CIS hard cold start and idle
 

Hi all, I know these CIS topics are talked about ad nauseam, but it seems everyone has a paticular circumstance with these systems.
The problem I am having is that my car starts hard when it is cold, starting only happens when my foot is to the floor on the gas, without it it seems like its flooding (rich). Once up to temp the car idles fine and starts fine again without any throttle manipulation. I followed Joe Engineers youtube video to tune the closed loop lamda system using an ocsiliscope and dialed in the FV duty cycle to spec (45-65%) at a healthy idle of 900 rpm. Timing is on checked multiple times. Once warm car pops out of the exhuast when blipping throttle. Popping does not occur when under heavy acceleration, but pops an extreme amount under deacceleration and engine braking. What gives? I am led to believe mixture is in spec from my instruments but clearly something is off.
Car Spec: 82SC 3.0 JE 9.5 pistons
964 cams
SSI exchangers and dansk sport muffler
New CIS airbox, hoses, vacuum lines, injector sleeves and orings, new FV and WUR, new TTS and 15 and 35 degree switches. Pulling o2 wire makes the car run significantly worse.
To me it seems that the car is too rich at cold start and lean when warm. Confirmed no vacuum leaks with smoke test and oil cap removal. Also to note in isolating varibles when tuning I will turn mixtutre screw rich when warm adjust idle , drive, and pops decrease, turn lean and pops increase in severity, no pops coming through the air box pop off valve. Any help is greatly appreciated! cheers.

1. Did you ever checked the WUR control pressure (cold and hot)? It must be in factory spec:
http://forums.pelicanparts.com/uploa...1726681713.png

2. How many turns do you need to close the idle screw? (remember original setting and set itback afterwards)
https://cdn4.pelicanparts.com/techar...s_med/pic1.JPG

3. Does the fuel pump run when you switch ignition on only?

4. Check the sensor plate height and adjust it if not in spec as described in the manuals:
http://forums.pelicanparts.com/uploa...1726682175.png

http://forums.pelicanparts.com/uploa...1726682206.png

5. Ignition timing is correct? Will say not too sharp, idle ~5°?

Thomas

Thomas, will check pressures when I get home, they were checked and within spec a couple months ago before I swapped in the previously mentioned new components. The fuel pump does run when ignition is switched on. when I plug the airflow sensor wire in the pump does not come on. currently leaving the wire unplugged. Ignition is set to 5 degrees. will also check idle screw for how far out it is and put a feeler in the sensor plate for correct spec, i have not disassembled any of those components but will ensure and adjust if necessary. Also might add that when the sensor plate is pulled down the engine sputters up more. Thank you for the pictures, are these from the bentley manual?
Joseph

Pictures taken from "Porsche Workshop Manual for the 911 from 1972-1983" and "Bosch K-Jetronic Workshop Manual". You can find both on the web.
The sensor plate height must meticously adjusted as the manual says!
After that you have to perform the basic adjustment as described above. The injection valves make a whining noise when injecting like in the following video:
https://youtu.be/4iWZdvOow0g

When he lifts the sensor plate manually you can hear it. Read the instructions on how to set the start of injection and the setting of the mixture screw in only this way! No exceptions. If the rest is in shape and spec your car will start hot and cold like mine (also other cams):

https://youtu.be/ynJhhmi4TCM

Thomas

As of this morning I hooked up my pressure gauges and tested system pressure which was at 66 psi so on the low end but in spec. Ambient temperature is around 50 degrees F.

Cold control checking with power to WUR unplugged and valve opened pressure did not change at all and stayed at the aforementioned 66 psi.

So clearly something is going wrong with either my pressure test or the WUR.

I followed the Joe engineer video and my Bentley manual on setting up the pressure gauges.

Still troubleshooting the issues with my fuel system. I am led to believe that the WUR is working properly once I put a reliable gauge on it and got in spec cold and warm control pressures.

Still car wants to take a second to fire up at cold start but its better for some reason now. The popping under decel is what I am looking at currently.

Bad decel valve? my car is an 82 so it has the pancake diaphragm on the right side of the engine with the small vacuum line going to it. It is plugged in and going to manifold vacuum on the back of the throttle body.

Will a bad one act as an air leak as well causing starting problems that get hidden by the lamda system kicking in adjusting things once the car has some temperature in it?

From what Ive read on here people plug them to get rid of rev hang which seems to add the pop and burble I am experiencing. Can this part be repaired with the squeeze it back together method? Is there a good way to confirm function with the part in or out of the car?

CIS troubleshooting…………
 

Hope you are not using a HF CIS pressure tester. Keep us posted.

Tony

It depends. Either it won't open and stay closed - then it's no vacuum leak. But exactly this will causes the popping in the exhaust under decel. The extra air from this valve helps burning the shorttime superfluous fuel in this manner which otherwise will pop in the exhaust. The valve opens up for just a second or so. You can also crosscheck this when releasing the throttle pedal and the rpms immediately falls to idle which it shouldn't. Instead the engine should stay above idle around 1500-1800rm for a second to make shifting a lot easier as this smoothes the gear change because of better fitting rpms transitions.

Otherwise - in case of a ripped diaphragm - then it's defintely a huge vacuum leak and supplies the engine in any situation with more air than required. You can work this partly by closing more the idle screw. But this won't solve the problem. Instead you move the whole system out of the working window with all negative outcomes: Less power, reduced efficiency, bad start etc.

How many turns do you currently need to fully close the idle screw? (reset it afterwards)

When the valve reacts too slow or is hanging it's probably glued with to fuel residues. Then I would flush it with some shots of silicon spray (no WD40 or similar oils as they will destroy the rubber diaphragm over time!)

You can take it out of the car and test it with a vacuum pump. JoeEngineer showed this on his page too:
https://j_o_e-e_n_g_i_n_e_e_r.com/porsche-911-cis-subcomponents-and-testing/ (without the underlines) or https://rb.gy/vave3m

Thomas

I was able to test the decel valve with a hand pump and it opened per spec at 15-18 inches of vacuum. So it is working as intended. cleaned out also with silicone spray.

Yesterday my CIS guage testing kit arrived and I proceded with checking the pressures, and yes Tony I was using a HF guage set that is now junk after one previous use where it worked fine lol.

Pressures were checked as following at 52 degrees F on a crisp montana morning.

System Pressure: 4.8 bar
http://forums.pelicanparts.com/uploa...1727452545.jpg


Cold Control Pressure: 1.4 Bar
http://forums.pelicanparts.com/uploa...1727452654.jpg

Warm Control Pressure: 2.5 Bar
http://forums.pelicanparts.com/uploa...1727452721.jpg

Residual Pressure:
10 min= 1.9 bar
30 min= 1.7 bar

I have a question about testing warm control pressure. The WUR has a vacuum port on it but it goes to atmosphere. Is this a vent. It is a 090. Just wondering if I performed the test correctly or missed a step?

How many turns do you currently need to fully close the idle screw? (reset it afterwards)

Thomas[/QUOTE]

Thomas, 3.5 turns open from fully closed oon the idle screw.

Hmm. This more I expected and as I heard from others. Standard setting for idle screw is about 2turns +/- 1/2 half turn.
So this let me suspect a too rich mixture due to a wrong sensor pate height setting. Could you post a picture from the downside of the sensor plate height and lever arm and best also from top side (removed rubber boot).

Ignition timing is in spec also?

Thomas

I put a light under the sensor plate when taking overhead picture. Pushing or pulling on sensor plate is smooth and not binding on anything.


http://forums.pelicanparts.com/uploa...1727459673.jpg
http://forums.pelicanparts.com/uploa...1727459673.jpg
http://forums.pelicanparts.com/uploa...1727459673.jpg

Looks not too bad.
Does your fuel pump run when just switching ignition on?

Yes it does. The plug on the back of airflow sensor is not plugged in. Pump won’t run when cranking it it’s plugged in. I was told I could unplug and just leave until I can figure a solution for the bad plug.

No. This is mainly a security feature which makes sure stopping the fuel pump in case of a stalled (hot) engine om a car crash with broken fuel lines to prevent engine fire...I wouldn't miss that anytime.

Nevertheless it also make the troubleshoot even harder because it also hides the (in)correct adjustment of the CIS...
Plug it back again. Then make a basic adjustment of the CIS as follows:

https://imgur.com/Cv2WRfv.png

Also make a crosscheck if your sensor plate is adjusted as the literature teels:
https://imgur.com/d9U8aW1.png

These both correct adjustments make the difference between a proper running (and starting) engine and a piece of **** when all other things have worked...

Thomas

ok I plugged in air sensor back in. set basic micture with injectors in little bottles.

set idle screw to 2 turns.

Car does not want to run. Will crank and then stumble at 200 rpm.

Can you explain what the air sensor plug does other than keep the FP off when the ignition is on and the engine isnt running? I think it has to with something. As when I turn the key the FP will run. I lean out mixture screw, turn key and no FP or FV running. Then go back to plate and move up ever so slightly it fires on again.

If it is that close to the trigger point is the mixture extremely rich?

No, the mixture screw is nothing else like fine tuning for the sensor pkate height. When you look closer on the lever from the sensor plate this not only a single lever arm but a lever system of two levers with a connection rod outside the pivot point. This results in an overproportional transmission of the travel from the sensor pkate to the plunger of the fuel distributor. That's this - I call it working point or injection begin - is so sensible because of this non proportional translation the system is able to enrich the mixture on higher speeds.

That particular point where the sensor pkate is being lifted due to the idling or starting engine defines the injection begin = where the fuel pump starts to pump gas and the engine starts or keep running.
When your car won't start with this basic setting then there's something else is wrong. You have to adjust both sensor plate height as well as the injection begin meticously as the documents describe. The basic adjustment procedure is taken from the 911 1972-1983 Porsche workshop manual, the sensor plate procedure is described in the Bosch K-Jetronic Workshop Manual. Both are linked here in the forum or available on the web.
So: With non running engine and ignition on fuel pump mist not run (security switch plugged in). When engine cranks the sensor pkate gets lifted so less but that enough to start the fuel pump and supplying the lambda control incl. the fequency valve if present.
That working point defines the working window of the whole CIS as this defines the (math.) function for the mixture.

Here's a video showing how it sounds when only the fuel runs when (security switch unplugged for demo) and then when the injection valves barely opens up and injecting fuel. You must follow all the steps described in the procedure precisely.

https://youtu.be/4iWZdvOow0g

Explanation of the lever system:
https://forums.pelicanparts.com/porsche-911-technical-forum/1158127-air-sensor-plate-height-why.html#post12203484

When you move that working window either too low or too high you enrichen or lean the mixture in wrong (working) areas (of the CIS) resulting in bad performance, poor mileage and bad up to impossible starts. Worn injection valves, bad / old fuel and other issues may worse the situation.

There is an excellent video series from Kurt from KlassikATS about the 2,7 911 with CIS incl. Troubleshooting and adjusting it:
https://www.youtube.com/playlist?list=PLkWFofpgz2OgF9FdliKnlzuY0pygqRkMj

Thomas

I had been waiting for my own smoke machine and fixed a rocker shaft leak in the meantime. Ive ensured there are no air leaks in the system. I also researched the above post about the lever arm height and I went to adjust the sensor plate height but it was in spec as described and checked with a 0.1mm feeler guage. I am still at a loss as to what the problem is. All fuel pressures are fine and no vacuum leaks. Adjusted valves again to make sure that all mechanical systems are in spec. Also made sure timing was dead on at 5 degrees BTDC. When cold starting it immediatly wants to fire up and then die. Wont start again until I floor the accelerator and keep it down until it warms up and then idles fine.

The car pulls hard to redline and has snappy throttle response, but when I lift off at any point the car starts to pop out of the exhaust.

My father has a 2013 911 with the sport exhaust and it does the same thing, but this popping doesnt sound right to me.

It doesnt pop when revving it cold. Richer when cold meaning no lean backfire in the exhaust?

Out of desperation should I check the order of fuel lines going to the FD? I assume they need to be in a specific order and I changed the lines out recently. Might be dumb to ask because they all send fuel at the same time based on the fuel piston moving up and down.

Out of ideas on where to turn next.

Long shot, but is the aar opening and closing? Getting 12v? Assuming an 82’ has one. Could be as simple as not plugged in or sticky and not opening or closing when hot or cold.

Sorry I can’t be any more help just throwing out some free easy ideas to check.

I checked mine with a mirror once and my phone a second time..

No, thats a good idea. I will certainly check it out and test. Should be fairly simple to pull the lines off and get a good look to see whats happening in there. Thanks for the suggestion!