Kroon Harness for 944 Turbo

[Tom]

I may have found the issue folks, but I'm not going to say it's 100% until I do a little more digging. And be sure to cut me some slack on my smoke testing. I could have done a better job of sealing off some of the lines, but it was getting late and I was getting tired. It's been days since I finished and my garage STILL STINKS.

Your bench test of the AFM looked fine to me in terms of voltages, but the dropped out areas may indeed be worn carbon traces on the pot tract. Those would cause hiccups when driving, but not make the idle consistently lean or the vacuum consistently low. The AFM's electrical signals do not directly control vacuum. The AFM did not look like the root cause of your issues, based on those tests anyway. Now if it also happens to have its own vacuum leak, that would be different.

The ChatGPT test, as described in the video, doesn't make sense. The barn door creates a voltage that the DME reads on Pin 7. If you open the door, that voltage is going to go up, and the DME will think there is more air, and the car will get richer. The more you open the barn door, the richer it will get, until it ultimate floods and stops running. To the extend chatgpt said otherwise, it was an AI swing and a miss. (Try CarBot here. It's powered by OpenAI but supplemented with 944-specific information.)

You discount the possibility of a vacuum leak because your idle is not high (and presumably the smoke tests), but the odds still point to that. My guess is your base idle is just set correctly and/or on the low side, and your ISV is able to keep the RPM down in spec. If you reset the base idle higher, you might get a fast idle once beyond the ISV's ability to correct.

That said, have you done a compression or leak down test? Vacuum needs to first be generated by a healthy engine, then preserved by the vacuum system. Smoke tests only test for the latter.... Also, I wouldn't pay much attention to the vacuum level until the car reaches full operating temp -- i.e., when the oil pressure is well below 4 at idle. Most cars will pick up 2 or 3 points from the time the coolant first reaches operating temp and later when the engine is truly warmed up and all the mechanical parts have fully expanded. The oil pressure is a much better proxy for that then the coolant temp gauge...

[chrischrischris]

Have you driven the car? Driving it may allow you to observe other symptoms. Like Tom said, you could have an issue with the engine and that's why a lack of vacuum. Any chance there's a bent valve or broken valve spring?

[NCGermerican]

I may have found the issue folks, but I'm not going to say it's 100% until I do a little more digging. And be sure to cut me some slack on my smoke testing. I could have done a better job of sealing off some of the lines, but it was getting late and I was getting tired. It's been days since I finished and my garage STILL STINKS.

Your bench test of the AFM looked fine to me in terms of voltages, but the dropped out areas may indeed be worn carbon traces on the pot tract. Those would cause hiccups when driving, but not make the idle consistently lean or the vacuum consistently low. The AFM's electrical signals do not directly control vacuum. The AFM did not look like the root cause of your issues, based on those tests anyway. Now if it also happens to have its own vacuum leak, that would be different.

The ChatGPT test, as described in the video, doesn't make sense. The barn door creates a voltage that the DME reads on Pin 7. If you open the door, that voltage is going to go up, and the DME will think there is more air, and the car will get richer. The more you open the barn door, the richer it will get, until it ultimate floods and stops running. To the extend chatgpt said otherwise, it was an AI swing and a miss. (Try CarBot here. It's powered by OpenAI but supplemented with 944-specific information.)
I'm not sure how I've never heard of CarBot, but I will definitely be using that moving forward!

In terms of the CGPT testing, my main point of focus was at open loop idle. According to Chat GPT, with the barn door totally closed, the reading should be .2 to .3v, which it was. However, when I opened the door just a tad to simulate idle air flow, the reading SHOULD have been (according to CGPT) between .5 - .9v, which it was not (26:45 in the video). It was only .256v, so it only BARELY increased the voltage. It IS possible that maybe I didn't have the door open enough to simulate where it would be at idle. I went off of what CGPT said "Slightly open (Idle Airflow)". Unfortunately, "Slightly Open" is a little vague and it's possible that the barn door is normally much more open at idle.

One thing I did NOT show in the video. When I push the barn door from fully closed to fully open, I can hear a significant "spring" noise. Sort of like when you open one of those attic access doors that are spring loaded. Hard to describe other than it sounds like a spring binding and letting loose suddenly..... In my next video, I'll try to replicate it and put the microphone right next to the potentiometer lid. I'm not sure if that is a common thing or not.

If the CGPT test was correct, then the symptoms would make sense. If the barn door is open further than the DME thinks it is (the voltage isn't as high as it SHOULD be), that means more air is entering the system at idle than the DME thinks there is. Therefore, it would act like a vacuum leak - the DME is only adding enough fuel for the voltage it's seeing from the AFM. My theory of why this is more prevalent at warm start is that there isn't a cold start enrichment from the open loop fuel map.

As for those brief OL moments - it would explain why I get random hesitations/misses when just low throttle cruising around 45 mph. Those started happening around the same time I started noticing the warm start lean issue.

All that being said, all the Storch MAF does is translate a MAF reading to an AFM signal, so in theory it should be an easy test to see if the AFM can be the true cause of the issue - it only takes a few minutes to install. If I'm still getting a lean condition, then it would probably point to something mechanical.

You discount the possibility of a vacuum leak because your idle is not high (and presumably the smoke tests), but the odds still point to that. My guess is your base idle is just set correctly and/or on the low side, and your ISV is able to keep the RPM down in spec. If you reset the base idle higher, you might get a fast idle once beyond the ISV's ability to correct.

RE: ISV - when I use the ODB+ software and check the duty cycle of the ICV, it was showing between 85-90% OPEN at warm idle. Shouldn't it be much less than that when warm? With it being that much open, it doesn't seem like it's trying to stabilize the idle downward.

RE: Vacuum Leak - I definitely discounted a vacuum leak, but still haven't ruled it out. The smoke tests strengthened my thinking, mainly because I'm not sure of where else to test for a vacuum leak.

That said, have you done a compression or leak down test? Vacuum needs to first be generated by a healthy engine, then preserved by the vacuum system. Smoke tests only test for the latter.... Also, I wouldn't pay much attention to the vacuum level until the car reaches full operating temp -- i.e., when the oil pressure is well below 4 at idle. Most cars will pick up 2 or 3 points from the time the coolant first reaches operating temp and later when the engine is truly warmed up and all the mechanical parts have fully expanded. The oil pressure is a much better proxy for that then the coolant temp gauge...
RE: vacuum - I have been reading this as well - my "bone cold" start reading at 10inHg should be somewhat discounted as a better representation would be when the oil is warm.

RE: Compression test - I did one a few months ago (off camera). I haven't posted the results because I did it with the engine cold. My plan for the next video is to install the Storch MAF and see if anything improves. If not, at least the motor will be together enough for me to start it and get it somewhat warm. Then, I'll start with a compression test on camera and see what that shows. If I have a cylinder that's lower than the others, I'll do a leak down.

[NCGermerican]

Have you driven the car? Driving it may allow you to observe other symptoms. Like Tom said, you could have an issue with the engine and that's why a lack of vacuum. Any chance there's a bent valve or broken valve spring?

I have not yet, but once I install the Storch MAF kit, I plan on taking it out on the road to see what's happening. As for bent valve or spring, I had the head rebuilt about 2 years ago and all was fine - at that time all new guides, seals, retainers and springs were installed.

Plus the motor is running lean, but extremely smooth. I would think if there's a bad valve, I would have a lot of misfires?

[Tom]

I don't think there's any practical way to know in your scenario how much to open the door enough, but not too much, to simulate idle... just my two cents, but I don't think there is any meaningful diagnostic value in that test. If it starts and ends at the right voltage and sweeps smoothly, that's the best available test without fancy test equipment. It's still possible (unusual but possible) that the door isn't swinging open as much as it should for a given air flow level and therefore giving a low voltage, especially if making weird noises, but...

But!! Low voltages from the AFM are't going to drop your vacuum as much as you are seeing unless it is so lean the motor starts to miss -- which does not seem to be the case (right?). The AFM is just reporting how hard air is being pulled through it, it does not control how hard air is being pulled through it. Focusing on that signal is akin to testing the thermostat in your house because the house has drafty windows and is too cold. Sure, it's possible the thermostat is inaccurate, but it's not making the house cold -- if that makes sense?


On the other hand, yes, those OL glitches in the AFM sweep are a likely cause of hiccups and hesitations while driving. Installing the MAF should smooth that out, if that system is well designed, but is unlikely to fix the low vacuum issue unless your AFM has a vacuum leak in it after the barn door. My concern for you is that you'd be introducing more variables with the MAF -- making it harder rather than easier to get to the root cause. There are often kinks to work out when making changes like that. You might get lucky and all the issues go away with the MAF, but I'd make the install easily reversible, so that if your problems remain (or morph into others) you can go back to the AFM to continue your hunt for the root cause.


One old school trick for finding vacuum leaks is to spray starter fluid at various connections/location under the hood. If the idle changes when you spray something, it means the starter fluid is getting sucked in at that point.

[NCGermerican]

I don't think there's any practical way to know in your scenario how much to open the door enough, but not too much, to simulate idle... just my two cents, but I don't think there is any meaningful diagnostic value in that test. If it starts and ends at the right voltage and sweeps smoothly, that's the best available test without fancy test equipment. It's still possible (unusual but possible) that the door isn't swinging open as much as it should for a given air flow level and therefore giving a low voltage, especially if making weird noises, but...

But!! Low voltages from the AFM are't going to drop your vacuum as much as you are seeing unless it is so lean the motor starts to miss -- which does not seem to be the case (right?). The AFM is just reporting how hard air is being pulled through it, it does not control how hard air is being pulled through it. Focusing on that signal is akin to testing the thermostat in your house because the house has drafty windows and is too cold. Sure, it's possible the thermostat is inaccurate, but it's not making the house cold -- if that makes sense?


On the other hand, yes, those OL glitches in the AFM sweep are a likely cause of hiccups and hesitations while driving. Installing the MAF should smooth that out, if that system is well designed, but is unlikely to fix the low vacuum issue unless your AFM has a vacuum leak in it after the barn door. My concern for you is that you'd be introducing more variables with the MAF -- making it harder rather than easier to get to the root cause. There are often kinks to work out when making changes like that. You might get lucky and all the issues go away with the MAF, but I'd make the install easily reversible, so that if your problems remain (or morph into others) you can go back to the AFM to continue your hunt for the root cause.

One old school trick for finding vacuum leaks is to spray starter fluid at various connections/location under the hood. If the idle changes when you spray something, it means the starter fluid is getting sucked in at that point.

My methodology was to separate my (2) main issues:
1) Low engine vacuum
2) Very lean on warm start (open loop)

I may be silly for separating them since they very well could be related. The reason I separated them is because I've had a low vacuum issue for YEARS. In fact, John Vitesse noticed it when I sent him a log file when I had his MAF setup a few years ago (that I couldn't get to work correctly). I uninstalled his MAF kit and went back to the factory AFM setup and all was well. Then came the motor replacement last fall and everything ran great for a few months. Then the weird stuff started happening - first I noticed the lean condition on warm start, then I started noticing the "misses" when just cruising at low throttle input....and then the big one hit. I was just driving to lunch and the car lost all power. I would give it gas and it just sputtered and sputtered......but the AFR was rock solid at 14.7! That was what sent me down the rabbit hole - new AOS, new Kroon Harness and now new MAF kit.

The problem is, I haven't driven the car to see if there's still any issue while driving, but I REALLY want to get this figured out before getting it on the road.

I think you are correct in that this HAS to be a vacuum issue. Going back to basics, my diagnostic vacuum gauge shows the following for vacuum readings between 10-15. 1) Vacuum leak in the manifold system somewhere 2) Retarded ignition timing. I have double checked my timing and the marks are spot on. Plus, would the DME retard the timing if it was detecting a lean condition from a vacuum leak, compounding the problem? I really need to take a step back and rethink all of this. I feel like I'm starting at ground zero all over again.

[Tom]

I don't think there's any practical way to know in your scenario how much to open the door enough, but not too much, to simulate idle... just my two cents, but I don't think there is any meaningful diagnostic value in that test. If it starts and ends at the right voltage and sweeps smoothly, that's the best available test without fancy test equipment. It's still possible (unusual but possible) that the door isn't swinging open as much as it should for a given air flow level and therefore giving a low voltage, especially if making weird noises, but...

But!! Low voltages from the AFM are't going to drop your vacuum as much as you are seeing unless it is so lean the motor starts to miss -- which does not seem to be the case (right?). The AFM is just reporting how hard air is being pulled through it, it does not control how hard air is being pulled through it. Focusing on that signal is akin to testing the thermostat in your house because the house has drafty windows and is too cold. Sure, it's possible the thermostat is inaccurate, but it's not making the house cold -- if that makes sense?


On the other hand, yes, those OL glitches in the AFM sweep are a likely cause of hiccups and hesitations while driving. Installing the MAF should smooth that out, if that system is well designed, but is unlikely to fix the low vacuum issue unless your AFM has a vacuum leak in it after the barn door. My concern for you is that you'd be introducing more variables with the MAF -- making it harder rather than easier to get to the root cause. There are often kinks to work out when making changes like that. You might get lucky and all the issues go away with the MAF, but I'd make the install easily reversible, so that if your problems remain (or morph into others) you can go back to the AFM to continue your hunt for the root cause.

One old school trick for finding vacuum leaks is to spray starter fluid at various connections/location under the hood. If the idle changes when you spray something, it means the starter fluid is getting sucked in at that point.

My methodology was to separate my (2) main issues:
1) Low engine vacuum
2) Very lean on warm start (open loop)

I may be silly for separating them since they very well could be related. The reason I separated them is because I've had a low vacuum issue for YEARS. In fact, John Vitesse noticed it when I sent him a log file when I had his MAF setup a few years ago (that I couldn't get to work correctly). I uninstalled his MAF kit and went back to the factory AFM setup and all was well. Then came the motor replacement last fall and everything ran great for a few months. Then the weird stuff started happening - first I noticed the lean condition on warm start, then I started noticing the "misses" when just cruising at low throttle input....and then the big one hit. I was just driving to lunch and the car lost all power. I would give it gas and it just sputtered and sputtered......but the AFR was rock solid at 14.7! That was what sent me down the rabbit hole - new AOS, new Kroon Harness and now new MAF kit.

The problem is, I haven't driven the car to see if there's still any issue while driving, but I REALLY want to get this figured out before getting it on the road.

I think you are correct in that this HAS to be a vacuum issue. Going back to basics, my diagnostic vacuum gauge shows the following for vacuum readings between 10-15. 1) Vacuum leak in the manifold system somewhere 2) Retarded ignition timing. I have double checked my timing and the marks are spot on. Plus, would the DME retard the timing if it was detecting a lean condition from a vacuum leak, compounding the problem? I really need to take a step back and rethink all of this. I feel like I'm starting at ground zero all over again.

Let the engine run long enough for the oil pressure to drop to 3 or lower at idle, then tell us what you are seeing for 1) RPM, 2) AFR, and 3) vacuum on the diagnostic gauge.

[NCGermerican]

Let the engine run long enough for the oil pressure to drop to 3 or lower at idle, then tell us what you are seeing for 1) RPM, 2) AFR, and 3) vacuum on the diagnostic gauge.

That's the plan - as well as a compression test.

[Dare]

Maybe this has been asked and answered in one of these threads*, but are you running with both a stock O2 sensor and a wide-band? If so, then maybe the lean condition is just an intermittent WB sensor/controller/harness giving you the wrong reading sometimes?


* -- It really is a shame you can't search for "O2" or "WB" in the search function.

[NCGermerican]

Let the engine run long enough for the oil pressure to drop to 3 or lower at idle, then tell us what you are seeing for 1) RPM, 2) AFR, and 3) vacuum on the diagnostic gauge.

Just did the above. Here are the results as well as a compression test:

Compression Test (Engine Warm)
Cyl 1: 155
Cyl 2: 155
Cyl 3: 149
Cyl 4: 150

After letting car run for 20 minutes
Idle RPM: 880 RPM
Vacuum: Steady at 15 in HG, dead steady - no bouncing
AFR: Between 14.7-15.0 (both AEM gauge and reading in ODB+
ICV: 93% Open (according to ODB+)
Ignition Advance: 4.1 (according to ODB+)

Here's another test I did. I removed the oil fill cap and NOTHING happened. No change in RPM, no change in AFR, nothing. According to Chat GPT (I looked for CarBot but couldn't find any such chatbot?) that would mean I'm already getting air introduced into the crankcase somewhere. I'm thinking it could be the dipstick tube as I doubt that seals properly (even though I have the o-ring and the plastic spacer and it's bolted to the IM. Any thoughts on that?

Lastly, if I were to go "old school" on this, the vacuum gauge literally says "slow ignition timing" between 14-17 in HG. I checked my timing marks and they seem "ok". Any thoughts on this?