CIS Component Testing - Long

[arbita1]

I've always had idle and cold start issues since I bought the car. When I first bought the car the O2 sensor was unplugged and the car had cold start issues (stumbling, backfiring, low idle). Since then I've gone back and forth between CO set to factory settings with O2 sensor plugged in and CO at 3.3-3.5% with O2 sensor unplugged. I've gone back and forth a few times now. It always seems like the car does better with the O2 sensory unplugged and CO set to around 3 - 3.5%.

Just recently I replaced the cat with a Dasnk premuffler. Since I had to check the CO after I installed the premuffler, I figured I'd go back and try the O2 sensor again with CO set at 2.0% (which is factory setting for non cat). Again I was back to cold start problems and it seemed to run very lean (rough).

Today I decided to go back to O2 sensor unplugged and set the CO to 3.3%. Immediately it was running like it should again. I'm not sure about cold start since I haven't been able to check that yet, but from past experience it will probably be better.

On top of all this, my idle has never been very consistent. It seems to move around a bit when the car is warm from 950, then 1000, then 1050, then maybe back to 950.

So it's alway been my intention to go through my CIS system to see what I can find. Here are my results thus far...

System Pressure - 5 bar (with-in spec)

Cold Control Pressure - 2 bar at 85 deg. F outside temperature (this came up lower than I was expecting...per Bentley I should have been between 2.6 and 3.1)

Warm Control Pressure - 3.6 bar (with-in spec)

Residual Pressure - 1.8 bar @ 10 mins, 1.7 bar @ 20 mins. (within spec.)

Timing is set to 5 deg. BTDC @ 950 RPM

I could not verify if Centrifigal Advance is within spec. (I didn't have a partner.) But I was able to see the timing mark moving while I raised the throttle from inside the engine compartment. So I think it is functional.

Vacuum Advance - does not seem to be working

Vacuum Retard - does not seem to be working

AAR - was partially opened during cold and fully closed when warm.

Duty Cycle seems to be non functioning. The dwell meter showed a constant 21% with and without the O2 sensor plugged in.

I checked for leaks with a propane torch. But my results were inconclusive because I did it before I richened up the mixure to 3.3% and it was idling very poor. So I have to try this again.

Sorry for the long read, but I appreciated any comments. Thanks.

[Gogar]

Make sure to check your Lambda relay too.

[T77911S]

+1
fues 18

if the FV is not working for what ever reason, that is why the PO unplugged the O2. you did not mention that you have dont a complete tune up. i would assume you have.

air leaks. lots of little places for unmetered air to sneak in. intake gaskets, are the intake bolts tight, the rubber boots on the runners, O rings and injector sleeves, blown air box. do you have a pop off valve, is it seated properly. vacuum lines. if the lines to the dist are connected, the diaphragm/s may be leaking.

without the O2, it will need to be set richer, like 3.3%.

cold starts- did you check the cold start valve and the thermo time switch?

a low cold control pressure is better than a high one.

[arbita1]

Originally Posted by T77911S

+1
fues 18

if the FV is not working for what ever reason, that is why the PO unplugged the O2. you did not mention that you have dont a complete tune up. i would assume you have.

air leaks. lots of little places for unmetered air to sneak in. intake gaskets, are the intake bolts tight, the rubber boots on the runners, O rings and injector sleeves, blown air box. do you have a pop off valve, is it seated properly. vacuum lines. if the lines to the dist are connected, the diaphragm/s may be leaking.

without the O2, it will need to be set richer, like 3.3%.

cold starts- did you check the cold start valve and the thermo time switch?

a low cold control pressure is better than a high one.

Originally Posted by Gogar

Make sure to check your Lambda relay too.

Lambda Relay - I believe it is functioning. The fuse that provides power to it is connected to the clock and dome light. Both are functioning. I'll verify the relay is working by pulling it out during idle and see if there is any change.

Frequency Valve - it is functioning. I can hear it and feel it buzzing. As far as tune up goes, my spark plugs are new, plug wires, distributor cap, rotor, and fuel filter all replaced about 3 years ago.

Air Leaks - yes I definitely need to verify this again. Is there a better way than using propane? Should I try carb cleaner too? I do have a pop off valve (installed by the PO) and it looks fine. Yesterday, I check all my vac lines that I could reach. A few of them were loose, so I tightened them up. The ones to the distributor I replaced with new hoses.

Cold Start Valve/Thermotime Switch - I didn't check those yet. I figure they are working since I don't have any cold hard startup issues. How do you even access the cold start valve to check it???

So my next plan of attack is to check the following: verify centrifigal advance, check again for vacuum leaks, check thermotime switch, check O2 sensor relay, check O2 sensor, check 15 deg. temperature switch, check 35 deg. temperature switch, check throttle position switch, and check 15 deg. throttle position switch.

[T77911S]

i thought idle and cold issues were in your first sentence???

[arbita1]

Originally Posted by T77911S

i thought idle and cold issues were in your first sentence???

Yes. Cold running issues. I think I worded that wrong in the first post. I don't have any starting problems.

My understanding of the cold start valve is that it only operates with the starter (when the thermo time switch tells it the engine is cold). Since the car always starts right up from cold, I figured it's all functioning.

[matt demaria]

For the 911 SC (american version) 1980 to 1983, the CO should be set (with lambda disconnected ) at 0.62 %. When you connect the O2 again the CO will drop to 0.22%. This is measured before the cat.
1979 and earlier is set to 1.5% to 2%. Do not use a Gunston analyzer, it is not accurate. It reads too rich by 1 to 2 points.
As long as the system goes into closed loop (.22%) the idle will stabilize when warm.
3% CO is too much gas. You run the risk of borewashing the cylinders and ruining the rings.

[arbita1]

Originally Posted by matt demaria

For the 911 SC (american version) 1980 to 1983, the CO should be set (with lambda disconnected ) at 0.62 %. When you connect the O2 again the CO will drop to 0.22%. This is measured before the cat.
1979 and earlier is set to 1.5% to 2%. Do not use a Gunston analyzer, it is not accurate. It reads too rich by 1 to 2 points.
As long as the system goes into closed loop (.22%) the idle will stabilize when warm.
3% CO is too much gas. You run the risk of borewashing the cylinders and ruining the rings.

Don't forget I am running a premuffler (not a Cat), so the CO spec should be 1.0 - 2.0% for that.

Unfortunately, the Gunson is all I have. By their own admission, they indicate it has a error margin of .5% (which is substantial when trying to set your CO to .6%). I usually try to shoot for the middle of the range this way if it is off, I should be still within spec.

My problem was that when I had the CO set to 1.5%+/- it idled very rough and seemed to be too lean. I don't think my Lambda system is doing it job. Plus, the duty cycle is 21% with the O2 sensor plugged in or out??? Something is definitely wrong with that. I think I need to go through all the lambda system components and verify they are functional.

[Scott R]

0.6% difference is not going to cause this kind of issue. I can be 0.2 to 4.0% and the idle will be about the same. Your cold control pressure is a tad low, which is going to effect your mixture setting in the end even after it's warmed up. You need to get that sorted by adjusting the WUR. (check the wur resistance first to insure the heater element is still working)

On to your rich mixture issue, (setting it rich it runs better) that's going to be a leak somewhere. Your FV is obviously working as the car will barely run with it disconnected. So it's time to start checking for vacuum issues.

1. Boots
2. Injector bung o-rings
3. Injector o-rings
4. Brake booster line
5. Oil tank line and cap
6. throttle bypass line
7. Air bypass lines to aar
8. Decel and aav diaphragm connections
9. Cruise servo connection
10. Throttle body o-ring
11. Pop-off valve o-ring
12. Air boot


I'm sure I'm missing one or two here. However, remember that this system is essentially run on vacuum. How much you have, or don't have, is going to effect the function of the sensor plate and ultimately the amount of metered fuel.


Start with the basics and the rest will come into line.

[arbita1]

Originally Posted by Scott R

On to your rich mixture issue, (setting it rich it runs better) that's going to be a leak somewhere. Your FV is obviously working as the car will barely run with it disconnected. So it's time to start checking for vacuum issues.

1. Boots
2. Injector bung o-rings
3. Injector o-rings
4. Brake booster line
5. Oil tank line and cap
6. throttle bypass line
7. Air bypass lines to aar
8. Decel and aav diaphragm connections
9. Cruise servo connection
10. Throttle body o-ring
11. Pop-off valve o-ring
12. Air boot


I'm sure I'm missing one or two here. However, remember that this system is essentially run on vacuum. How much you have, or don't have, is going to effect the function of the sensor plate and ultimately the amount of metered fuel.


Start with the basics and the rest will come into line.

This lean issue at cold starts I'm having is very similar to symtoms I had on a old Ford Explorer I used to have. The problem was a vacuum leak from the manifold gaskets. I'm thinking that maybe there is definitely a leak that gets seals up as the engine comes to operating temperature.

I was never able to find the leak with carb cleaner on the Explorer, but the manifold gaskets were a known weak point on that car. So replaced them and the problem was fixed. If I can't find it, I may have to just replace some of the old rings, etc. that are known problems.

[boyt911sc]

Matt,

I'm impressed in your patience and perseverance in tackling your CIS problem/s. You have done a lot of work and spent an enormous time and effort in locating the culprit/s. But you're not closer from where you started several weeks ago. Some just mentioned to you about fuel pressures and vacuum.

Why are you adjusting the mixture screw? Tinkering this sensitive mechanical setting without verifying the integrity of the vacuum of your engine is a band-aid fix. Sometimes people get lucky!!!!! But what happens if you have a significant amount of unmetered air going into the system? Stop the guess work and determine where unmetered air is going in. Even a well running CIS engines exhibit some sort of vacuum leak/s some where. It is a question of how much unmetered is tolerable!!!!! Even a working lambda unit will not make your car run good if you have significant vacuum leak/s some where. You may or not have an air leak. Only a test could confirm this suspicion. Keep us posted.

Tony

[arbita1]

Originally Posted by boyt911sc

Matt,

I'm impressed in your patience and perseverance in tackling your CIS problem/s. You have done a lot of work and spent an enormous time and effort in locating the culprit/s. But you're not closer from where you started several weeks ago. Some just mentioned to you about fuel pressures and vacuum.

Why are you adjusting the mixture screw? Tinkering this sensitive mechanical setting without verifying the integrity of the vacuum of your engine is a band-aid fix. Sometimes people get lucky!!!!! But what happens if you have a significant amount of unmetered air going into the system? Stop the guess work and determine where unmetered air is going in. Even a well running CIS engines exhibit some sort of vacuum leak/s some where. It is a question of how much unmetered is tolerable!!!!! Even a working lambda unit will not make your car run good if you have significant vacuum leak/s some where. You may or not have an air leak. Only a test could confirm this suspicion. Keep us posted.

Tony

Thanks Tony. Yes you are right, it may be a band aid fix, but it allows me to enjoy my car in the interim, while I try to figure this out. Unfortunately I don't have lots of time to dedicate to it, so I have to try to balance between driving and wrenching.

What is the best way to search for vacuum leaks? I'm thinking I'm going to try spraying carb cleaning on the suspect areas.

[boyt911sc]

Originally Posted by arbita1

Thanks Tony. Yes you are right, it may be a band aid fix, but it allows me to enjoy my car in the interim, while I try to figure this out. Unfortunately I don't have lots of time to dedicate to it, so I have to try to balance between driving and wrenching.

What is the best way to search for vacuum leaks? I'm thinking I'm going to try spraying carb cleaning on the suspect areas.

Matt,

A good device is the smoke machine but most shops don't have this available. I searched and looked for one around but found one that was inoperable!!!!! So I built one using compressed air (5-10 psi) which is more than sufficient. It takes a few seconds of test to locate the leaking areas after spraying aqueous solution (detergent)!!!!!

This is probably the reason why I find CIS troubleshooting a piece of cake. I am not talented enough to work without the use of pressure gauges in troubleshooting fuel injection systems. There are numerous spots where you could inject the pressurized gas into the system. Pick your choice and block the exits!!!!! A latex glove or balloon work great for this test (????).

I wished you are closer, so I could the test myself for you. I'm away from home and won't be back home in three (3) weeks. So I could not post the picture of the home-made gadget we're talking about. Keep us posted.

Tony

[Scott R]

Tony where did you supply the compressed air to during testing? I have a "bench device" for testing airboxes but I've never tried an air test in the car.