What causes in-gear and out-of-gear idle behaviour

[jedwards]

Gentlemen,
I am seeking help in understanding the differences between in-gear and out-of gear idle mixture in cars fitted with automatic transmission and an electrical Constant Speed Solenoid (CSS). Mine is a  '67 250SL automatic.

We have recently had a extended discussion on  Linkage Adjustment, because I thought that was where my problem lay.  However ja17, (always the consummate and generous expert) has made me realize that linkage adjustment is not my problem, so now I am trying to understand what is happening.

My car currently has smooth idle in park, good mixture setting, good mid-rpm mixtures and good high speed mixtures.  I have the CSS set to maintain rpm, against the increased drag of the transmission when put into gear.  So in theory, if the CSS is doing its job, when I move from park to in-gear,  the motor, timing, advance, FIP and everything mechanic should be still be spinning at the same speed, whether in gear or not. Only the load should change.

From what I understand, fuel supply is determined by FIP as a product of the centrifugal forces on the FIP rack, so if RPM remains constant, (irrespective of load) then fuel supply should remain constant.  The CSS moves the linkage assembly in their pre-determined ratio set by the relative lengths of the FIP and Venturi rods, so if RPM is constant,  fuel air ratios should also remain un-changed. However, my park idle is smooth and stable and  produces around 4.3% CO but as soon as I put it into gear, and the CSS restores RPM, idle becomes rough and CO drops to 1.8%. 

Other than raising the park idle mixture to around 6.0-6.5% to compensate, I cannot find any advice about what is actually causing this temporary lean situation.

Can anyone explain what is actually happening with the increase in load that is creating this overly lean situation at idle, which is resolved the moment the motor comes off idle?

[ja17]

Don't worry much about the park idle mixture. You hardly spend any time with the engine running in park! Strive for the best in-gear idle. 

When you put the car in gear, load on the engine increases and the fuel requirements increase. Sounds like your in gear idle is too lean. Do you have a vacuum dashpot on your venturi?

[jedwards]

Hi ja14,
No, I only have the CSS. Mine is an early 250SL (see image attached). I believe the vacuum dashpot was only introduced with 280SL.
Ignoring the off-load idle sounds like very good advice as always.  Setting the in-gear idle using the split method is a bit of an issue for me as I am doing all this alone and have no-one to stand on the brake for me. This might need to be an iterative process, so would 6.0-6.5% CO at Park idle be a fair starting point in your view?

Has anyone measured a well behaving car to see what a good in-park  versus in-gear idle CO measurement should be?

[stickandrudderman]

One of the great benefits of fitting a wide-band O2 sensor is to identify the cause of problems like this but you've already deduced that the co changes dramatically between in and out of gear. I have seen exactly this problem and the cause was the tiniest amount of play in the ball joints which was causing the throttle to be opened fractionally before the FP when commanded by the CSS. A simple half turn on the throttled control rod length adjustment was the very effective cure.

[jedwards]

Thanks Stick,
I've order a Gunson exhaust analyzer that will hopefully arrive this week. I've replaced all my linkages so fine tuning, as you suggest, should be that much more reliable.
I think it is a case of persistence eroding resistance in the end.

[wwheeler]

I have this problem to some extent on my W111 w/ a CSS. Except is the reverse. My idle in gear is good and out of gear shows a bit more engine shake. RPM is the same in both cases. I have not used the analyzer to determine what the % is though. Maybe more people have this issue but are just not trying to tune the engine to the Nth degree.

I agree with stick about the socket movement. While my entire linkage is basically new and well lubed, there is still a small amount of play.The one thing you did not mention is that when the CSS kicks in, it moves the throttle flap and IP off their stops. In out-of-gear idle, they should be on their stops. As you may know, a lot of things happen as soon as the air flap opens. One is that the air screw's effectiveness is diminished. So sounds like the air flap is opening more than the IP thus creating the lean mixture. Sounds like stick's idea is the cure.

Another but more difficult solution may be to replace ALL of the ball pins including the riveted ones. I believe you can grind the rivet head off and use the standard threaded type ball pin. I think only about half are easily replaced with the threaded pin. The other half are "permanent".   

[ctaylor738]

I had the same issue on my 280 and sort of resolved it by going a couple of clicks richer on the idle mixture to get the in-gear CO up to 3%.

But I have never seen a good answer to the question of why the mixture goes lean under load.  What is happening inside the engine that reduces the CO % when you drop the trans into gear? 

[garymand]

$0.02: the rods are not adjusted correctly yet.  The CSS is well placed to push at a place in the linkage that can instantly open the butterfly and the FIP at the exact same time, but your symptom is proving that is not yet the case in your linkage adjustment.  I will trust the experts here that the butterfly is opening a little ahead of the FIP.  My process is to do exactly what the CCS is doing: push the linkage to just take out the play in all the joints and eyeball the butterrfly lever and tghe FIP lever as they just start to move.  It sounds like your butterfly is actuated just ahead of the FIP.  And be sure to tighten all the locknuts once you are happy with the simultaineousness of the actuation.  It is really frustrating when the mixture drifts off as the rod rotates 

[jedwards]

Thanks Gary,
Good point. I'll have a chance to check that over the weekend.

The other interesting part of the puzzle is that it does not do exhibit this behaviour when cold and under the influence of the WRD, and only shows this leaning-out condition under CSS when fully hot, which with my car takes a quite a long time.  Its a 250 so has a good cooling anyway plus a newly cored radiator so it generally struggles to get up to 180.

So in the back of my mind there is a festering thought that this might actually be a combination of the thermostat keeping the coolant too cool to fully close off the WRD  until its really hot, which takes about 15 minutes of driving in my car. So when I work on my car, it may not be fully fully hot, and may have a slight contribution from the WRD to the idle mixture. Eventually it heats up enough to close the WRD and it goes lean under load.

So a quick test might be to remove an oval shim from under the WRD (lean it out a little) and see if the problem occurs earlier?  Any thoughts or comments? 
What temp thermostat is recommended for the 250?

[stickandrudderman]

So a quick test might be to remove an oval shim from under the WRD (lean it out a little) and see if the problem occurs earlier?  Any thoughts or comments? 
What temp thermostat is recommended for the 250?

"A quick test"? It's nowhere near as quick as a simple half turn on the butterfly control rod. Garymand has explained in different words exactly what I suggest is your problem. The fact that it doesn't happen when the WRD is active further proves the point because the extra air being bled in by the WRD negates the small effect of butterfly/FP differential.

[jedwards]

Hi Stick,
Certainly that will be the first thing I'll try tomorrow morning.
You are right in suggesting the impact of the additional WRD air, as I thinking only in terms of the increased fuel.
I'll work on it and report back. I've ordered a new MB 87 degree thermostat just to be on the safe side as well.

[ctaylor738]

OK, but what's the explanation for my car?  No CSS, but goes lean when dropped into gear?

[jedwards]

Hi Stick,
I had a chance to check out your suggestion with the linkage over the weekend and it did help things a little. As my linkages are all new and have vitally no play there was not much room for preferentially favouring the FIP without lifting it off its stop.  But I did succeed in getting a friend to help so was able to conduct a split test in gear, which was a first for me. No clear answer. Giving it more fuel or more air did not smooth things out much at all, only changed the RPM.
Things improve with more RPM, but clearly issues still needing working through.
The exhaust analyser should be here this week and I'll install a new thermostat just to be on the side side, so I hope to have some more useful feedback this time next week.