When the engine is cranking on the starter, the speed of the engine is not 'smooth', it is accelerating/decelerating with the compression. This makes it hard to get a stable reading from the strobe.

Digital strobes with an advance feature are worse in this respect as they need stable RPM in order to calculate the timing for the strobe to flash.

Right, but it's all relative to the way the CDI unit triggers, e.g. the Bosch CDI signal requires a sharp negative
edge pulse relative to pin 31 (scope ground). Some other CDIs use a positive sharp rising pulse edge.

Well given its the correct permatune for year I have to assume the box uses the same pulse trigger conditions as standard? So how can the pulse trigger apparently move so much?

Johnny - agreed cranking cyclic acceleration is not uniform however light gives a stable repeatable signal. Also the sound of the engine as it cranks, fires and tries ro run supports the timing light reported timing I. e Some of the combustion is actually slowing the engine down due to retarded spark.....

Is the Permatune you have programmable for igntion advance?

If the trigger polarity is correct, then you need to check the CDI output pulse for multiple sparks.
That will cause strange timing degree indications from the timing light. By the way, how is the tach
working, i.e. does the needle jump around? Raise the RPM to 2K and the spark period should be stable
with 10ms pulses at the CDI output.

Can you even get the car running? You should be able to get the static timing close enough to get it started. If not, your problems are elsewhere. My guess is that you don't actually have erratic timing but it doesn't run well enough to get it consistent. Again, check that relay. My 82 would barely start at time and sometimes not at all. The fuel mixture would vary wildly which could cause the backfiring. It's a cheap part. Try it.

OK thats interesting. I did reconnect the tacho yesterday and it does jump around a fair bit when blipping throttle, it also did this before.

Jonny - NAFAIK, just a replacement standard permatune.

Lambda system is working correctly.

So the collective thinking is that timing via light is not accurately represented during cranking and as such cannot be trusted? Poor starting is down to another issue and only set / check timing with hot engine?

Yes could get the car running, and when running it ran well. The hard part is getting it started in the first place. From the cold I don't think the lambda circuit has any input (its not doing anything until the sensor has warmed up)?

im not really saying the lambda isn't working. The relay supplies power to the AFR sensor as well as the frequency valve. I'm not guaranteeing it will fix the issue but it's cheap and a common failure that can cause the issues you are having. It's worth a shot.

The Lambda ECU only monitors the oil temp switch and the O2 sensor. So for troubleshooting, one can
just disconnect the ECU and richen the mixture at the FD. Then the mixture is just affected by the WUR
for cold starting/running.

Lamda circuit runs freq valve at 50% duty when cold (open loop). When O2 warms it adjusts duty based on O2 input, a throttle switch or two and temp sensor(s?). I usually defer to mysocal911, but pretty sure off idle and WOT switches come into play.

There is a chart available in these forums that shows what it does and when. If you wish, I can find it. It's in tirwin thread.

Yes, the Lambda ECU does have an accel relay input for enrichment, but since the OP is troubleshooting
just a simple running condition, the accel relay will not be an issue. Furthermore, the ECU does not
have any temp sensor input other than the oil temp switch input. Please provide a wiring diagram
where a temp sensor is shown as a direct input to the Lambda ECU, if that's the case. The Lambda ECU
is the same basic design as used by M/B for their CIS system during those years. Other than for fuel
enrichment under accel and emissions, the engine will run fine without the ECU once the mixture is
enriched after the ECU is disconnected.

Yep, temp switch is not a temp sensor. Thanks for clarifying. It is what I intended to relay but wrong terminology. It is the only temp feed to the Lambda box.

For discussion only - there appears to be one temp switch to the main box and one to the Control Unit Acceleration Enrichment. I made an assumption they both went to the main box.

There is also a throttle switch to the Lambda box and a micro switch to the accel enrichment unit.

Here's the links to the PP factory wiring diagram.

http://www.pelicanparts.com/911/911_parts/Electrical/911_electrical_82SC_Part6-1.jpg

http://www.pelicanparts.com/911/911_parts/Electrical/911_electrical_82SC_Part6-2.jpg

That's right, i.e. two enrichment inputs to the ECU. One enriches more than the other.
At the moment I don't remember which has the greater effect.

So I will update with a status, much testing, some results still not right however!

I am reasonably happy the ignition system is now functioning correctly.

I have leak checked the s**t out of the intake system and only found a small amount of bypass around the pop-off seat, this is with positive pressure in the manifold. I bought a smoke machine as any soapy water type testing is a bit of a waste of time IMHO, particularly with everything still in the car. I've plugged brake booster feed. When the engine is running. The pop of seat is sealed.(some thick grease and of course the vacuum pulling it against the seal).

I have reset WUR pressure to within spec; 1.7bar at 15c and 3.4bar fully hot. It was 0.5 and 4.1 when I started testing.

System pressure is around 4.3bar, bottom of spec.

I have checked CSV clicks when 12v applied. I've checked Thermotime switch for earth path.

When fully hot I had to adjust the mixture half turn rich, using my lambda sensor in place of the OEM. Set to 0.95lambda.

Duty cycle at frequency valve 65% at start, 50% after a bit of engine heat. Sensor not connected.

Current condition: It will not start easily cold or hot. It will cough into life after around 20-30sec cranking with some manipulation of throttle pedal.

When it's hot it will rev nicely and hold idle.

If I lift the airflow plate very slightly AFR gets richer, if I pull it down it gets leaner. It will do this at any engine speed I can hold with the throttle. If I slowly rev the engine in neutral it will hold 0.97ish lambda.

I measured around 10inHg (35ish kPa) of vacuum at the brake booster port at idle, nearly 30in when revving. If I remove the oil filler cap at hot idle I get a slight drop in rpm.

So I think I have confirmed:
No air leaks.
WUR is functioning as intended.
Fuel flow is all within spec (main flow into system, return to tank, and fuel flow from FD to WUR.)
Residual pressure is fine.
Lambda system is working as it should.
Airflow plate and FD working as intended?

Why will it not start nicely? CSV is obvious culprit but when doing a hot start the CSV is not activated anyway?

Today I tried resetting cold control pressure back to how it was I.e 0.5 bar (so richer at start). It still took 20-30sec cranking to get anything, then it was indeed richer at around 0.65-0.7 lambda.

All thoughts welcome!

CIS start up........
 

James,

Do the basic CIS start up preparation. And you will find out why you can not get your engine to start and run when cold.
a). WUR, FD, CSV, AAR, fuel injectors, FP, and air box all tested good.
b). Ignition timing, cam timing, valve clearance checked.
c). CDI, ECU, relays tested good.
d). Others.

One of the most neglected test is confirming that the air box and pop off valve could hold vacuum. You can not do this test with the air box installed. Unless your air box is new, it could be leaking. If you don't test and confirm it's condition, you would be hoping everything is OK. The trick for a successful CIS troubleshooting is reducing the variables to the minimum. You are very systematic in your work.

Ask this question to yourself, "What could possibly cause this problem?" If you could identify the most likely culprits, you are on the right direction. Identifying the 'culprit' is the key for a successful investigation. Sometimes you get lucky. But you can not depend on luck all the time.

This is what I would do if I am doing the test:
Disconnect all the fuel injectors including the CSV?
Place the injectors in a bottle or vessel to collect fuel, if any.
Connect an inductive timing light.
Secure everything (bottles and injectors) from falling.
Turn the ignition switch to START. Observe very closely the fuel deliveries of the 7 injectors. Repeat a couple of times. The amount and characteristics of fuel delivered is critical during cranking. How strong was the ignition signal?

Keep us posted. Thanks.

Tony

Hi Tony.

I think I have verified basic operation. System, control pressures. Air flap setting. Hot idle speed and timing, hot O2.

Now when it's running hot it runs well, although this is also stationary, I can't drive the thing yet with no plates!

However you point ref most likely culprit is most correct. June 2016 fully running perfectly starting to August 2016 complete mess.

So my most 'likely contendors' were fuel flow blockage or air leak.

I think fuel flow and pressures check out.

So back to airleak! Everything so far points to a very lean condition. From initial full backfires and no starting at all to now both richer control pressure hot (3.4 vs 4.1) AND a half turn of mixture screw to get hot condition to 0.96lambda.

Everything I've read says that a vacuum leak that prevents a start is a large leak, like one intake runner boot not connected.

I've pressure tested the airbox, lid on and off, vacuum tested it and finally checked with smoke machine. Pressure and vacuum were inconclusive due to limited space to see all the potential areas. Smoke test showed very small leakage at pop off valve.

So I'm struggling with the idea of a vacuum leak big enough to prevent starting?

However next phase has to be compression check and full removal of cis, unfortunately!

Question - assuming for a moment ignition side is good, shouldn't the csv provide enough fuel on its own at the intake pipes to at least allow an initial fire or two? Regardless of what the FD etc is doing?

Many thanks for continued support......

So I pulled the whole cis assembly today. Time to stop messing around and have a proper look!

Actually wasn't to bad a job.

CIS air box tests..........
 

James,

Now that you have the CIS unit off the engine, test and confirm that at the airbox and POV are not leaking. Some pictures to share:

http://forums.pelicanparts.com/uploa...1487798293.jpg

http://forums.pelicanparts.com/uploa...1487798392.jpg

http://forums.pelicanparts.com/uploa...1487798473.jpg

You don't need a lot of air pressure to do the test. A bicycle pump or tire compressure is more than sufficient. Take notice about the inflated balloon in the set up. If the balloon maintains the pressure for a long time you will confirm that the airbox is not leaking. A more appropriate test is a vacuum test. If there is an air leak, perform the pressure test with the air box submerged in water. Keep the pressure low (2~3 psi.) to prevent the POV from opening or releasing air. Or slightly press down the POV, works better with vacuum.

You mentioned about the fuel mixture going LEANER. Either you are getting unmetered air or less fuel is being delivered. Pick your choice and locate the culprit. In your case, it could be unmetered air. But we don't know that as a fact. So test and confirm.

Tony

Can you give an idea of the amount of fuel delivered by one injector when cranking for 10sec? I did test the two front injectors into pots and received a very small amount of fuel, like maybe 10-15cc.

Will be stripping components and testing. I think a new airbox makes sense as well but we shall see.