Unstable Spica idle speed

[paulsle]

Here’s a problem solving opportunity for the Spica experts among us.

Starting near the end of last season the engine in my ’72 GTV seemed to acquire what I will call a 2 stage idle. The idle speed during and after initial warm up was normally about 850 to 900 rpm, but then once fully warmed up, would only occasionally idle down lower than 1,100 rpm. All spica adjustments and other engine tuning settings were carefully adjusted by the books. (Wes Ingram’s, Pat Braden’s Owner’s Bible, AROC tech notes, Factory fuel injection manual. I culled what I thought to be the best info from each manual, with most weight given to Wes I’s.) No vacuum leaks were found, and the Marelli Plex distributor advance mechanism seemed to be nice and free. I tested the thermostatic actuator in a pot of hot water and it seemed to be performing pretty much right to spec.
I decided to just live with the sometimes low, sometimes high idle through the rest of the 2002 season.

While putting the car to bed for the winter in December, I discovered that the original 140,000 mile Spica pump (with missing reference screw cover), while seeming to function ok, was loaded with very diluted fuel. Since I had a spare pump, I decided to send it out to Wes Ingram to be tested as a replacement. I didn’t want to pay for a full rebuild at the time since I’d already planned another expensive engine project for the winter. The new, used pump tested out ok and Wes pronounced it fit for re-use with a life expectancy of 20-40k miles.

Over the winter the engine received new cams (old ones were worn), a very complete reworking of the head, a new oil pump, and of course the new, used Spica pump.

All of the engine work is proving out fine after 600 miles (although I think that the power band might benefit from another cam timing tweak) but the exact same idle problem is still present. The idle speed after warm-up is usually running about 1,100 , but only occasionally settles down to it’s base setting of 850.

As before, the Spica system is carefully set to spec, the TA seems ok, ref gap set to .019 @175 degrees, the idle air o-ring is good, no detectable vacuum leaks, the throttle cable and links are free, and the distributor is ok.

So my question is what could I be overlooking? Is a component of the Spica pump malfunctioning, perhaps an incorrect setting, or could the problem be unrelated to the pump?

Any suggestions, either on or off the forum, are welcome and will be appreciated.


Grazie mille,

[Roadtrip]

Paul -
It seems you are well-versed on the SPICA system. If your pump gap is .019" or less at hot idle, then I don't think there's anything wrong with the FI pump or T/A. I also assume that your running mixture is correct (light brown spark plugs).

It sounds like an air leak or self-mis-adjusting idle system somehow. The SPICA idle system has never been terribly precise, but the following may help in further diagnosis:

Check:

1. Idle O-ring in air distribution tower on manifold is hardened, missing, or not adjusted. (likely, but I don't think it'd cause your symptoms . . . . just the inablilty to affect any change. Idle speed would be wrong, but consistent.)

2. Air leak at manifold gaskets (not likely)

3. Air leak past very worn throttle butterfly shafts (likely in a high mileage car).

Try these things.

With engine warmed up and at idle, pinch off each idle air supply hose and note and difference between individual cylinders.

Next, also at idle seal off each intake manifold bore one at a time and listen/feel for any difference between cylinders. It should make no difference since at idle the engine should be getting ALL it's air through the idle air tubes. If the butterflies are slightly open, then you will probably feel a drop in idle as you cut off the extra air. Next, at the same time you might pinch off the idle air hose for that cylinder to see if the cylinder is leaking significantly through the throttle shafts. Maybe some heavy grease around the pivot point seals may seal it enough to make a diagnosis. You also may try just manually manipulating the shafts to see if there's any obvious play as well.

I'd be interested to know what you find.

[paulsle]

John,
Thanks for reply to my post. Your tips are quite insightful and right along the lines of where I’ve been thinking the problem must be. I made time today to pop the hood and check through the items you suggested. Here’s what I found….

Pump gap:
Pretty much .019 at 175 degrees, then closes up to .000 when fully warmed up. I’ve set and reset this a number of times, as the gap seems to be a moving target when the h2o temp is climbing. I’m now satisfied that it’s right.

Fuel mixture:
After reassembling the engine and changing the Spica pump this Spring I discovered that the mixture was extremely lean. The clue was a very low idle that could not be increased, and the proof was the white spark plug color. At that point I did the 2400 rpm mixture tuning procedure described in most of the tune-up guides. Upon completion and road testing the plug color was then gray, so I twice more richened the mixture by an eight turn. The plug color has been running medium to dark gray, bordering on brown. I’ve been reluctant to increase it further since I seem to be getting away from the setting arrived at with the 2400 rpm method, and the engine has been running very well. No backfiring on decel and no pinging during accel.
All that said, today I gave the solenoid another 1/8 turn to richen it a bit. I’ll pull the plugs to read the color after a run down the highway. Fortunately (at least for this purpose ) I live about 45 seconds off of the highway with no stops between the ramp and the garage, so the plugs won’t foul from idling. I’m hoping to adjust the mixture using the exhaust gas analyzer at the inspection station when I go in for the annual sticker renewal due during Sept.

Idle o-ring:
Recently replaced with a new one, it works the way it’s supposed to.

Air leak at manifold gaskets:
The manifolds were attached to the head after port matching at Glynn Motorsports where the head work was done. They’re fastidious, but it’s always a possibility, although low on my checklist.

Air leak at butterfly shafts:
After brushing grease around the end of the #4 butterfly shaft, I think I detected a barely discernable drop in idle speed. I’ll do this again to see if it’s a repeatable event.

The engine behaved as it should when sealing off each intake port individually and pinching the idle air hoses with the engine idling.

Nothing much else to report now, but once I track down the cause of this erratic idle, I’ll make it know to the board. Good luck on your Spica manual edits.

Thanks again.

[paulsle]

John,
Thanks for reply to my post. Your tips are quite insightful and right along the lines of where I’ve been thinking the problem must be. I made time today to pop the hood and check through the items you suggested. Here’s what I found….

Pump gap:
Pretty much .019 at 175 degrees, then closes up to .000 when fully warmed up. I’ve set and reset this a number of times, as the gap seems to be a moving target when the h2o temp is climbing. I’m now satisfied that it’s right.

Fuel mixture:
After reassembling the engine and changing the Spica pump this Spring I discovered that the mixture was extremely lean. The clue was a very low idle that could not be increased, and the proof was the white spark plug color. At that point I did the 2400 rpm mixture tuning procedure described in most of the tune-up guides. Upon completion and road testing the plug color was then gray, so I twice more richened the mixture by an eight turn. The plug color has been running medium to dark gray, bordering on brown. I’ve been reluctant to increase it further since I seem to be getting away from the setting arrived at with the 2400 rpm method, and the engine has been running very well. No backfiring on decel and no pinging during accel.
All that said, today I gave the solenoid another 1/8 turn to richen it a bit. I’ll pull the plugs to read the color after a run down the highway. Fortunately (at least for this purpose ) I live about 45 seconds off of the highway with no stops between the ramp and the garage, so the plugs won’t foul from idling. I’m hoping to adjust the mixture using the exhaust gas analyzer at the inspection station when I go in for the annual sticker renewal due during Sept.

Idle o-ring:
Recently replaced with a new one, it works the way it’s supposed to.

Air leak at manifold gaskets:
The manifolds were attached to the head after port matching at Glynn Motorsports where the head work was done. They’re fastidious, but it’s always a possibility, although low on my checklist.

Air leak at butterfly shafts:
After brushing grease around the end of the #4 butterfly shaft, I think I detected a barely discernable drop in idle speed. I’ll do this again to see if it’s a repeatable event.

The engine behaved as it should when sealing off each intake port individually and pinching the idle air hoses with the engine idling.

Nothing much else to report now, but once I track down the cause of this erratic idle, I’ll make it know to the board. Good luck on your Spica manual edits.

Thanks again.

[Roadtrip]

Paul -
Hmmmmmmm . . . . . .

Sounds like the pump is in excellent tune. It's normal for the gap to decrease to less than .019" when the engine is at operating temperature (about 190 deg F) since the T/A piston continues to extend even after 170 deg F. As a point of fact, the mixture does not change with gaps of less than .019".

1. Check the plastic rod ends (both short and long) for cracks, slippage, roughness, or looseness. Look closely. They're not easy to detect. With the engine hot, the short rod should not be holding the throttle butterflies open at all.
2. Check binding, roughness, or shaft play in the throttle bellcrank. Have you taken the bellcrank off and regreased the shaft recently? Does it rest against the throttle stop screw at idle? You already said the throttle cable has the nominal slack at idle.
3. Check the oil separator and hoses not plugged or gunked up. Try pinching off the hose and see if that makes any difference.
4. Check for cracks or leaks in the short hose from the air box to the idle air distribution tower on the manifold.
5. With the engine at idle and you remove the hose from the air box to the idle air distribution tower, what happens? Pinch or put a small hose clamp on the hose and tighen it slowly to restrict the idle air intake. Does that bring the idle speed down to a steady, consistant level?

Past that, the only thing I can suggest is to swing a rubber chicken over your head in the front yard and scream like a banshee. This is an extreme measure to be used in SPICA emergencies only.

Keep us posted.

[papajam]

"Swing a rubber chicken over your head". I like that! But it might not be necessary. There's two other things I've run across that you might want to check.
When the idle hangs up, what is the ignition timing? Even though "... the Marelli Plex distributor advance mechanism seemed to be nice and free", it may be starting to advance too soon. This can be caused by a weak primary spring in the advance mechanism. Try retarding the timing, for testing purposes, to see if the symptoms change.
The other is cam timing. You mention installing new cams but not a new chain. Assuming the chain has the same 140k miles as the Spica pump, then that puppy is stretched. Are the new cams stock? If not, what is the makers lobe center timing?

[Roadtrip]

Those are both good points. I assumed that the ignition and timing checked good. Did you put a timing light on it and see if the timing seems to vary at idle? I suppose you could duct tape the advance weights and see if that made any difference. Retarding it as a check is a good idea.

A grossly stretched timing chain sounds like an interesting cause also. I would assume that puppy would really be making some noise too.

It's common to blame the fuel system (carbs especially) for poor running engines when the real culprit is the ignition system.

Swing the chicken just to be sure, however.

[paulsle]

John S. and Papajam,

Thanks Gentlemen. It doesn’t look as though the rubber chicken act will be needed. With your help it looks like this little mystery is solved, or at least about to be. But first let’s run down your list of suggested items to check and I’ll comment on each…

Plastic rod ends:
Thanks for thinking of this. They were checked, cleaned and lubed during the recent top end assembly work, but I did find that the short rod needed to be adjusted ½ turn to fully close the butterflies. Naturally this change slowed the idle a bit. The locknuts needed to be tightened also. Good, one small step in the right direction.

Bellcrank & Idle stop screw:
No problems here. Bellcrank cleaned and made nice during the headwork. Idle screw stops the bellcrank at the prescribed 10 degree angle.

Oil separator and hoses gunked up:
They’re not. Oil separator removed and cleaned out a while back, and the hoses are reasonably fresh. All hoses, including the one running from the oil canister to the dipstick tube, are free of clogs.

Pinching off the hose from the idle air tower to the oil separator:
Yes, this lowers the idle slightly. But that’s to be expected, isn’t it?

Marelli Plex distributor / Ignition timing:
Here is where I believe the problem lies. This distributor was apparently installed by a previous owner and I don’t know it’s mileage or wear. What I discovered back when I set the timing is that if the static or idle mark is observed, there is nowhere near full advance at 5,000 rpm. I’ve read that this is a common problem with this distributor. I set the timing for full advance at 5k rpm. Today I watched the timing marks at idle and noticed something interesting. The advance changes with any small increase or decrease of idle speed. I may just be under the wrong impression, but I always thought that the advance should stay put at and near idle, and then start to increase at some point above a very fast idle. I didn’t try retarding the timing (not sure I want to), but Jim’s idea of a weak primary advance spring may be on the mark.

Where the heck does one find proper replacement springs?

Actually, due to the unknown wear and limited advance of my Marelli, I’ve been looking into the other ignition alternatives available. There is a message thread on the Giulia and GTV forum (conversion to electronic ignition) discussing many of the options. At this point I’m leaning towards the Bosch distributor. Ereminas Import’s monthly ad in the Alfa Owner magazine offers this with a “Hall effect ( inductive pickup)”option. I’ve not heard of the Hall effect before seeing this ad, does anyone know what it is?

Cam chain / Cam timing:
Upper chain is new with the head work. I suspect that the lower chain is probably original with 140,000 miles on it. I have an itemized invoice describing the engine rebuild at 100,000 miles. No mention of either chain being replaced.

The new cams are Colombo and Bariani with 10.9mm lift, 280 degree duration. The cam maker doesn’t specify the lobe center settings, but since timing for these cams is 30/70, I believe 110 degrees is the setting intended. I chose this particular grind as it very closely matched the cams this engine previously ran. The old cams were unmarked (probably Shankle / Ricambi) regrinds, installed when the car was no more than a few years old. With those old cams the engine came on like a buzz saw above 4,000 rpm, and retained enough low end power to allow driving at “around town” speeds without constantly dipping down into 3rd gear.
I can’t give a precise answer to Jim’s question about the cam timing setting, although I think that they’re set near 102 degrees. The problem is my cam caps have “additional” timing marks incised into them and it’s hard to know which are the factory marks. (see photos). As best as I can make out, the old cams were set with the intake a few degrees advanced and the exhaust retarded just a click off the original mark. (That click may have been due to the old chain being stretched.) On the advice of Paul Glynn who did the head work, I set up the new cams to what I think are the factory marks as a safe starting point. After a few hundred miles I tried advancing the intake cam to the newer, scratched in mark. This change caused a noticeable rough spot at 2,200 rpm so the intake is now back at the safe setting. I’m not sure where to go from here with this. I’d like to try 110 degrees, but due to the original timing marks being messed up, I don’t feel like I have a reliable reference point to start with. Alfa Ricambi at one time sold a cam timing kit that’s now NLA. This is another area where I’d welcome advice from anybody who’s coped with this problem.

I’ll keep you posted on the progress.

Paul

[Roadtrip]

I didn't know you were running non-stock cams. Do you have the 10.2 pistons as well? If so your probably a candidate for one of Wes Ingram's high-performance FI pumps. I bought one off Ebay a while back that's labled HP-150 for 150 horsepower. It's meant feed the appetite of a engine with hi compression pistons and 11mm cams.

If you're trying to boost the performance, that Marelli breaker point distributor should probably go into a storage box. It's barely up the task of a standard engine. Centerline has some custom advance curve Bosch distributors with electonic ignition. Not cheap, but not out of line.

Do the springs on the advance look broken or really stretched out? Did you try taping the weights to see if the idle settles down?

[papajam]

I automatically check the centrifugal advance for sticking/binding everytime I tinker around under the hood. Very simple process. With the dist cap off, grab hold of the rotor and turn it clockwise until it stops (about 10 - 15mm). It should turn quite easily without binding. Now, very gradually, release your hand pressure on the rotor to let the advance weight springs return the rotor to rest. Now try to turn the rotor counter-clockwise. If it moves at all, the advance mechanism is not fully returning or the primary spring is kaput. Simply remove the springs and weights, clean out all the gunk, reassemble and retest. Resist the temptation to tighten up the springs; the secondary spring is fitted so loose it's a wonder the bloody thing doesn't fly off into the boonies somewhere.
Try IAP for the cam lobe center cards. If NLA from them too, I'll copy the ones I have and mail 'em to you.

[paulsle]

Hello Roadtrip & Papajam (& anyone else who cares to jump in),

Apologies on the delayed reply to your most recent suggestions.
I had to take some time away from the BB to finish a few work related projects. Also managed to make it to Lime Rock for the Vintage races on Labor day. Reasons aside, there’s some progress on the Spica idle problem to report…

Quote:

Do the springs on the advance look broken or really stretched out? Did you try taping the weights to see if the idle settles down?

[Roadtrip]

Wes Ingram has a lot of experience with HP engines and SPICA. You might drop him an email and ask him about your setup and recommendations on ignition and pump timing.

I'm still fuzzy on what kind of cams you have on the engine presently.

With a high compression pistons and hi lift cams, I think you'd definately need a hi perf SPICA pump to go with it.

I have a spare Ingram "HP-150" (150 horsepower) injection pump. I'm not sure of the condition, but I plan on fitting it to my stock Alfa engine just to confirm it works. Although I think I'd like to keep it, and assuming it works, you'd be welcome to at least try it out and see if it performs well for your setup before you invest in a new one.