Jay, if you sprayed starting fluid around the throttle shaft and the idle went down, then that's a significant air leak that you should fix. Because the spray goes over a wide area, also look at the nearby parts, such as the vacuum nipples and hoses, to be sure the starting spray wasn't finding an air leak there.

Anyone else have various used parts sitting here and there? Ever decide you want to reduce the pack rat mode, and lighten things up - only to later find, damn it, that you got rid of something you really should have saved? It's definitely happened to me. BUT NOT THIS TIME!

I don't know how or where I got it. But I found I have a 73+ throttle body in the garage. It's not perfect. But its throttle shaft and bushings seem very tight. So I will be swapping it in tomorrow to see what happens.

It has a groove around the top rim, and a nut at the end of the shaft rather than a pinch - while my 73 does not have the groove, and does have the pinch. But otherwise it certainly appears identical. Same throat size for sure. PET says that the 73.5 - 75 all have the same throttle body. I would have thought it got bigger going from 2.4 - 2.7. But I'm not complaining.

I want to thank again PeteKz, Boyd911SC, HarryD, ESully, PMax, and everyone else who has continued to help me dig and dig on this. As a long time owner of this 911, as well as a 356 and 912, I know from first hand experience, if something on a Porsche is not smooth, without oddities, clicks, clacks or clunks (except the doors) and plenty of power, there is something wrong. And I just never want to give up on even the minor things (because my car truly continues to run very well, just these nits that I keep focusing on).

Jay. Have you tested your fuel pump? Recently, at a track day, my car would not rev over 5000 rpm. I replaced the fuel filter and the screen in the fuel tank.

The car seemed to rev easier at my next track day but did not feel 100%.

On the advice of my (curmudgeon) mechanic, I replaced the fuel pump. Suddenly, my car revved easily to 6500 rpm as evidenced by the rev limiter kicking in. I also noticed that my cold and warm starts were much better (fewer cranks and less initial sputtering).

https://forums.pelicanparts.com/porsche-911-technical-forum/1159652-fuel-pump-replacements-1973-5-cis.html

May be worth a look.

Funny you should ask. A few years ago, my fuel pump failed as I was driving to a race track! I didn't know what it was. Just flying along the 101 up to Laguna Seca, when the car just quit. I had it towed to a shop. And they finally found the fuel pump failed. They said it was crazy unlikely that a pump fails while in use. But that's what it was.

Well... unfortunately, while the new pop off valve and the replacement throttle both eliminated the whisp of smoke, and spraying starter fluid now has no impact, the car is still basically doing the same things: (1) if I set the CO at a close to factory spec number, the car runs gruff; (2) I have to richen it up to make it run smoother (about a quarter turn), but then idle wants to hunt from being too rich; (3) I'm having a very hard time finding a fuel richness adjustment that gets it to be completely smooth at light throttle; and (4) it doesn't love starting 5 min+ out from stopping - needs me to start engine spinning without any throttle, then floor it after a bit, and it usually starts. Otherwise, it runs great. High rpm, full throttle, idle, etc., all great.

I did just run the injectors in glasses at light throttle to see if they are somehow uneven (because I've never done anything with the fuel distributor). But they are spot on exact.

I did notice that when I did that, when I initially turned on the ignition so I could lift the throttle, 5 of the 6 injectors stayed basically dry. But one of them pushed out a bit of gas. Not a lot. But I wonder if that's enough to make the engine too rich and gives it trouble starting. Probably not.

Hmmm. I just made it richer and it seems all pretty good right now. Smooth and idling fine - but at 1,000 rpm. I'll have to see how it goes for a little bit. Richer than the book says - but maybe my reader is off.

I got up this morning, packed up two WURs to send to Boyd for testing. But my drive to the post office was so glorious, I didn't send them! I tried very hard to see if I sensed anything wrong, and maybe once it wasn't sewing machine smooth - and that could have been a jiggle in the road. No dips in idle either, holding at a hair under 1,000 rpm. We'll see how long this lasts!

I don't know about the starting improving or staying the same yet.

Jay, if enriching the CO/mixture screw above the spec made it run smoother, that is a pretty good indication that you still have an air leak (or several). I can't guarantee that, but it is typical. Hate to say it, but keep looking for air leaks.

Well that begs the question how did Porsche believe that the CO would be measured when setting it at idle? I am using an AEM AFR wideband gauge using a short Innovate tube that clamps on the exhaust pipe - like about 4-5" long, with a narrow pipe. The sensor is screwed into the end of the pipe outside of the exhaust pipe. I understand that it is not accurate at low RPM. I would have assumed that means it reads lean. But maybe it reads rich - just unreliable no matter what. Do I need a longer snake that puts the sensor further inside? Do I need a bung welded on? (Certainly Porsche did not anticipate a bung would be welded on)

The blue CIS book says that it is supposed to be 1.5-2.0 %CO, at 900 rpm idle.

According to online charts I've seen, that translates to 13.95-13.77 AFR.
So question #2 is if that conversion is accurate.

But, now perhaps more to the point. As I understand it 12.7:1 is considered the optimum fuel ratio. The way the 73-75 CIS with throttle position valve works, is it is richer at idle and WOT, and leaner in mid RPM - but at WOT it doesn't actually ever seem to get to as rich as idle, which is odd. So if the car is set to start at a lean 13.95, it goes leaner from there, like to 15 during light throttle, and NEVER comes close to the optimum. I would think that at some point in the range, it would be close to 12.7:1. I know the rising air plate makes it richer as it goes up too. But still, the AFR gauge tracks the same - richest at idle, lean in the middle, richer at higher rpm, but not to 12.7:1.

I also understand that the 1.5-2.0 at 900 rpm is the factory setting because that is what is required for it to meet the then smog requirements. But that the car, even then, ran better more rich. Is that not true? I seem to recall hearing over the years "CIS cars like to run a little rich." If I have heard that, I wonder if that is the same thing - that people knew they had to set the cars richer to run nicely.

I've smoked the hell out of the engine now. There isn't a wiff anywhere. So I can't imagine where else there could be air getting in.

here is a picture of the tester I have. It's about 4-5" long, and the sensor screws into the larger diameter piece next to the knurled tightening knob.

http://forums.pelicanparts.com/uploa...1726531257.jpg

It's a little more complicated than "12.7 is the optimum ratio."
According to Bosch, The best power is Lambda=.90 (about 13.2 AFR). As a practical matter, it's better to set up max power a little richer, down to .85, becuase you lose very little power by going richer, but you gain more resistance to detonation and cooler operating temps. Also according to Bosch, the best fuel mileage is Lambda=1.05 (about 16.0 AFR)so you should shoot for that at cruise throttle and speeds. For idle, the lambda should theoretically be 1.0, but in practice that can cause some roughness or burbling, so a little richer helps settle that.

I'll take your word that you have smoke tested it and there aren't any air leaks.

At this point, I would say that if it runs well, then just drive it.

It may be worth noting that I'm in California, where there is at least 10% ethanol in the gas. I think I'm seeing online that it does mean the settings need to be richer.

I'll try another smoke test. I actually haven't since I found the throttle shaft leak, and fixed it. So perhaps there is an even smaller leak that I didn't see.

Thinking about this some more. I have done quite a few smoke tests. But on hindsight, it seems each time I found another small leak I didn't see before. First it was one of the new rubber boot/hoses that join the airbox to the intake pipes that needed a little more tightening. Then it was the pop off valve that was replaced/better glue. Then it was the throttle shaft, that I changed with a new throttle body. It has never been the airbox - and it has never been a lot of smoke like an airbox I saw some guy video that smoked like crazy from all over the place.

So I wonder if smoke goes to the path of least resistance and each time, I just see the worst offender, and once I fix that one, the next one down starts leaking. If so, we'll see what my next test does!

I should point out that the car idles fine at the leaner numbers. It also seems fine at hard throttle. It's the light constant throttle that is rough. So I need to richen it up for that. When I get to a richness that is good for that, it means the idle is richer than the 1.5-2.0 CO (13.95 - 13.77 AFR, .95-.93 Lamba).

Tail pipe sniffers can give readings off a bit than an O2 sensor. Check out this video at 2:30 min.
https://youtu.be/IDh7gzfotLA?si=RewCnpR2e2kAuF2Y

Thanks Ed. The bummer is that their result shows what I was concerned about: the tailpipe sniffer is saying it's lean, when it isn't. So when they richened it to make up for the false lean reading, they ended up making it too rich.

For me, it is reading richer than the factory says for idle. But as I explained in my last post, my car idles fine at the factory setting. That is not the problem. But at that setting, it runs too lean at the midrange. So I need to have an overly rich idle to get a good mid range.

Perhaps that is a necessary evil of the 73.5 - 75 CIS with Throttle Position Sensor and static WUR.

Jay, given your discussion so far, I'm nearly certain that at some point you will end up buying an AFR meter with the Wideband sensor and bung to weld into the headers. You might as well bite that bullet now and save yourself the time and trouble of continuing to get inaccurate reading off the tailpipe. You may be able to use your current AEM meter and install the sensor on the header. If not, then I advise you to buy a 2-channel meter, that comes with two sensors and bungs--one for each bank of cylinders and headers. I cheeped out the first time and got a 1-channel AEM meter, and now I wish I had spent the extra $120 or so to buy the 2-channel gadget.

Another important point: The exhaust gaskets must seal well to the exhaust ports and the muffler. Otherwise you will get air sucked into the exhaust, which will make the readings inaccurate too.

Well, I'm embarrassed to say I bought the wrong part. The fuel distributor on my 911 is marked 006. Someone had an 006 that was rebuilt by Bosch and never used. So I bought it from him. But then when I got it, I found it was not the same 006. How the heck do they have the same part number on different parts? Mine has three ports - all threaded. The new one I got has only two ports. And I have now seen a third 006 which has three ports like mine, but one of them is blocked off. The new 006 I got has no ports blocked off. Just two molded ports.

I ordered some parts to make it work with my three ports just to test to see if it makes any difference on the remaining hard start. (TPS, gas in, gas out)

I'm also amazed at the absence of something. That leaking throttle shaft must have been playing into my mixture issues for a very long time. I keep going out for drives in my car to experience how nice it is now that I found that and fixed it. Subtle but so much.

I posted this on another thread, and I'm doing it here too in case someone later searches and comes across one or the other looking for this info:

I'm following up on my quest to fix my hot starting problems - started fine for the first 4 minutes after stopping, then had troubles starting for at least a couple hours. I just put in the Bosch 62274, 0 437 502 047 injectors, from Pelican, and it solved it. (Though not really a heat issue, more of an issue of time since last that allows more raw fuel to dribble down into the engine to make it too rich to start until it gets pumped out)

I haven't been able to drive it yet (hip injury from running). So I can't yet comment on how it impacted the rest of the driving.


The symptoms I had were that cold it started OK, but not instantaneous. It also started right away after turning it off, and for about 4-5 minutes after turning it off. But after 5 minutes, it seemed like it had vapor lock. But, after a lot of experimenting, I found that if I tried to start it without any throttle (that is, the cold start valve was not spraying) and just let is spin until it showed sign of ignition, then pulled the throttle, it would regularly start - but still more cranking than I used to have. And if I tried to start with the hand throttle - which is what I normally did, it would never seem to start.

I tested my fuel pump check valve, and it was fine, but I changed it anyway (and tested again).

I tested my pressure leak down. It was totally within spec.

I put all my injectors in beakers to see if they leaked after shut down. Three of them did a little, but I ignored it because (1) the leak down was within spec, and (2) I was told the injectors leak because of the fuel distributor, not the injectors. I got another fuel distributor to test. Didn't change anything.

But because things were pointing to too much fuel on start up, I pulled the trigger and got the $17 each Mercedes injectors from Pelican. They arrived today and I hobbled as fast as I could to install them. I first tested them to see if they were leaking on shut of. Nope. Then started it, and it started right up. Warmed it up, and let it sit an hour or so. Just went to start it, and snapped right on.

SUCH A RELIEF!!!!

FYI, I started some thoughts on if the higher fuel pressure of the Mercedes influences the injectors in the 911 with lower pressure. https://forums.pelicanparts.com/porsche-911-technical-forum/1025109-1973-5-911-fuel-injector-4.html#post12339811

I think you are coming to the conclusion, but chasing smog specs is a fruitless endeavor. The engine wants what it wants for fuel.

My 74 is happy at ~13.4-.6 for idle.

I would not trust that 02 sensor setup, either weld one in or look up the bodges other people have used with that tailpipe adaptor, usual stuffing something in the tailpipe to prevent fresh air reversion.