Cold Start Woes

[Bikesmith01]

Hey Everyone!
I have a 1975 911s that is VERY hard to start when the car has been sitting for more than a couple days.

Here's what happens: The starter will crank the engine just fine, but I'll get zero combustion for a while. It's not coughing or almost starting, just cranking. Like the engine is getting literally no gas at all. (In the rest of this post I'll call that the "zero-gas issue") It will do that for anywhere from 30 seconds to a couple of minutes, and then suddenly start. Like it just suddenly found its gas supply.

It's quite rough at first, sputtering and coughing. Sometimes it will die a second or two after starting. When I try again it starts right away again. It doesn't go back to the zero-gas condition. After 15-20 seconds of running rough, it will very suddenly switch to running basically fine. Still sounding like a cold engine, but not like an engine with issues.

Once it warms up it runs amazingly. No issues at all. It starts warm perfectly.

Here's the weird part: The next day, stone-cold, it will start immediately.

It's only after two or three days that it will go back to the zero-gas condition where the starter revs with no combustion again. What's the difference between a one-day cold car versus a three-day cold car?

Starter fluid starts it right up so I'm pretty sure it's a gas issue.

I know I should check all the fuel pressures, but the problem is so weird I thought I might be able to get a head start on diagnosis by thinking it through.

The previous owner had the engine rebuilt by a Porsche dealership and then put 700 miles on it before selling it. That 700 miles was over a period of 5 years, so it did quite a bit of sitting. But I highly doubt there's anything substantially wrong in the engine based on the recentness of the rebuild and how well it runs once warm.

I pulled the time/temperature switch and confirmed that it's working properly, but I haven't checked the cold start valve yet. I could see that contributing to the initial rough running, but it doesn't seem likely to be causing the zero-gas issue, don't you think?

I did discover that the hand throttle lever was not adjusted correctly. It was set to bring a warm engine only up to about 1500 RPM. Adjusting that properly helped. It made the engine start better once it "finds its gas," but did nothing for the zero-gas issue.

I wondered if the fuel system was draining past the check valve over the course of a couple days, so I tried letting the fuel pump run for 30 seconds before starting. No difference.

I wondered if the injectors might be sticking closed after they sit for a couple days and it was taking a minute or two for them to unstick. So I tried pushing the air flow lever by hand while the fuel pump was running, to see if i could unstick the injectors. That worked!

So I replaced the injectors. It improved the running of the car more than I expected. Once started it goes to a smooth running condition much more quickly. Even when fully warm, the car idles more smoothly (it had wandered a little before), it has quicker throttle response, and it's making noticeably more power, especially below 3000rpm.

But the car still has the zero-gas issue.

Pulling the air filter and pushing the air lever by hand with the fuel pump running still fixes the issue. Car starts right away. Still runs rough for a few seconds, but pushing the air lever for a couple seconds fixes the zero-gas condition.

Just now I was wondering if a faulty fuel accumulator could be letting the system drain, but a quick bit of googling seems to indicate bad accumulators only cause hot-start issues, not cold-start issues.

So I decided to see if you guys have any thoughts. Fuel pressure? I can totally see incorrect pressure causing the 15-20 seconds of rough running once the car starts. But the zero-gas issue? Seems to me to be something more than just an incorrect pressure.

What do you all think?

Thanks! -Jon

[Funracer]

FD plunger sticking either at the bottom (no fuel) or top (too much fuel for start) maybe? You having to lift the plunger by hand to get it going makes me thing it is sticking at the bottom after shutdown.

[Schulisco]

Check valve at the fuel pump is either missing or non working any.more.

[PeteKz]

Jon, you have done an excellent job of troubleshooting so far and providing diagnostic information.

Fuel accumulator: You are generally correct that hard warm starting indicates a bad FA, however, if the internal diaphragm is really bad, it can leak so much fuel that it takes awhile for the fuel pump to build up pressure. If it looks like the original FA is still in the car, I can almost guarantee that it is bad, and you should replace it, even if it's not the main problem.

Further troubleshooting: Start the car and get it running smoothly. Shut it down. Wait 30 minutes. Have fire extinguisher handy. Crack open the fuel line screwed to the FA or fuel filter. it should squirt out a couple teaspoons of fuel. If it does, then the FA is holding residual pressure and both it and the check valve at the pump are good, and so are the pressure regulator valves at the fuel distributor and the WUR.

If it doesn't hold pressure, then you will focus your subsequent troubleshooting on those components.

Question: Your car should have a fuel accumulator with two screw-on fittings on the top, and none on the bottom. IS that what is in your car?

[47silver]

My 75.
I run the fuel pump for 10 seconds before starting if it is cold but if i do not pull up the left hand throttle lever for the cold start injection valve then it is hard to start. I would check the cold start solenoid first and that circuit.

[Bikesmith01]

Thanks for the replies, everyone!

After reading your replies I decided to try replacing the fuel pump check valve. It's so cheap I figured, why not? Maybe it would help, maybe it wouldn't. Either way I'd be happy to have a known-good part there.

I got my first surprise when I pulled the belly pan and found... No fuel pump! Where it was supposed to be I found a hose coming out of the fuel tank and going straight back into the car. So I turned the ignition on and traced the sound of the fuel pump to near the left rear wheel. It's rubber-mounted to a bracket welded to the tube that the torsion bar goes through. I don't know if that's an alternate original position or previous-owner shenanigans.

Also, the fuel pump is the type with no external check valve. The type that's $835 on Pelican. Okay, maybe I won't replace that unless I can prove that it's faulty.

Some specific responses: Funracer, The plunger lever feels very smooth. Not sticking at all.

Schulisco, I agree with you. Bad check valve at the pump seems like a reasonable cause for the issue, but the lack of external valve on the pump complicates things. I'm willing to spend $835 on a new fuel pump if I need to, but not as a diagnostic tool. I'll only do it if I can be pretty sure that's the issue.

47silver, I've tried letting the pump run for as much as 30 seconds before trying to start the car. It doesn't help. And I am using the throttle lever.

PeteKz, This is helpful info. Thanks. I'm tempted to replace the FA to see what that does. I don't know if mine is original. It's green and does have two screw-on fittings on the top, none on the bottom. I like your idea of testing for fuel pressure at the FA, but I wonder about the timing. Half an hour after running my car starts instantly. Six hours after running it still starts almost perfectly. I suspect that half an hour after running I'll absolutely get a squirt of fuel out of there.

It's the difference between one day cold and five days cold that I'm looking for. So I might try cracking that fitting after the car has been sitting for several days. Even after five days I would still expect to find the fuel line between the FA/filter and the engine full of gas, right? Just not under pressure anymore? I'm wondering if it's possible that I'll find the whole system empty of fuel after the car sits for several days.

I'll check that and report back.
-Jon

[Schulisco]

Jon,

The early 911's indeed have had the fuel pump next to the left rear wheel as yours too. From 1976 and on the fuel pump was mounted underneath the fuel tank in the front.



I also know that there are some types of fuel pumps have the check valve integrated within the fuel pump, e.g. some from Pierburg and also Bosch.

I posted here a link to a Bosch catalogue of fuel pumps, check if yours in there and if there's a replacement type available, often the compatible ones are much cheaper:
https://forums.pelicanparts.com/porsche-911-technical-forum/1125870-fuel-pump-question.html#post11790041

Here is a cross reference list of Bosch fuel pumps from the friends of Rennlist:
https://rennlist.com/forums/attachments/928-forum/849430d1404219342-help-with-bosch-fuel-pump-reference-bosch-fuel-pumps-cross-referrance-guide.pdf

Another Bosch fuel pump catalogue:
https://www.boschaftermarket.com/xrm/media/images/country_specific/gb/parts_1/parts_pdfs/fuel_pumps_2pp_aug_2017_lr.pdf

Here's a link to the Porsche / Bosch K-Jetronic Workshop Manual: http://members.rennlist.org/vilhuer/K%20Jetronic%20Workshop%20Manual.pdf

It tells that your 1975 car got a Bosch 0580254991 fuel pump and a Bosch 0438170015 fuel accumulator. Porsche PET says it's 911 608 110 02 or 03. Currently cannot find a replacement pump for that, but I want to encourage you to find one.

Before replacing the fuel pump and/or the fuel accumulator I personally check the pressures if all works as designed. Get out for a fuel pressure gauge for CIS and check first the control and system fuel pressure and then the holding pressure after shutting down the engine. It must hold approx. 2bars in the beginning dropping slowly down to 1 bar in an hour or so....refer to the Porsche / Bosch values in the manuals.
If these checks fail only then start to replace parts.
The pressure on an intact system releases slowly. After replacing the FP and the FA on my SC it took hours / overnight to release the pressure completely. Now after few years it still holds the pressure far more longer than the specification, but not that long like in the beginning.

The FA is a bit special. On most cases everyone follows the instructions from Bosch and Porsche to test the holding pressure. If it does, everything is ok. Really? Imho no. This is only the half the truth. The FA is moving any time when engine is running. It's working permanently! In Bosch manuals they write "it silences the noise of the fuel pump". I say - it smoothes the fuel pressure anytime! This means depending on pushing or releasing the throttle and how fast youpress or depress the pedal the amount of fuel varies that fast that the fuel pump cannot deliver the extra amount of fuel that quick. In consequence the FA also smoothes the mixture, saying it helps to deliver the fuel flow far more continuously and it guarantees are way more precise mixture and therefore a far better running culture of the engine. There are reports that after installing a new FA the mixture must be recalibrated...
Over time the FA corrodes inside with the consequence that it cannot move that freely and smoothly as before. So it starts to hang from time to time. You can check it by taking it out of the car and testing it on an air pump or better with a compressor: Close the output (middle) and put pressure up to 5 bar/72psi on the input (outside). Then suddenly release the pressure, e.g. by opening the pressurized line! The FA contains a huge spring inside, it must follow the released pressure immediately without any hesitation! Mine did not, it was tight, but it took more than a second before it suddenly released with a big bang in my hand... So it couldn't work any more as it should. This issue is in my eyes one of the most overseen issues on CIS engines and can cause the craziest and hard to find symptoms...

Thomas

[BoxsterGT]

The later Bosch Flow-Through Fuel Pump is a FAR BETTER option for your 75 911. Our host has this pump for around $150 if I remember correctly.

This Bosch Pump can also be installed on earlier 66 - 74 models using this hose routing above the rear crossmember. It eliminates the Hose clamps and multiple Steel lines Under & Through the Frame rail.



There are additional postings about this later pump upgrade here on Pelican.

I have made up a number of these upgrade hoses, all in E85 Rated hose that will last for the lifetime of your 911.

For more info, please email me.

len.cummings @ verizon.net but without the spaces.

Good luck.

Len

[PeteKz]

Jon: DON'T throw parts at the problem! Troubleshoot first. If you add new parts, just becuase it was easy, you complicate your troubleshooting because, 1) the new part may not be any better, and may be worse than your existing part, and 2) if the problem "goes away" you don't know if it was because you replaced the part, or changed something else in the process of replacing it.

It takes 1 minute to check for residual fuel pressure. Do it. Then you will know whether your FA is working and the other parts are holding pressure. You are probably right that they are OK, because the car starts fine after sitting for 1/2 hour or 5 hours, but verify it.

Yes, as Thomas noted, the CIS fuel pump is in the rear from 1973.5 to 1975. I don't think it really matters whether it's in the back or the front. So leave it where it is. it's mounted on two rubber isolators to reduce pump noise to the frame. My original pump was getting noisy and a couple times I needed to whack it to get it going, so I replaced it with the generic replacement from Pelican 4 years ago. It has worked fine since.

Len's suggestion to change the fuel line to one that comes up ahead of the rear crossmember is a good one. I second his suggestion. I used Goodyear branded rubber fuel hose and hose clamps on mine, but his nicely fabricated line is a better solution.

CIS gauges: Order them from Amazon. There's a whole set for about $50 that has adapters to almost every FI system on every car. So you can use it on your BMW's, Mercedes, gm, Ford, etc.

[fanaudical]

I wouldn't change a thing on the CIS until you have a full CIS pressure test done. Get a test gauge kit (this won't be the only time you use it) or find a local Pelican with the gauge kit to come visit and run through the test with you. Report back results and then we can help.

One thing that is critical on the cars with the fuel pump in the rear is to have open fuel lines and no crud in the tank. No gas symptoms may be a clogged pick-up.

[PeteKz]

And one other thing: If you don't have an in-line fuel filter between the gas tank and the fuel pump, install one. The less crud from the gas tank that goes through your fuel pump, the longer it will last. I installed one between the tunnel pipe exit at the firewall and the pump.

[47silver]

I noted that the supply line from the tunnel to the fuel pump that i replaced (it was the original) was installed up and over then down to the fuel pump. I think that the factory wanted to have a trap to make sure the fuel pump always had fuel on the suction side. My 75 pupm is in the rear. Never have any problems with fuel issues

[Bikesmith01]

Hey Everyone!

I have an update for you. I did the fuel pressure tests. Here's what I found:

First, I found the ground wire to the heater on my control pressure regulator was broken. I don't think it was a huge deal. Just means control pressure didn't come up to maximum until the engine had heated up the regulator. It's fixed now regardless.

Fuel Tests!

Fuel Delivery volume: The pump is supposed to put out 750ml per 30 seconds and turned out to do EXACTLY that. So far, so good.

System fuel pressure: Supposed to be 4.5-5.2 bar and I found it to be a bit low, about 4.1 bar. I don't know if that's low enough to be problematic. Thoughts?

Cold Control pressure: Supposed to be 1-1.5 bar (tested at just about exactly 10 degrees C). Mine was 1.3 bar. Good!

Warm Control Pressure (after fixing the wire to the regulator): Supposed to be 2.9 to 3.1 bar at idle, 3.4 to 3.9 under partial load. I tested with the car off, just fuel pump running and found it stabilized at 3.1 bar.

Residual Pressure: Supposed to hold at least 1 bar for twenty minutes. With the valve on the test gauge open to the control pressure regulator, the pressure dropped to zero in 6-7 minutes. First big red flag! With the valve on the test setup closed to eliminate the control pressure regulator, the pressure dropped quickly to about 1.5 bar and then dropped very slowly. After twenty minutes it was just a touch under 1 bar. Call it 0.95. (I wonder how pressure-tight the valve on the test setup is. Maybe lost pressure slowly through there into the control pressure regulator?)

Fuel pump pressure/residual pressure: I put the pressure tester inline both between the fuel pump and the fuel accumulator, and between the FA and the fuel filter. On this test I'm not as sure about what I was looking for. The test procedure says to run the fuel pump until pressure reaches 5 bar (78 psi) and then make sure residual pressure doesn't fall below specification. First off, 5 bar is not 78 psi. It's more like 72 psi. Also, my specification book doesn't mention residual pressure specs specifically for the fuel pump. So I'm not sure exactly what I was looking for, but here's what I found: Fuel pressure maxed out at 4.8 bar, about 69 psi. When I had the tester between the fuel accumulator and the filter, the pressure dropped very quickly to about 1.5 bar, then slowly from there. When I had the tester between the pump and the FA, the pressure again dropped very quickly to 1.5 bar and then slowly but not as slowly as the previous test. This time pressure was down to zero in less than five minutes.

So where does that leave me? It seems to me that the tests indicate that my fuel pump is questionable. Maybe low pressure. Maybe that low pressure is causing the system pressure to be a little low as well? The fuel pump check valve seems to be pretty much dead as well. My fuel pump has the internal check valve, so it looks like I'm replacing it. BoxsterGT said the Bosch pump is a good upgrade and it's a lot cheaper, so that seems like a no-brainer.

Then there seems to be a pressure leak through the control pressure regulator as well. Do you guys think that's worth replacing as well, or will getting a pump with a good check valve eliminate that issue?

Thanks once again for all your help!
-Jon

[fanaudical]

Which WUR do you have installed? Is it the -009?

[boyt911sc]

Jon,

Your WUR is leaking and causing the residual pressure to bleed too fast. This problem is common to CIS pre-SC (1978). If I put your WUR in my test jig, you could tell in less than 10 sec. if the WUR would pass the leak test. You need a good working WUR to fix your cold start problem. To make matter worse, your primary pressure valve inside the FD is leaking too.

Tony

[PeteKz]

I wasn't sure how you tested the residual pressure between the FP and FA, and then between the FA and FF. Did you use the CIS test valve to isolate the fuel pump from the rest of system pressure? And then to isolate the FA and FP from the rest of the system? It sounds like you have the fittings to screw into the FA and FF, correct?

Another test you can try: pressurize the system and then shut off the pump and quickly remove the return line from the top of the WUR. You should not get fuel out the port during the residual pressure test. If you do, the WUR definitely is leaking. Your other tests indicate this. Fix the WUR before you replace other parts.

I still think you should remove your FA and make sure it's not leaking internally. Even though it has no return line like later CIS systems, if the internal diaphragm is damaged, the FA won't provide the reservoir of pressure to hold residual pressure in the system. At this age, and with ethanol contaminated fuel, the FA's seem to be failing frequently. Do you know if it has been replaced?

[Bikesmith01]

Okay, seems like I'm closing in on the issue!

Fanaudical, you're right! I forgot to mention which WUR I have! I don't know the part numbers, but I have the one with a single vacuum port on top for load enrichment. No altitude compensation.

Boyt911sc, Yup, I agree the WUR seems highly problematic. The system pressure regulator looked pretty close to correct, though. After 20 minutes it was still holding just about the pressure it was supposed to. Do you feel that's bad enough to warrant replacing?

PeteKz, So in reviewing my process for the fuel pump residual pressure test this morning I realized I messed up in a blindingly obvious way. I put the tester right at the output of the FA, one hose of the tester on the FA fitting and one on the hose that would have connected to it. Basically just used the tester to increase the length of the hose between the FA and the filter. Then I did the same thing on the input fitting of the FA.

Where I really messed up, though, was that I left the valve on the tester open during the residual pressure test. So obviously (in retrospect) the gauge was connected to the whole fuel system and the pressure was bleeding off through the WUR.

What I think I should have done was hook the tester up in the same way (making sure the side with the valve is toward the FD). Then I can pressure everything up, close the valve, shut off the fuel pump, and see if the isolated pump can hold pressure. I'll try that today and report back.

I'm also going to try the other test of the WUR you describe, but I'm a little confused when you say "during the residual pressure test." If I have the car in normal running configuration, run the fuel pump for a minute, and then disconnect the WUR return line, is that doing the test correctly? I shouldn't see any gas coming out of the return line fitting of the WUR? That makes sense to me, but I want to make sure I do it correctly.

As for the FA, the Probst fuel injection book says you can check for a leaking FA by removing the screw on the bottom and seeing if gas comes out. I did that and mine was dry. I don't know if my FA has been replaced. Is there a way I can tell by looking at it?


Overall, It sounds like my WUR is almost certainly my biggest issue. I'm going to redo my fuel pump check valve test today, and try PeteKz's idea of seeing if I have fuel coming out of the WUR return fitting.

But it looks like replacing the WUR is my next step. Pelican doesn't appear to have the correct part, but a place called Sierra Madre Collection does. Have you guys dealt with them before? Are they reliable? Is there another source you'd recommend?

Thanks again!
-Jon

[fanaudical]

Here are the -009 WUR specs I have collected for the CIS system on my '75 Targa:

System pressure:
Acceptable range when testing: 4.5 to 5.2 bar
When adjusting set within this range: 4.7 to 4.9 bar

Control pressure - "warm" - Vacuum controlled version:
Testing at atmospheric pressure (no vacuum): 2.7 to 3.1 bar
With vacuum: 3.4 to 3.8 bar
Apply 520 to 546 mbar (390 to 410 mm Hg) to control pressure regulator (WUR) to perform this test.

Leak test (engine warm):
Min. pressure after 10 min: 1.3 bar
after 20 min: 1.1 bar

Fuel injectors opening pressure: 2.5 to 3.6 bar.


This thread has a pic of the pressure/temperature curve for the -009 WUR for cold control pressure with no vacuum applied:

http://forums.pelicanparts.com/porsche-911-technical-forum/780952-sorry-yet-another-cis-troubleshooting-thread.html


My opinion - You need to fix all the leak-by issues first.

Other than that: Your specs and my specs differ somewhat for your pressure tests. It's not necessary to run the engine to test warm control pressures (you do have to jumper some stuff to get power to the WUR heater). Not sure what you mean by "partial load" for your warm control pressure test - the prescribed test is with either vacuum applied or not. Idling has vacuum applied to the WUR. So I do suspect your WUR could use calibration as well as repair for a potential leak.

I have not used Sierra Madre; I had my WUR rebuilt and calibrated by Tony (boyt911sc - Tony has forgotten more about CIS than I will ever know...) and it came back perfect. You might ask him if he still offers this service.

[Bikesmith01]

Fanautical, here's the relevant page from my technical specifications book. I don't know how important the distinctions are between engine-on versus engine-off-fuel-pump-running control pressures. But it sounds like it's moot if my WUP is having issues

[Bikesmith01]

All right, I have a couple more updates:

I did the leak from the WUR test. With the car in normal running configuration I ran the fuel pump for about 30 seconds. Then I disconnected the tank return line from the WUR. There was a bit of dribbling from the hose, but then nothing more came out of the hole in the WUR. Not what I expected. The residual pressure test seemed to indicate pretty definitively that it's not holding pressure, so where is that pressure going?

Also, I redid the pressure test of the FA and fuel pump check valve. I hooked the tester up between the FA and the filter again, but as soon as I turned off the pump I closed the valve on the tester to isolate the FA and pump. The pressure had dropped to ~1.5 bar before I could flip the valve, but then the pressure actually came back up to between 1.8-1.9 bar. And it stayed there. It's been over half an hour and it's still at that pressure. So that makes it look like FA and pump are good. hmmm.

This is totally looking like mostly a WUR issue, and maybe an iffy FD as well. I had assumed replacement would be the only way to deal with that, but having them rebuilt is an interesting option. Fanautical's endorsement of Tony (boy911sc) has me intrigued. Tony, is that something you'd be willing to do? I'll send you a private message about it.

Thanks! -Jon