MFI idle oscillation help please

[SP2]

Howdy. The 69E is running great. New spark plugs seemed to make a big difference. My car starts well and idles well initially, but after it is warmed up the engine idle goes up and down, up and down. I wanted to try my first tune up for the basic things like new distributor cap and points, valve adjustment.

JW inspected my car today (thanks!) before I embark on my restore. Incidentally the compression was 180-185 for all, so I was pretty stoked. He noted that the cold start is not functioning because there is either a hose or a wire on the cold start to the engine that is not connected. I haven't had a chance to look at this in daylight, but John said the cold start is near the fuel filter. The only other thing that might be related is that the ignition wires are weird blue NGK wires. My SC wires have a wire loom on the outside that are grounds. Does this mean that the NGK wires are not grounded? The engine also cam with an MSD ignition and the tachometer is not working. Not sure if this would contribute to the problem.

The car never got warm enough for JW to hear. But today wasn't a tune-up day. That will have to come later.

My background is pretty new guy stuff. Anything I can adjust to get rid of the idle oscillation? Thanks.

[Grady Clay]

James,

The oscillation is the combination of idle speed and the micro-switch setting. You can make it quit oscillating by pulling one of the wires off the micro-switch. The switch is on the leftmost end of the throttle cross shaft. You can also just leave it alone. It is only annoying, not causing a problem.

Don’t adjust anything. Let John do that. It would be advisable for you to have him do a complete CMA procedure. For him to do it correctly, it would be good if you put some highway miles on it, fresh oil and several tanks of fresh fuel.


“The 69E is running great. New spark plugs seemed to make a big difference. My car starts well and idles well initially….”

This usually means the mixture is too rich.


“He noted that the cold start is not functioning because there is either a hose or a wire on the cold start to the engine that is not connected.”

If it still starts and runs, this usually means the mixture is too rich.

Best,
Grady

[Jeff Higgins]

How fast is the idle? Sometimes it will be at just the right speed to where the mechanical advance wants to start advancing, which will speed up the idle. Sometimes that will make it oscilate.

The cold start system consists of a solenoid on the fuel filter console that when activated re-directs fuel flow. It sends the fuel to fuel lines right on the stacks, providing a priming charge, like pumping the accelerator pump on carbs. This solenoid is controled by the thermo-time switch located on the oval breather cover just to the right of the MFI pump. The thermo-time gets its power from the starter solenoid when you turn the key to engage the starter. When it goes "hot" with juice from the starter solenoid, it sends power to the fuel bypass solenoid on the filter console, sending fuel to those stacks. The thermo-time will cut out after a few seconds of cranking when it heats up from the current going through it, ending the direct fuel flow into the stacks.

It's not uncommon to see the hot wire to the solenoid on the filter console disconnected. The thermo-time infamously fails quite often, and either sends fuel for too long into the stacks or not at all. It's kind of a fussy system. If the car starts just fine without it, and in temperate climates like ours a lot of cars do, just leave it disconnected. About all it will do for you to reconnect that bypass solenoid is to potentially dump unneeded fuel in on cold start, to wash down your cylinder walls and belch black smoke when it does light.

[SP2]

Grady and Jeff,

Thanks for the replies. I eventually will get the car tuned up by JW, but not til I get it repainted to stop the surface rust from getting penetrative. He did try to increase the idle because I told him the car seemed to idle too low. I don't know the MFI nomenclature yet, but he said it was a little screw on each stack that points to the right. I know it is only a bandaid but it is all I wanted him to do on the car yesterday. (I had scheduled a PPI-type inspection on the 69 as well as my other car's 915 for work.)

I don't really know what the idle is because the tach doesn't work. John thinks it is because of an MSD ignition in there, and there is sloppy wiring in there. There is a second red electronic box about 3x3 inches which I thought was for the signal to the tach. It sounds like I will just leave the hot wire unplugged since it has nothing to do with the idle up and down.

Jeff if it helps, when I started the car yesterday for pick up when it was cold, it started up right away without pulling up on the hand throttle and the idle sounded perfect. Later it started to oscillate. Maybe that means too rich. It's a new car and I will need to sort it out. I did a search and found Grady's post with a big compilation of links including a CMA pdf. I am interested in learning this. I am about to finish another project, then I can concentrate on the 69E.

The car hadn't been driven since 1996. I drove it home from San Francisco a month ago so it got 900 highway miles then. Incidentally, the idle problem surfaced a week later.

Thanks!

[Jeff Higgins]

I will certainly defer to Grady on MFI advice; I have learned everything I know about it right here on this site with his helpful, patient advice.

The microswitch he refers to is on the #1 cylinder stack; the throttle arm has a phillips head screw that is adjustable and pushes down on that switch. Working in concert with an RPM sensor, it is supposed to cut of fuel when the throttles close at high rpm by firing a solenoid on the back of the MFI pump. When RPM's drop below a certain speed (1200 RPM or so?) the RPM sensor shuts off power to that solenoid, even if that micro switch is closed, returning the pump rack to a position to supply idle fuel. "Hunting" up and down might mean it is turning that solenoid on an off. Anyway, if you are looking for that switch, look on the upper left side of the #1 intake stack.

Running too rich with MFI is no trivial thing. This system pours a lot of fuel in to begin with, and it is notorious for diluting your oil supply with fuel. It will do this both by running rich all of the time, or by actually leaking internally (it is pressure-fed by engine oil for lubrication). So don't assume you can just keep driving it until you have time to chase this down. If it really is rich enough to just "fire right up" when cold, you probably have a problem. The guys I'm refering to that live without the cold start system working properly have accepted a certain compromise in cold starting; they have to crank the hell out of them.

When you go through CMA, go through the WHOLE thing. This includes the valve adjustment (you have alread done the compression check) and general tune-up before starting. Set aside a nice long afternoon, and relax and take your time. Get a synchrometer and the long adjustment tools (our host has the ones you want); the protractors are very difficult to find and, frankly, optional. The dirty little secret about them is that once the rod lengths are set, the pump and throttles return to their stops with no pre-load or slop, the geometry you would check with the protractors pretty much has to be darn close.

Grady, please chime in on this if you would be so kind. This is a bit of advice I picked up locally from some MFI tuners, but I don't believe I have ever heard it from you. I actually saw a set of protractors at a shop one day when we were having a local R Gruppe MFI tech session. I asked the guy if he knew where I could get a set, and he said he hasn't used his in 30 years. Wouldn't give them up, though...

Once all of the rod lengths are set properly and air flow is synchronized, you will want to bring it to some one with an exhaust analyzer to set the mixture. I went to the Carb Connection in Kirkland and ran it on their dyno at full throttle, full load, and partial throttle for the low-speed.

Anyway, see if you can get this done before driving it much more. It's worth it to keep a very close eye on this MFI stuff and particularly that mixture. Lots of sub-systems control it under various conditions of cold/hot motor, acceleration/deceleration, and whatnot. If they are not working properly, there is a very real danger of pouring in too much fuel. This will wash down cylinder walls, dilute oil, and foul plugs.

[SP2]

Thanks Jeff. Looks like I should do some homework and read up on this stuff. I understand what you are saying about the gas in the oil. Lack of lubrication wouldn't be good for this almost 40 yo car. I haven't seen any black smoke when I start the car. Just white smoke for about 30 seconds for the first start of the morning. Looking forward to figuring this out.

[Grady Clay]

Jeff,

Thanks for your contribution. It seems that the Pelican progression is working. My years ago posts are used and refined and we have a new generation of MFI experts. Thank you for continuing the heritage and knowledge. It seems that also James is on his way to becoming another expert and teacher for the next generations.

I see this as part of our responsibility as current caretakers of these 911s


One of the most difficult aspects of the current existence for MFI 911s is the fact that too many rest quietly. Recovering from that state is difficult and expensive.
These 911s work best when driven regularly and aggressively.


James, one of your missions is to protect your engine while you have the 911 idle during painting. A good solution can be to loan your engine to another owner while his is being rebuilt. A little cooperative effort here benefits everyone. Our 911s and engines loathe sitting. The work best with regular and energetic use.

The worst thing for a 911 engine (particularly MFI) is to be started cold, not warmed up and then shut off for long periods. There are several Pelican threads on preservation.


I think this thread should focus on James’ procedure for getting the 911E running properly and not have it suffer from disuse or misuse during painting. I would never permit my car from being started when at a paint shop.

So how does James do this?

Just having completed a 900 mile trip is a good start. I would complete a through tune and CMA. I would also drive it a lot. If paint & rust are an immediate issue, spray on some primer. The key is to have the 911 in regular service and driven aggressively.

With a proper running MFI 911, preservation is much easier. I have written about this in other threads.

Best,
Grady

[SP2]

Thanks! My ultimate goal is to be diagnostician and mechanic for this car. JW has also identified a few spots of rust on the floor pan that need to be resected and patched. I am considering learning how to weld. It will be a keeper. No flipping when I am done.

Aside from the CMA, what do you consider a proper tune.

My rough plan is to refresh the suspension, pull the engine and tranny, strip to carcass then have towed to paint. Then interior, then glass. My car wasn't driven for 11 years, so I think the engine can wait. If I drive it to much too rich, Jeff points out I might do some harm. Thanks again Grady and Jeff.

[SP2]

Hello again. I have revised my restore plan. My earlier question about a "proper tune" is answered in CMA. While I am still worried about leaving the surface rust, I think I will, as Grady suggested, just take care of the small areas that need rust attention now and save the repaint for later. (small penetrative stuff on floor pan and R front fender near bottow of door, and lots of surface rust on bumpers and over L rear taillight. Probably will get the bumpers stripped of rust then primed and painted single stage slate gray.) Probably need to get the mechanicals tip top first. I have made one pass at all the MFI literature in the Pelican Tech articles under MFI. As Jeff suggested, I will get a synchrometer. I read the CMA and the use of a protractor seems important, and yet I hear you can do this without one. I would like to buy a set, but my search button tells me they are rare.

I like to be as quantitative as possible, and for a new guy it might be better to go by the book (CMA) with a protractor. So my questions are:

1) Which protractor set do I need for a 1969 911E originally 2.0, but top end rebuilt to 2.2 911S P/C in 1978?

2) Where can I buy them?

3) If I can't buy them because they are not available or beyond my budget, how do you CMA without them?

Thank you

[SP2]

Jeff and Grady,

I think this is the switch you are referring to? I am going to pull one of the wires. (The up and down is very annoying, and everyone in any parking lot seems ready to give me a diagnsis and advice. )

[Early_S_Man]

James,

Yes, that is the 'Microswitch' that controls the output from your speed switch/relay to the cutoff solenoid. At least you know the speed switch is working correctly!

[Cornpanzer]

My car has doen the same thing for the past couple years. Whenever the engine gets up to operating temps combined with cool ambiant temps, my idle hunts.

It will rev up to about 1400 rpm to the point that the over-run shutoff controlled by that microswitch kills fuel to the system. The engine starts to die, gets fuel once again and starts reving back up to 1400.....rinse ......repeat. You are right, this is very annoying.

What I found is that the throttle shafts on my car are worn to the point that they vibrate in the throttle bodies causing the throttle plates to flutter. The fluttering of the plates allows extra air into the engine and causes the rpm to drift. If you start your engine, you can probably hear the clicking of the throttle plates. Exaspirating (sp?) this problem was the fact that the plates were not being firmly held shut by the return springs.

In my case, I reached in and manually held each bank of throttle plates shut. What I found was that the right side bank was the problem area and as soon as I manualy closed and held the throttle plates shut, the idle returned to normal.

Not sure about you, but my arms are not quite long enough to reach the engine at stop lights, so I simply popped the main actuating linkage off of that bank of throttle bodies and adjusted the length of the rod approx 1/8" so that the return spring of the throttle linkage closed this side more firmly - voila, problem solved. No more hunting idle., and both throttle bodies are still in synch.

To be honest, I still need to budget for a throttle body re-bush, but for the time being, this will help me retain my sanity. I am on vacation now, but I can send you pics of the piece of linkage that I adjusted to accomplish this next week when I return.

[SP2]

Thanks panzer,
I will check my throttle arms. If they are worn, can you just replace them. I have been reading a little about the throttle bodies. How do you check if they need to be rebuilt.

From what I have been reading, MFI is an excellent fuel system used on Porsche's racing car before, so I want to keep it, and keep it running well. But I want to see how good I can get it before I start taking it apart.

I bought brand new P228b protractor set today. They said it has been discontinued and I bought the "last set in the world." That is pretty cool.

Anyone know why the fuel filter replacement for a 69 is so expensive? 60 bucks. Is there an alternative, like a new fuel filter with different connections?

[Cornpanzer]

Unfortunately, you have to have the magnesium throttle bodies re-bushed. Mag is very soft and this is what is wearing. Plan on $1000 + easy.

However, these are old cars and as much as we want them to be perfect.....its not always possible. SOme throttle shaft wear is to be expected. The fluttering throttle plates is simply a side effect of the wear. By ensuring that they snap back and close tightly, you are possibly giving yourself a couple more years of annoyance free operation until you can adress everything more thoroughly.

[Jeff Higgins]

If the throttle bodies are worn they will be leaking air around the butterfly shafts. The old trick of spraying carb cleaner at the ends of the shafts where you suspect a leak should tell you if or where they are leaking. The idle will speed up momentarily when the cleaner gets sucked in.

Another thing that might indicate a leak is if the air bleed screws have no affect on air volume in the stack. That could also mean the air bleed is gummed up, but you should clean them thoroughly before starting CMA anyway. The reason the air bleed screws would have little affect on a leaky throttle plate is because the air bleed would be somewaht "overwhelmed" by the volume of air passing through the bushings, and would not contribute proportionaly enough more or less to make up for that.

The six individual throttle plate springs Cornpanzer refers to are both very weak and prone to gumming up. Yes, you can "cheat" like he has and use the pressure of the big spring on the MFI pump itself to help close them. Like he says, not ideal, but it will get you by until you can fix them properly. Ideally every throttle rests on its own stop and the pump on its own, with no preload in the system. In other words, you can take the 114mm rod from the pump to the cross bar off and on and neither moves. You should also be able to take the two (also 114mm) rods from the cross bar to the throttle bodies off of the cross bar and it should not move.

That's probably not all that clear, but it leads into my "how do you do CMA without protractors" spiel. The pump to rack rod has to be 114mm, +/- 0.2mm; pretty darn tight tolerance. Make sure that one is right. Now the other two, from the cross shaft to the throttle bodies, are also 114mm but allow a +/- 5mm tolerance on that length. Those throttle bodies, in a perfect world, were set on a flow bench to equalize flow through each throttle. They should be equal with the individual butterfly's shafts on their respective stops, assuming some one hasn't messed with it. That is the key bit of information here; they should be good when resting on their stops. The obvious thing to do, then, is to adjust the right and left side "114mm rod" lengths so that the individual throttle arms rest on their stops. All three on one side should hit their stops together. So, set the rod length so the ball end will snap over the ball on the cross shaft, with the pump at its idle stop, and its rod to the cross shaft engaged. No preload: the cross shaft should not move.

Now everything is set all "loosy goosey" at the idle position, with no preload on any of the rods. Pump is on its stop, left and right throttles are on theirs; none of the three influencing each other. It pretty much has to be correlated at this point.

Oh, and that fuel filter - welcome to MFI. James. There ain't no substitute (for the filter, that is.)

[SP2]

Thanks Jeff. That is alot of info. I am sure the more I get into this, the more it will begin to make sense to me.

What is special about the fuel filter? Shouldn't it just be a filter between two hoses? Since I am not using that cold start thing above the fuel filter, can I eventually delete it some day?

With regards to the 114 mm, how do you measure this correctly. The photos in the manuals show calipers applied at both ends of the rod, but it also says the measurement is taken from the *center* of the ball. Wouldn't placing the caliper at the ends of the ball lead to a rod too short?

Thanks for the explanation of the CMA adjustment. It will help me when I dig into the MFI later. First I need to finish up my SC project, well at least the engine/tranny part of it.

[Jeff Higgins]

Mic the width of the ball cup and subtract that from the overall end-to-end length of the rod with the cups on it. That will give you the center-to-center of the ball cups.

You will want to eventually reconnect that cold start. Whether it is fired by the thermo-time switch or a button on the dash, it really is nice to have. The cars that do without it in any form really need to be cranked for awhile to get enough fuel into them to start when cold. Like Grady mentioned, it is likely yours starts easily now because it is very rich. Fix that, and the cold start will get noticably more difficult without the bypass circut working. So unless you are willing to put up with a less than convenient cold start, you will want all of that working and, therefore, that spendy filter element.