"Bubba" joins the stable; sorting out EFI issues

[70SWT]

I recently bought an interesting project car, which I have learned the original owner named "Bubba." It is a Petrol Blue 1980 911SC targa that was purchased new by a physician in the southeast, who soon after that relocated here to the Las Vegas area. I am the third owner.

Here is a pic of when the second owner bought it, dusty from having been stored for a while before that:



Here is how it looks in my hands, when cleaned up a bit; the main difference now it that the second owner reupholstered the targa top:



I was able to locate the OO and talk to him a bit about his Bubba. He says he lovingly owned and maintained the car, the latter of which appears certainly true based also on the stack of service receipts I received. He also apparently had a good time with the car in club racing. He says he racked up a number of podium finishes throughout the southwest during the 80s. Here he is racing at Willow Springs circa 1989, exiting turn 6:



To support his weekend racing fun, the OO equipped the car with a number of functional upgrades, including a through-body sway bar, Bremtek brakes and 930 body modifications.



Perhaps the most interesting mod is this extremely large AC condenser, located under the front passenger compartment. From talking to the OO, I understand he was very serious about keeping it cool in the Vegas heat! It shows:



The car then changed hands in October 2015 to another Vegas resident, a muscle car enthusiast with a restoration shop full of projects. During 2015-16, he removed the CIS system and put together an EFI setup consisting of mechanical items from Tbitz, a custom-fabricated aluminum intake manifold, and a "FAST" brand EFI system. Here is a pic of the system. I have a video of the car running (which is very smoothly once warm – more to come on that), but am not sure how to best post it, as it is too big.



My plan for this car revolves around the fact that, while I am not a club (or any other sort of) real racer, I do want enhanced street performance and generally like a light, nimble car. As the car is rust-free and complete, including the original CIS components (which will NOT be going back on the car, but will be preserved), a lot of body work is not in the cards. So, as this will be primarily a weekend toy for me, my approach to the car will involve:

(1) Optimize the running state without doing any major mods per se for now, although PMO/Jenvey/whatever ITBs or maybe even carbs might be on order later. I respect what was done to create this EFI set-up, but it's not exactly pretty, and I do like the PMO-type look;

(2) Lighten the car without destroying its practicality as a street driver. In particular, I am not a big whale tail fan for either the high-rear weight involved, or the look on a narrow-body car, so that has to go in favor of a ducktail, which is my favorite look. And then,

(3) Address cosmetic items. The car has minor dings and scratches all over, and needs fresh paint. I'll address all of that once 1 & 2 are done.

[70SWT]

So, to the items needing attention. I need to get the EFI tuned, as it starts and runs but is rich. I may need to just take it to a tuner, as it is a highly custom installation combining FAST and a few Bitz Racing components. There is so much I want to do to get the car up to snuff that I need to decide.

Here are some pics of the current EFI system:

[75 911s]

I love this color! Interesting brake and AC upgrades. There are a couple good threads with AFR tables and Bitz kit diagnosing. Since it's such a custom setup, it's going to take some time to understand and tune. Might be too much fuel at cold start if you are getting that fuel smell. What's the AFR at startup? Happy to see another custom air cooled in Vegas. Feel free to reach out, though my tuning knowledge on EFI is extremely limited.

[GaryR]

Originally Posted by 70SWT

I was able to locate the OO and talk to him a bit about his Bubba. He says he lovingly owned and maintained the car, the latter of which appears certainly true based also on the stack of service receipts I received. He also apparently had a good time with the car in PCA club racing. He says he racked up a number of podium finishes throughout the southwest during the 80s. Here he is racing at Willow Springs circa 1989, exiting turn 6:

Great story, but the PO may be confused/exaggerating a bit if he says he was doing PCA Club Racing in the 80's. PCA had Driver's Education events on track then (this is far from racing) and the only amateur "Racing" was with SCCA. The first PCA Club Race was at Summit Point in 1992.. It IS possible he ran Time trials at that time with PCA, that's another possibility.

[70SWT]

Originally Posted by GaryR

Great story, but the PO may be confused/exaggerating a bit if he says he was doing PCA Club Racing in the 80's. PCA had Driver's Education events on track then (this is far from racing) and the only amateur "Racing" was with SCCA. The first PCA race was at Summit Point in 1992..

Good to know...I don't know if he specifically said PCA, so maybe that is on me. I've alerted the OO to this thread, edited the post above, and will make any further adjustments needed to the history.

[70SWT]

Originally Posted by 75 911s

I love this color! Interesting brake and AC upgrades. There are a couple good threads with AFR tables and Bitz kit diagnosing. Since it's such a custom setup, it's going to take some time to understand and tune. Might be too much fuel at cold start if you are getting that fuel smell. What's the AFR at startup? Happy to see another custom air cooled in Vegas. Feel free to reach out, though my tuning knowledge on EFI is extremely limited.

Yes, the custom nature of this is likely to present some challenges. For example, I understand the PO used a creative sensor (in an oil line?) for engine temperature. Still trying to figure out some basic items at this point.

I will post some pics of what the AFR monitor shows when I tackle the project again later today.

Hoping to have this beast at a Cars und Kaffee soon, if I can at least get through plan item (2)...

[jpnovak]

Overall, it looks like a tidy install. You are correct that FAST is not a commonly installed EFI system in the Porsche world. But, like all the others it is perfectly capable of controlling fuel requirements.

The FAST system uses a GM TPI plunger style intake Air Controller (IAC) valve. This should be PWM controlled and require 2 different signals. The IAC should have 4 wires going to it to power coil 1 and coil 2 that operate the position of the plunger. Think of it like Coil 1 pushed down to close the valve; coil 2 pulls up to open the valve and they compete against each other to control position.

The FAST system does not allow for individual target AFR and cold idle speeds. HOwver, I suspect that the valve is not opening to allow for extra air during cold operation. So, When the ECU detects cold start it squirts extra fuel without the extra air and the car runs rich and sputters.

You will need to find the wires for the IAC and then remove the IAC from the car. Verify that the plunger pulls up into the IAC when the car is cold. If it doesn't move then the wiring is wrong or the IAC is bad. Next steps depend on what you find from this test.

Step 1 - remove the IAC from the airbox. It will unscrew. The IAC will unscrew from the filter element.

You can test the sensor with 12V power supply. YOu will have to Ohm out (measure resistance) to first determine which pins go to the 2 separate coils. Then make sure they have resistance and are not open indicating a burned coil. From there, apply 12V to the coil wire and the plunger should move in different directions depending on Coil 1 or Coil 2. If no movement or only movement in one direction then replace the IAC.

To test the wiring, you should have a scope but a DMM (Digital Multi meter) will also work. Matching the wiring position to your coil assignments above you should see some voltage when the car is cold on each side. This is PWM controlled so it will be "less than 12V" depending on how much duty cycle is applied to achieve a desired IAC position.

Can't help much more than this without being at the car.

If you go to a shop make sure they are familiar with the FAST system. Shouldn't matter what kind of car it is on.

[70SWT]

Thanks much, Jamie - will explore as you suggest and report back...

[eastbay]

Don't see it mentioned, but first thing I would do is install a wideband sensor, that way you can see what is going on.

[spuggy]

Originally Posted by 70SWT

I understand the PO used a creative sensor (in an oil line?) for engine temperature.

You should probably find where that is; where it is located will have a big effect on warmup strategy/accuracy.

Oil feed from tank will lag engine temps considerably. Oil return from motor could actually work fairly well - but would tend to mostly reflect internal thermostat opening. The more common sensor in the chain case also lags engine temps considerably.

Methods that lag the "real" engine temps can leave the ECU providing warm-up compensation long after the motor has ceased to want/need it. And crude "delay after start" enrichments aren't great either - consider stopping for gas and a hot restart.

Fitting a CHT is the probably the best way. TurboKraft and Clewett both provide a kit, with a slightly different approach.

In my experience, once the CHT is reading 60C and over, warmup enrichment is unnecessary and should be out - chaincase oil sensor would show only a tiny amount higher than ambient at that point, BTW.

Without good sensor information, it's very hard to tune properly for warmup...

[70SWT]

Originally Posted by spuggy

You should probably find where that is; where it is located will have a big effect on warmup strategy/accuracy.
...

Without good sensor information, it's very hard to tune properly for warmup...

This was my concern as well; I wondered exactly what temperature it was sensing, from where, and where folks have found to be the best location/source on these engines (vs. the 3.2's CHT arrangement...still reading up...)

I have a mountain of receipts I am sifting through to see what components went into this system - at least I have good documentation. I recall seeing a Clewett receipt, will see what that was.

I had to work late last night, and had no time to explore further...hoping to find time today.

[70SWT]

Originally Posted by eastbay

Don't see it mentioned, but first thing I would do is install a wideband sensor, that way you can see what is going on.

I will attempt to see what it has, and what the FAST system will accept.

[tirwin]

Get a copy of Engine Management by Greg Banish to help get started.

[70SWT]

Small progress during the workday: so the PO had just purchased a new IAC and provided it with the boxes of parts and documents that came with the car. It tests out for resistance (about 50 ohms on each of the two circuits, if I read my meter right).

Will also remove the IAC in the system, and test both of them including for operation with 12V tonight...

[70SWT]

Made it into the garage tonight - progress (or lack thereof):

Testing IACs: both IACs test out at around 50 ohms on all circuits. However, it doesn't appear these can be tested for function by applying juice. I understand from web searches that these are pulse activated and won't function for constant current. Neither moves when 12V is applied off of a good car battery.

Checking AFRs: Can't do this yet, as the IAC circuit needs to be sorted out before I hook it back up. I suspect the PO is right and he has it miswired - the new IAC was already in the manifold, so I bet he tested it.

The issue appears to be that while the wiring diagram for the IAC from FAST is color-coded (from his writing, it looks like he tried some trial and error):



...the four-wire GM-type connector has all black wires, and I find no diagram for this plug online or in the paperwork with the car:



So it is very possible this is miswired. I will keep searching to find a pin-out diagram for this connector online. Basically, I need to discover the equivalents between the four pins on this connector and the FAST color-coded wires. The closest thing I found was another guy with essentially the same problem (https://www.supraforums.com/threads/using-gm-style-4-wire-iac-to-wire-into-aem.522797/). I may just call FAST tomorrow.

Once I have that info, I can verify the wiring, fix if needed, and THEN start the car up again with the IAC connected to see what the idling AFR looks like.

[jpnovak]

You have everything you need...

Using the diagram for the IAC stepper.



You will not the following looking down at the IAC connector. There is a tab that notes orientation.

Position A - Coil 1 Low goes to FAST C - Blue
Position B - Coil 1 High goes to FAST D - Red
Position C - Coil 2 Low goes to FAST A - BLK
Position D - Coil 2 High goes to FAST B - YEL

Funny that I don't see this in his combinitorial notes.

From your picture you need to swap the black and blue wires going into the connector.

[70SWT]

Thanks again, Jamie - couldn't find that diagram online myself last night. Will see what I can accomplish tonight.

Other follow-up:

eastbay: It looks like this system uses a wideband O2 sensor.

tirwin: Thanks, will check out that resource.

[spuggy]

Originally Posted by 70SWT

location/source on these engines (vs. the 3.2's CHT arrangement...still reading up...)

Having struggled trying to tune an oil temperature-based warmup, IMO cylinder head temp is the only information to use for controlling the warmup cycle. There is no substitute for reading that directly - and cylinder head is the absolute best location for that sensor

Bottom line, there is nowhere on a dry sumped motor you can read the oil temperature and treat it the same way as head coolant temperature on a liquid-cooled motor; they behave very differently.

Oil feed is useless; the overall oil system takes 10, 15 minutes to heat up. The oil return is not useful because the internal engine thermostat opens after 10, 20 seconds after engine start (very obvious on a car with a 10 bar gauge). So that return temperature stabilises around thermostat opening threshold very soon after starting, and rises slowly over a much longer period of time - eg as the entire system comes up and input oil temps rise too. A chaincase sensor is also not useful; initial lag is far too long to use for warmup.

So, CHT. It's the information you need to control warmup, don't fight it, just fit a sensor that reads it

TK's approach for the CHT uses a plate between the head and the top fin; minor mod to the engine tin, (also need to trim a corner of the plate for twin-plugged). Clewett suggests a drilled/tapped hole (in the cam tower, IIRC). Apparently there's a void, so the shavings can't go anywhere that matter.

I used the TK setup with a liberal smear of Arctic Silver. You can remove/refit the tin with the motor in the car - although one of those jobs where you have to know it can be done otherwise you'd probably give up...

Cylinder head temperatures climb immediately/linearly after start (and reach eye-opening #'s - like over 200C - if you leave the motor hot idling for 5-10 minutes). Once warmed up/moving, 105-110C is typical even under boost - smidge higher in town.


Regardless of where/how the ECU is getting temperature information, if it runs rough and smells too rich, it almost certainly is. In my opinion, get an accurate temperature reference (preferably) and adjust (lean) the warmup compensation until it runs better - making sure that it is not so lean that it objects to throttle tip-in.

IIRC, I use warmup comp fuel trims of 40% @ 20C, and 0% @ 60C, based on CHT. Perhaps still a little overrich on initial start - but you only get to play with this once per day, and it warms up fast... When CHT reads 60C, the block is still cold and so is the oil - but motor responds normally and doesn't seem to need/want any extra rich. YMMV.

Almost all of the warmup comp on an air-cooled motor will happen before the wideband is up to temperature, unless you use "fast heat" mode - which heats the current pump (to 780C) as soon/as long as the ECU has power. This is useful for dyno runs or initial setup, as you can heat up the sensor beforehand and get lambda readings immediately - LSU 4.2s don't seem to care.

But this isn't a good long-term practice; the Bosch data sheet explicitly says not to do this; and LSU 4.9s don't like it much (it can kill them, and will decrease service life if it doesn't).

In "normal" mode, at least some ECUs won't heat the sensor until the motor has been running for at least 10 seconds (and kill the heater as soon as it stops).

The ICV has no real effect on cold running other than idle speed, you don't strictly even need an ICV at all. I don't have one fitted right now (although I plan to), right now I just don't let the motor idle at 800 RPM cold - where it wants to, even with some extra advance.

I just start and drive. Literally 3-4 blocks and warmup comps are all out and idle is a solid 1050 (which seems to be the sweet spot for my car/cams). Still slightly too soon for a wideband reading (which seems to take somewhere around 2-4 minutes), so it's best guess based on other cues...

[70SWT]

Originally Posted by spuggy

...cylinder head is the absolute best location for that sensor...

Great info - thanks for writing all of that up. Figuring out the sensor scheme is definitely on the list of next items to do.

[spuggy]

Originally Posted by 70SWT

Great info - thanks for writing all of that up.

You're welcome; bit of a learning cliff at first eh?

It's a lot of reading/understanding out of the gate. But most all the concepts are transferable to any programmable ECU - and understanding how it's working is priceless, as is the flexibility.

Originally Posted by spuggy

Regardless of where/how the ECU is getting temperature information, if it runs rough and smells too rich, it almost certainly is. In my opinion, get an accurate temperature reference (preferably) and adjust (lean) the warmup compensation until it runs better - making sure that it is not so lean that it objects to throttle tip-in.

...

The ICV has no real effect on cold running other than idle speed, you don't strictly even need an ICV at all.

Just in case it isn't obvious from the above; if your warmup comps are too rich, simply adding air (via the ICV) will just result in a faster idle - exactly the same as if you opened the throttle - and the same overrich fuel delivery, assuming a half-sensible map.

Because the fuel map is based on load or VE or kPA or something (in my case, MAP/BARO) against RPM, with whatever warmup trim it thinks the temp sensor indicates it needs added on top.

Add more air via the ICV and it'll idle faster. And then it'll just be sitting in a different load cell in the map, with (almost certainly) exactly the same warmup comp added. And will still be too rich.