Troubleshooting CIS

[tirwin]

Ok, Pelicaners -

Here's the deal. I purchased a 1983 911 SC a couple of months ago.

Here's a pic for those interested:



I've been slowly and methodically working my way through a list of things that needed some attention. Today the UPS man is bringing me a new ignition lock assembly. With any luck I will have that sorted out this evening.

Next on the list is to sort out a cold running problem. (More on that in a bit.)

I have been reading the Bosch Fuel Injection & Engine Management book by Charles Probst and I have been browsing this forum for a while trying to familiarize myself with the in's and out's of CIS K-Lambda. Like any good engineer, I am compelled to start with a hypothesis and a good test plan before I start messing around with anything.

Wisdom Point #1: Always start with a good test plan. (Corollary: It can be really expensive to start changing things in a CIS system without having a damn good reason.)

Now, I have read lots of good advice on this forum but I have to admit that some of it is not organized in such a way as to be useful. A lot of good ideas -- but few seem to have a proper procedure for process of elimination. Even the Bosch book has stupid statements in it like "it could be [this] or [that] or any number of other components or connections." Well, duh!

So, I thought maybe I could kill two birds with one stone. Maybe while I'm solving my problem I could try to do this in a way as to be useful to others and post it when I'm done. It might not be perfect, but it would be a start. If people think this would be generally useful, I'd be happy to take this on.

Ok, back to my problem for a bit...

When the engine is cold, I don't have a problem with initial ignition. The engine starts but coughs and wheezes and dies out. Sometimes I get backfire in the air box but thankfully a pop-off valve is installed. If I turn the ignition and AFTER the engine has started I give gas, one of two things will happen:

1. Engine seems to have no reaction to the gas pedal. Doesn't seem to matter how little or how much I press the pedal -- it just coughs and wheezes until it dies.
2. The engine will react to the fact that I am giving gas. If I keep the RPMs at 2k for 20-30s then everything seems fine. However, if I don't keep the RPMs up for long enough, the engine sputters out again.

In summary, I don't think I have a cold start problem. I think I have a cold running problem.

One other important piece of information is that the PO disconnected the O2 sensor so it is always operating in open-loop mode. The PO also says that the mixture is set a little rich.

Once the engine is warmed up, I don't usually have any issues with a warm start. I have had a hard start issue once or twice but I suspect that it was due to the fact that I didn't drive it very long before I turned it off, so that still has me thinking cold running.

Based on all this, I am thinking of three possible problem areas:

1. Warm-Up Regulator (WUR)
2. Fuel Distributor
3. Vacuum Leak

Does this list seem about right? Any other thoughts on things to focus on?

Now, on to the test plan...

First step will be to break out my handy-dandy CIS gauges a run System Pressure, Cold Control Pressure and Warm Control Pressure tests.

Next step would be to check electrical on WUR.

Does this seem like a reasonable approach to get started? Obviously, once I start getting some results that will hopefully start pointing me in other directions.

Thanks very much in advance!

[Kemo]

sounds good! welcome to the board. Lets see the pressure results and move forward from there. looks like you are on the right track. the 83 SC is a really good year IMHO. Nice purchase...

[5String43]

These symptoms are precisely those that present themselves when I cold-start my '82. I've been suspecting that the WUR probably is the culprit, but when I called the guy who runs Fuel Injection Corp. to see about a replacement/rebuild, he told me that the fuel distributor more likely is the problem. Haven't done anything about this yet. But I'll be very interested in seeing what you find.

[ossiblue]

Welcome to the forum, tirwin, and congratulations on your "new" 911.

I agree, your proposed procedure is on target for your symptoms--first step, take the fuel pressure measurements and post your results if you aren't sure of their significance.

It sounds like you have a good handle on diagnostic procedures, which is necessary with dealing with CIS problems. Yes, there are many reasons why a symptom may appear and the book refers to that (your "duh" comment is duly noted) and that's why a systematic plan to eliminate one variable at a time is essential. Typically, with the symptoms you describe, the first step is to eliminate excessive fuel pressure as a factor, thus the pressure tests. Then, you go from there.

I don't know how much you've read up on the tests so forgive me if this is something you already know. After hooking up the gauge set, you will need to unplug the electrical connector from the wur. When you jump the terminals at the relay to run the fuel pump, current is sent to the wur, so for cold pressure testing the plug needs to be removed.

Once you've gotten your results for system and cold pressure, you can plug the connector back in, turn on the fuel pump, and watch the pressure gauge. If the pressure begins to slowly rise over the next few minutes, the heating element is working correctly--make note of the time it takes to raise the pressure, or note if the pressure doesn't change. This test checks two things at once--if the heating element is good and your warm control pressure.

As far as your three likely suspects, they are definitely in the mix but if you listed them in order of priority, I would move the vacuum leaks to position two given your symptoms.

Once again, welcome to the forum and report on your progress regularly as it will help many others with similar problems--and we will all have them sooner or later.

[DDD]

Keep us posted. One of my pet peeves is people trying to diagnose problems and then the thread never shows a resolution. Who knows, maybe they never get resolved? I am in the middle of trying to sort out a high idle problem on my '83. Still don't have it figured out, but through all of my searching and research elsewhere, I have learned (I THINK) that the O2 sensor doesn't do anything during starting or until the engine reaches a certain operating temp (I think this is partly to make sure the O2 sensor isn't giving data to the ECU when it is cold, and I think prone to giving false readings as a result). I think it needs a 15-degree celcius temp reading from the sensor in the pass-side timing chain cover before the ECU will allow the O2 sensor to have any effect. You may know this already, but FYI since you mentioned it...I don't think the disconnected O2 sensor has any impact.

And, keep in mind that the O2 sensor compensates for any mixture adjustments to keep the air-fuel ratio at Stoichometric at idle or cruise, so my guess is it is disconnected because the PO wanted a rich mixture and didn't want the O2 sensor counteracting his setting. But, based on my own experience (because when I disconnct my O2 my car idles fine), it could also be disconnected because by doing so other problems are covered up (like a vac leak, which I think is my problem, as the O2 tries to counter the resulting lean mixture). Might be worth plugging it in just to see how the car behaves (again, shouldn't change your start condition at all based on how the ECU is programmed).

All that said, I think cold WUR pressure could be the culprit here. Let us know. -DDD

[T77911S]

the most critical thing when starting to sort out CIS is AIR LEAKS.
this is one of those places it does not hurt to spend money first since there are so many places for air leaks.
get the hard stuff out of the way:
injector sleeves
injector seals
manifold gaskets
intake runner boots
vacuum hoses
vacuum hoses to and from the oil tank
seal on oil tank cap
O ring under throttle body

you have some plumbing behind the CIS that you cant see and dont really need to replace but should be inspected.

oh, also a quick way to check out the vacum hose to the brake booster and the booster is to use a hand vac pump and pull a vacuum to see if it holds.

then you need to check fuel pressures.

[Cdnone1]

subscribed

[MotoSook]

Originally Posted by 5String43

These symptoms are precisely those that present themselves when I cold-start my '82. I've been suspecting that the WUR probably is the culprit, but when I called the guy who runs Fuel Injection Corp. to see about a replacement/rebuild, he told me that the fuel distributor more likely is the problem. Haven't done anything about this yet. But I'll be very interested in seeing what you find.

He's a magician if he can narrow it down to the FD from the above.




Anyhow, to the OP, the fact that you are operating fine when warm is a good sign. How does the CIS behave under deceleration, like fron 4K+?

It sounds like your cold start injector is working, hence the momentary cold start. The fact that you can't easily coax it to idle high with throttle input makes me suspect an high control pressure when cold. With an open throttle, the airplate still has to lift to add air and fuel (the air plate indirectly lifts the fuel plunger).

If you have the CIS pressure tester, you really should post numbers. From cold it shouldn't take you more than 30 mins to get the numbers needed to troubleshoot.

An unplug O2 sensor won't make the CIS go crazy and you won't need to over compensate throwing the system out of whack....it simply takes one variable out of the system.

You should know as an engineer....there is no guessing to solve a problem...get the fuel pressures.

[Bob Kontak]

Originally Posted by T77911S

you have some plumbing behind the CIS that you cant see and dont really need to replace but should be inspected.

Check the brass tube that enters the throttle body from the driver's side - behind the electric heater blower fan. I have my blower removed and I can reach the union easily.

My car had the symptoms you describe when this pipe came loose. The single backfire was really pronounced, though.

[kiwiokie]

Originally Posted by 5String43

These symptoms are precisely those that present themselves when I cold-start my '82.

Same problem on my '83SC. Always idles fine after the 3rd start so I have not been too concerned but am intrigued as to the cause.

[psalt]

tirwin,

Welcome to the list. The 911SC's CIS lambda is different from K Basic and complete information on how the system works is hard to find. Probst's book is a good start, but it is missing a few 911SC specific devices and the Bentley manual has several errors in this section. My advice is to search this site for "CIS lambda" and read the archives. You will also find lots of advice about K Basic on this site, but most of it is irrelevent to the problem you are describing. The cold running stumble and stall is a common problem with the later lambda SC's because of the revised leaner fuel calibration in combination with the usual change in vacuum from wear, compression drops and leaks. My experience is that this is a tuning issue, not a parts swapping problem. Others with no experience may give you contradictory advice. I have solved this problem on several cars over the past 20 years and here is what worked for me. First check that your fuel pressures are in spec and that the lambda ECU is producing the correct cold open loop duty cycle with a dwell meter. The O2 sensor is irrelevent until the engine reaches 15C, but the ECU must be pulsing the FV at the default duty cycle . The 02 sensor is also irrelevent to performance, you can run the engine without the benefits of closed loop, but you need the FV and ECU to get anywhere near factory driveability. My advice is to get the system working as designed, leave the O2 sensor connected for mileage at cruise and cat life, and adjust the open loop mixture for best performance at WOT. Test for vacuum leaks at the injectors, intake boots, gaskets and throttle shaft with an unlit propane torch. Wiggle the injectors while testings. Most engines at this age will have leaks at the injector sleeve Orings. The seals are protected from the heat of the runners, but the sleeve Orings bake. You need to remove the stakes in the runners to replace the sleeve Orings and the procedure can be found in the archives. All of the cars I've seen with this problem that you are describing have had perfect pressure, a working WUR, and a working lambda system. Once you are sure that the system is operating to factory spec, change the spark plugs to the normal Bosch WR5 plugs. Warm up the engine by driving at least 10 miles. Disconnect and plug both vacuum lines to the distributor, reset the idle to 950 rpm with the large thumb screw on the throttle body and set the ignition timing to 7 degrees BTDC. Reconnect the front advance line (orange), leave the rear vacuum retard line (grey) disconnected and plugged. Reset the idle to 950 rpm. Hook up a dwell meter to the lambda test port under the left side engine cover and use an allen wrench to set the mixture to 20-30 dwell at idle. Let the car cool overnite and try a cold start. You really only get one shot at this test a day because subsequent attempts are altered by the CSV spraying each time you crank the engine. A 911SC with CIS lambda can be tuned to fire instantly on the key, settle into a steady 1500 rpm cold idle without touching the throttle, have good WOT performance and give near 30 mpg, but you need a properly working ECU, FV and O2 sensor to have it all. The cold running issue is a known problem for many years. One of the other parts vendors that also sold and serviced these cars new, markets a "super switch" with different specs from another car with CIS lambda to improve cold drivability. I have tried several of these, they do help with other issues, but not the initial cold stumble problem. Good luck.

[tirwin]

Well, boys... got a head scratcher here.

As I said in my original post, my plan was to get the new ignition switch assembly installed before I began troubleshooting the CIS system. Well, late last night I finished installing the new ignition switch assembly. When I got done I hooked up the battery and turned the ignition. Weird thing is that the car 'sounded' different when I cranked it. It still sputtered and died. Tried again, same thing. On the third crank, it started right up and settled at a perfect idle. Whoa!

Thinking I might have been tired or imaging things, I decided to try again today with a cold start. 1st crank sounded much better than usual, but died after about 5s. 2nd crank and it started right up and idled perfect!

This seems to be a new and possibly important symptom. New ignition assembly means new ignition switch, so that eliminates a variable but why would that improve the start up?

Last night the temp was 57 F. This afternoon it was 75 F.

Is this coincidence, significant or am I crazy?

[tirwin]

Originally Posted by tirwin

Thinking I might have been tired or imaging things, I decided to try again today with a cold start.

Meant to say 'imagining things'.

[psalt]

tirwin,

Yes, it is imagination. When you start a cold engine, it needs a very rich mixture to fire, down to a 6:1 AFR, depending on ambient. With CIS, the CSV supplies the extra fuel, but once the engine fires, other CIS components are supposed to take over and supply the rich mixture needed to make what is called the transition period to cold running. If your engine fires then stalls, it is not working as designed. You can imagine that repeated cranking attempts energize the CSV, and more fuel is dumped until there is enough hanging around to make the transition to cold running. This is an uncontrolled variable that differs in every engine. If your engine fires immediately, you get little help from the extra cranking fuel, if your engine take a long time to fire, you get more. All of this occurs before the heater in the TTS terminates the CSV. There can be several reasons that your engine can no longer supply a rich enough mixture to make the transition as designed, you have to test each one separately if you want to find the solution.

[Bob Kontak]

Paul,

There may be something to what he is saying regarding the new switch in post 12. I am thinking there could simply be better electrical connections carrying more voltage to the engine.

Last year, when I took your advice to plug the retard hose on my 81, it now normally starts on the first turn. If it sits in the garage a couple of weeks and the battery gets a little tired it will do the start and poop out in 4-5 seconds. Then start and idle fine. If I charge the battery up it starts first time every time.

I don't think the switch is the final solution based on the first post info, but it may have helped some.

[scotricker]

thanks tirwin, for volunteering to post results, etc in your first post. I have subscribed and have been checking back. My 911 is a 78 so the CIS components are slightly different, but I will keep checking here for more mystery solving ideas.

[j911brick]

Don't overlook the AAR and Thermo-time switch, and I'm leaning towards the TTS.

[trond]

A simple solution would be to replace the ageing CIS injection with EFI from tbitz racing. Its only a long days work with no specialist work at all, good economy (you can sell the old CIS) and you'll get both improved throttle response and slightly more power. I went this route when my CIS needed work and would choose the same solution again.

Look into it

[jmod]

Just thinking about a way that the ignition switch could affect the cold start, do you have a standard starter motor or do you have a hight torque substitute? If you have a high torque starter the wiring may have been altered in a way that feeds the cold start valve and thermo switch from the ignition start contact. Some high torque starters don't have the extra contact for the cold start of the standard starter.

[tirwin]

Fellas -

Apologies for the delay in getting back to this topic. I got tied up with work and family stuff last week and didn't get back to this as quickly as I hoped.

Tonight I completed the CIS tests. The brief synopsis is that it looks like I have a bad WUR.

Here are my results:

1) System pressure test: 4.7 bar normal
2) Cold control pressure test: 2.5 bar @ 26 degrees C (81 F) - normal
3) Warm control pressure test: 2.5 bar after 3 minutes - abnormal
4) Residual pressure: 1.5 bar @ 10 min, 1.4 bar @ 30 min - normal

So, looks like the WUR is bad. I should've had a reading between 3.4-3.8 bar. Looks to me like the bi-metallic strip isn't warming up like it's supposed to. Can this part of the WUR be replaced? Looking up the part on our host, it doesn't appear new ones are available for my year model. Suggestions? Am I on the right track?

I will post my test plan/procedure stuff in a separate post shortly.