum leak was in the distributor advance, that would neatly mesh with the question of why my idle seems so strongly affected on ignition timing.

So, having switched to manifold vacuum advance, it still isn't enough to avoid the skipping/bucking. Seems like I do need to advance my timing, but I'd wonder why. Perhaps that would be explained by timing slop (as described in 3 above).

[theholycow]

Originally Posted by theholycow

I did try pushing the throttle closed by hand {...} I need to get a chance to try it at 1500.

Tried it. No luck. It's not a sticking throttle cable. It was in exactly the same position, almost but not quite touching the A/C throttle solenoid. Obviously I'd still need to check the butterflies themselves.

This morning was awful, I had to drive deep in congested hilly Worcester and the skipping/bucking was just awful.

This afternoon I switched back to the TVS-ish thing behind the check valve (or whatever it is that's sticking out of the manifold, with the photo posted earlier in the thread). I bought 93 octane. Spark knock is back as expected, seems reduced as expected, but is definitely not gone.

Some observations...

Distributor connected to manifold vacuum:
- 1200-1600 rpm idle (usually 1500)
- ported vacuum measures 15-17 in/hg at the highest step of cold idle (immediately after starting) with my foot off the gas
- ported vacuum measures 6.5 in/hg at the lower step of cold idle (after using the accelerator to cancel choke but before engine warms up) with my foot off the gas
- skipping/bucking seems at its worst
- no spark knock, probably because the skipping/bucking happens first

Distributor connected to ported vacuum:
- 400-1200 rpm idle (usually 500 or 900)
- skipping/bucking sucks
- no spark knock, probably because the skipping/bucking happens first

Distributor connected to original "manifold" connection which I think is a TVS behind a check valve:
- 1500-1600 rpm idle
- spark knock with heavy throttle

I think I ought to use ported vacuum and either add more centrifugal advance or advance the initial timing.

[theholycow]

A couple random observations:
- When the skipping/bucking is most severe, the vacuum gauge can jump into pressure as high as 5psi
- Gauge jumps as high as 5psi pressure sometimes when dieseling during shutdown

Originally Posted by Erik

There have to be air and fuel for you to have such a high idle.

2. You have an intermittent vacuum leak (brake booster, etc.) and a rich condition at idle- together the extra air and extra fuel give you about the right mixture to increase the idle speed. (To test this- If you disconnect a small manifold vacuum hose does the idle rpm go up or down?)

I wonder is my charcoal canister could be saturated and that's providing extra fuel. Unrelatedly, I would likely implicate the evap system in the vacuum leak, or maybe the cruise control. I might want to try disconnecting both, and maybe the Thermac too.

Abbreviated list of things to check per that post's list:
1. Actual butterfly operation (binding inside carburetor?)
2a. Idle change when I disconnect a small manifold vacuum hose
2b. Remove every vacuum accessory/line/etc possible in case one is responsible for intermittent leak
2c. Try leak detection again, maybe try to identify inconsistencies and test in both circumstances
3. Timing slop from worn distributor or chain
4. Not sure how to check for misfire

[GasSavers_Erik]

To test #4 when the idle is extra slow, pull off the spark plug wires one at a time and see if idle speed drops lower.

If you are getting pressure in the intake during skipping and bucking- that is an intake backfire. It can be caused by lean conditions, or by bad plug wires/dist cap (you may be getting spark to a cylinder that is on intake stroke).