Wideband data for 3.2 getting 10 mpg city - what do you think?

[Discseven]

Checks just completed...

• All intake nuts. All tight. Replaced barrels with 10mm nuts after rebuild. Find these easier to deal with...

• Rubber sleeve joining manifolds. Loosened and looked all around sleeve. This has bias towards 4-5-6 so a leak here??? ...Was tight. No cracks in rubber. Mating surface on manifolds are without blemish.
  
• ICV rubber. Not a contender for 4-5-6 but right in hand. All hoses are pliable and connections are tight.
  
• Intake manifolds. Cleaned & inspected during rebuild. I know they are not cracked and flanges to heads are all flat, aligned, and without blemish.
  
• Ignition wires. All oriented on dizzy correctly and run to where they belong.

Cylinder balance... that will happen when ECU's back in service. Results should be interesting.

Arriving tomorrow is a narrow wall 12mm socket. This is to fit into the channels in the SSIs that lead to head nuts. With the engine out and flipped over, it's one thing to access these nuts. Laying below the installed engine it's another. Regular 12mm socket will enter the channels but there's a seam at the other end that prevents regular socket from making complete passage. Fiddling with putting socket on an extension that's passed through the channel... not worth saving a few dollars now. I'll check exhaust nuts when narrow socket's in hand.

Must be closing in on what ails this engine.
.

[Discseven]

Thanks for 964 coil info Sal.

[76FJ55]

One other concept that I don't remember if it was mentioned is if you have a misfire, or dead spark on a cylinder it could cause your lean measurement even though you may have the correct fueling. an O2 sensor only measures oxygen content in the exhaust stream, so if you have uncommuted fuel with the associated unburned oxygen you will have a lean reading on the WBO2.

this may also account for the poor fuel economy, since you will be spending fuel that is not generating the associated amount of work.

[ahh911]

Originally Posted by 76FJ55

One other concept that I don't remember if it was mentioned is if you have a misfire, or dead spark on a cylinder it could cause your lean measurement even though you may have the correct fueling. an O2 sensor only measures oxygen content in the exhaust stream, so if you have uncommuted fuel with the associated unburned oxygen you will have a lean reading on the WBO2.

this may also account for the poor fuel economy, since you will be spending fuel that is not generating the associated amount of work.

Had this, pin hole leak below the spark plug wire dustboot so couldn't see sparks at night, partial spark on one cylinder caused the O2 sensor to feed more fuel to hit 50%.

How are the ground return paths for the left and right side injectors? I read you checked the signal, maybe I missed it.

Phil

[wazzz]

Originally Posted by ahh911

How are the ground return paths for the left and right side injectors? I read you checked the signal, maybe I missed it.

Phil

There is no real ground for injectors. They are permanently powered on one side (+12V) and their other side is the driving signal from ECM, which is a pulsed ground signal with spikes to open and close them.

[Discseven]

Originally Posted by 76FJ55

One other concept that I don't remember if it was mentioned is if you have a misfire, or dead spark on a cylinder it could cause your lean measurement even though you may have the correct fueling. an O2 sensor only measures oxygen content in the exhaust stream, so if you have uncommuted fuel with the associated unburned oxygen you will have a lean reading on the WBO2.

this may also account for the poor fuel economy, since you will be spending fuel that is not generating the associated amount of work.

76... I've not looked at each plug's spark. What's been relied on is plug appearance together with engine's performance. Have had a 5 cylinder 911 under foot. Drivable but relegated to sloth mode. Well said on uncommuted fuel. Is a curious topic given Porsche including a single O2 sensor (for 3.2) and some having disconnected it.

Originally Posted by ahh911

Had this, pin hole leak below the spark plug wire dustboot so couldn't see sparks at night, partial spark on one cylinder caused the O2 sensor to feed more fuel to hit 50%.

How are the ground return paths for the left and right side injectors? I read you checked the signal, maybe I missed it.

Phil

Phil... I tested injector harness for a potential fault that would cut the signal to any injector(s). Ran 12v through the circuit for each injector and harness is good. Ground in this case was the test battery's ground. Gilles has elaborated on injector's working ground.

[Discseven]

Just finished checking vacuum hoses & connections. Roughly 90% of the hoses are relatively new. Ones that are not, they're not brittle. Fuel feed system on engine was refreshed and customized with a spin-on fuel filter in 2020. No problems with that system. Flow & pressure at the rail have been tested. Engine with the new fuel feed did deliver 18/27 mpg pre-rebuild. (27 highway am not historically 100% sure about but know it was in twenties. 18 city I'm positive about.) "Does not exist" is due to 3.2 Motronic engine being with '80 oil tank.



With the maniacal focus given the engine during the rebuild, thought I had every connection done without leaving a doubt behind. Am eating crow here... found a loose clamp ("L") on the brake booster circuit. Were it not for repeated "air leak" comments, I'd not have checked this circuit. Thank you to anyone who mentioned "air-leak." This find is to your credit. Is this loose clamp the holy grail here...?

...This would allow some measure of air to bypass AFM. And I think there'd be a bias of flow into the 1-2-3 side through the booster vacuum passage in that manifold. If so, there's more air entering the engine than the AFM knows of and with the bias 1-2-3 side should run lean. The opposite is the case... 4-5-6 is the lean side. If I'm not mistaken, added air and no O2 sensors (telling the ECU of this condition as is the case here) would lead to improved fuel mileage and not necessarily a good condition for the engine---too lean. Yet the opposite is being proven at the fuel pump. The odo was confirmed being accurate from the get go.

Have yet to check exhaust nuts. Temp gun brake discs to check if any are dragging. And ECU remains in question. With loose clamp now tight, am eager to get ECU reinstalled and trace the engine to see where lambda is with this change in condition.
.

[scarceller]

The leak you found at the Venturi Tube is not likely the case for the issue. As you mentioned it would cause 1-2-3 to read lean and you have the opposite. But nice to fix this anyway.

Please do these 2 next:
1 - Double check ignition cables and firing order, I've seen crazy things in the past.
2 - Swap injectors between banks and recheck AFR, trust me on this one, it's important.

[scarceller]

Originally Posted by wazzz

There is no real ground for injectors. They are permanently powered on one side (+12V) and their other side is the driving signal from ECM, which is a pulsed ground signal with spikes to open and close them.

As mentioned, injectors have 12v at all times on one pin. Then the other pin goes back into the DME via DME Pins 14-15 then internally in the DME those pins go to one single transistor and that transistor applies ground to the injector. The ground is applied using peak-and-hold method. This means it's a ground pulsed signal, can easily be seen using o-scope.

I highly doubt your issue is the DME nor it's wiring. I really suspect air leak, faulty injectors or some other mechanical issue causing the lean out on 4-5-6 side.

[scarceller]

Originally Posted by 76FJ55

One other concept that I don't remember if it was mentioned is if you have a misfire, or dead spark on a cylinder it could cause your lean measurement even though you may have the correct fueling. an O2 sensor only measures oxygen content in the exhaust stream, so if you have uncommuted fuel with the associated unburned oxygen you will have a lean reading on the WBO2.

this may also account for the poor fuel economy, since you will be spending fuel that is not generating the associated amount of work.

Correct, and a cyl balance test by simply unplugging an injector one at a time can easily find the cyl not firing.

[ErrorMargin]

Originally Posted by Discseven

Temp gun brake discs to check if any are dragging…

I’m not sure if you’ve already tried it,
… after a short drive to warm things up a bit, get under the engine with your temp gun and try to get a read on each exhaust pipe where it exits the head, or at least a temperature on each cylinder. This will quickly show if one or more cylinders are not firing.

[wazzz]

Originally Posted by scarceller

One more thought, double check firing order of the dizzy and be sure the spark cables are going to correct cyl. Firing order is 1-6-3-4-2-5 (clockwise rotor rotation). No disrespect here, but I've seen stupid things cause issues in the past

Also the 964 coils are exact same part number as the stock 3.2 and you can find plenty of used good ones on ebay and you get 2 coils for a decent price. They are the good German black ones, just search for "porsche 964 coils" I see a set of 2 right now for $49

Enjoy!

Sal, you may want to edit your post and swap 3 and 2 in the firing order. Someone might read it and think this is the firing order to apply.

[Discseven]

Updated to do:

- Exhaust nuts - Ordered narrow wall 12mm socket and 9/16"is received... redo.
- Cylinder balance / disconnect one injector at a time and compare wideband traces
- Temp gun on down pipes / ErrorMargin... this was done with Lucas injectors. Will do again with Bosch injectors now installed
- Temp gun on rotors / check is for dragging brake related to 10 mpg
- Recheck ignition cables / firing 162435 - Correct connections at distrib and plugs
- Swap injectors side-to-side

Engine is disabled for time being with ECU sent out. Exhaust nuts and recheck of ig cables will do today.

[Discseven]

Ignition parts just tested for resistance…

Clewett spark plug cables end-to-end
multi-meter @ 20k ohms

1 = 0.51
2 = 0.56
3 = 0.60
4 = 1.31
5 = 1.33
6 = 1.64

Blaster coil

Primary @ 200 = 1.4 - mm resistance @ 200 of 0.7 = 0.7
Secondary @ 20k, both = 6.14 (no mm resistance reading @ 20k)
Case grounding to circuits @ 2000k = none

Distrib & rotor

Cap’s coil lead (center pole) outside to inside @ 200k = 0.8
Each circuit in cap (inside rotor contacts to outside leads) @ 200k = all 0.8
Rotor’s center point to outside contacting edge @ 200k = 0.8

Distrib was rebuilt, shimmed, and lubricated aprox 2 years prior to rebuild. Where I am, if this dries out---it's getting close---it rusts quickly.




Ig cables were rerouted in 2020. Did so to see top of fan housing absent that hardware. Emptiness that resulted came with a price—is more complicated to deal with covert cables whenever needing to do so. Having fiddled with ig cables a number of times during this MPG expedition, was time to do something...



ID collars added to all cable-ends. Simple / works nicely. Benefit is mainly 4-5-6 side but is now altogether easier. Rotting fuel lines were dealt with in 2020. Never fond of changing stock fuel filter. New fuel run was designed including spin-on. Not simple / works nicely. Filter head cnc'd by Xometry. Fine job they did. Jim is credited with the spot on CAD files that went to Xometry. Thanks Jim.

ECU inspection news is just in and it's not good. Unit tested in working order / no faults—according to ProgRama / Joseph. Given wideband data to date, I was thinking there was a signal fault to the injectors generated in the ECU. That would have been the problem and onto fixing it. And so my disappoint in "no faults." Sal has held it not being the ECU but I had to check it. (There's a sordid novel behind this ECU.)

Hit list when ECU is back---and after exhaust flanges are checked for tightness:

- Cylinder balance / disconnect one injector at a time and compare wideband traces
- Temp gun on warm engine down pipes. This was done with Lucas injectors. Will do again with Bosch injectors.
- Temp gun on rotors / check is for dragging brake related to 10 mpg
- Swap injectors side-to-side and compare wideband traces

.

[Discseven]

Continued...

Exhaust manifold nuts = All tight


ECU

Fuel injector signal issue did not appear in ECU as I thought was a possibility. Nothing wrong with ECU according to bench test by ProgRama. Am a bit surprised they did not---in the least---attempt to upsell me. Or, find ANY fault.

If you're wanting to bench test your ECU, speak with Sal. I didn't know Sal did this or I’d have gladly sent my ECU his way rather than to ProgRama.

Receipt from ProgRama:



Altitude sensor

With this sensor disconnected or connected… no difference in O2.


Cylinder balance

Disconnecting one injector at a time starting on 1-2-3 side…



Is immediately apparent I f’d up the ID color codes (in LogWorks.) They’re flipped. Am eating crow here. Apologies for this. Doesn’t change solving 10 mpg or the imbalance. Does change the interpretation of when the Bosch-Lucas injector combo was running/tested. Correctly diagnosed, it is the Bosch injector side (1-2-3) that shifted richer and into near alignment with the Lucas side (4-5-6.) This injector combo with idle mixture screw at 5 turns out has resulted in most aligned O2 readings to date.

Were it not for Sal suggesting this balance test, ID error may not have been discovered. Thank you Sal. With IDs now corrected…

Green = 1-2-3
Blue = 4-5-6
Engine at warm idle
Mixture screw currently 7 3/4 turns out






Newly installed 4.9 sensors

Have another pair of Bosch 4.9 sensors on the shelf. Just installed them in place of 4.9 sensors that have been running.

Idle mixture screw was turned fully in and is now 5 turns out. Engine @ op temp…



7 3/4 turns changed to 5 turns (idle mixture screw) makes a difference. Spike = #1 injector disconnected/reconnected.


Rotor temps

Car was roll-tested at the outset of 10 mpg puzzle solving. Was easy hand-push and car came to gentle stop. I subsequently took brakes off the menu. Was suggested to temp-shoot the rotors… good idea. Shot them twice, each at a different stop. Temps in dF.

LF: 96 RF: 99
LR: 111 RR: 221

LF: 100 RF: 117
LR: 128 RR: 136

Delta between LF and RR seems out of place.

Notes: Under hard braking, no pull to either side. Mintex pads, drilled rotors, flushed system and new fluid installed when engine rebuild was in progress. Brake drag brings up…


Hand/parking brake

Very rarely do I use parking brake—handful of times a year if that. Always park car in gear. Just checked: set hand brake and car would not budge. Released hand brake… car would not budge. Backed up slightly and brakes released. Then hand-pushed car… easy and rolled to gentle stop. Call hand-brake a potential 10 mpg contributor, perhaps along with wheel bearings. After posting this, am diving into the rear drums and checking all bearings.


Down pipe temps



There are temps here from a few weeks ago (when the Lucas injectors and W6DPO plugs were installed.) A second set of temps were taken yesterday (with the Bosch and Iridium plugs in.) All temp taking sessions are at random moments. Gun shots are always aimed just above the heat exchanger collar and on the naked down pipe near the flanges. Aim is at apex of pipe’s curve relative to where the gun is. Moving aim even slightly left or right of apex alters readings—so none of these temps should be considered precise. “Ball park” is more appropriate.

Below are the numbers charted above.

Temps taken in 7.24 - Lucas injectors and W6DPO spark plugs:

3 520 6 481
2 521 5 515
1 524 4 457

3 497 6 484
2 445 5 472
1 423 4 442

Temps taken now 11.24 - Bosch injectors, Iridium spark plugs:

3 305 6 377
2 290 5 358
1 283 4 316

3 515 6 521
2 485 5 541
1 501 4 510

Fuel

Have run Shell exclusively for years. Is not a brand preference, rather it is the station’s location that works best for me. 98% of time I’ve filled from this same location.

Now, tank has been run empty and filled at Westar, 22nd & Coral Way. City mpg test in progress. Engine config running with this new fuel brand:

• Iridium spark plugs - took place of W6DPO
• All Bosch injectors - took place of Lucas
• Clamp at brake booster Y tightened - found loose
• ECU handled by ProgRama - benched / no faults found
• Idle mixture screw now 5 turns out - was most recently 7 ¾ turns out - did not keep track of setting for early mpg tests.

Lucas injectors

On their way to Idaho Bill for testing. Given earlier results with Bosch-Lucas combo installed and idle mixture at 5 turns out, when Lucas return, am going to put Lucas back in 4-5-6, leave Bosch in 1-2-3 and run city and highway mpg tests. While this may not be the fix for 10 mpg, it may—oddly—resolve the imbalance.


Still on test menu:

Swapping injectors side-to-side
Parking brake inspection
Wheel bearings

.

[Discseven]

Left side parking brake. Looks guilty. Mostly the hub.

City mpg test in progress is contaminated with this work. After both sides are rebuilt, will restart city run. Get to right side tomorrow. Discovered these drilled rotors have no adjustment hole in hub. Complicates things.

Bearings on front left feel good. Wheel spins freely. No axial play. Did not get both sides of azz end in the air to check rear bearings.

.

[scarceller]

Karl, I just looked at the results from cyl balance test and cyl 4 and 5 are under fueling! That's why the blue line averages out and runs lean.

Swap the injector from bank to bank and test again.

If you look at the blue line results you can see that inj 4 and 5 did NOT lean out mixture as much as the other injectors did. This points to those 2 cyls 4-5 running lean and most likely it's the actual injectors in 4-5.

Swapping them will prove this. You could also just swap 4 with 6 and that way you only mess with one fuel rail.

And injector 4 is the worse of them all.

Hope I'm making sense here.

[scarceller]

The temps on the brake rotors seem to indicate a possible issue, how old are the brake hoses? A collapsed brake hose will cause a rotor to drag because it can't allow the brake fluid back out of the caliper. Keep that in mind.

[Discseven]

Originally Posted by scarceller

Karl, I just looked at the results from cyl balance test and cyl 4 and 5 are under fueling! That's why the blue line averages out and runs lean.

Swap the injector from bank to bank and test again.

If you look at the blue line results you can see that inj 4 and 5 did NOT lean out mixture as much as the other injectors did. This points to those 2 cyls 4-5 running lean and most likely it's the actual injectors in 4-5.

Swapping them will prove this. You could also just swap 4 with 6 and that way you only mess with one fuel rail.

And injector 4 is the worse of them all.

Hope I'm making sense here.

Sal... I see it. Makes sense. Swapping 4 and 6 Friday.

Originally Posted by scarceller

The temps on the brake rotors seem to indicate a possible issue, how old are the brake hoses? A collapsed brake hose will cause a rotor to drag because it can't allow the brake fluid back out of the caliper. Keep that in mind.

Soft lines were replaced a number of years ago. Will take a close look at them tomorrow while also dealing with RR parking brake rebuild. Thank you Sal.
.

[Discseven]

Originally Posted by scarceller

...

And injector 4 is the worse of them all.

...

From my notes, #4 cylinder has the replacement injector I got just prior to reinstalling the original 5 Bosch 158s. Bosch installed in place of Lucas. #4 inj is supposedly also 158. Bosch part number is on the neck like the original 5. What sets #4 apart from the rest is the metal prong at the discharge end---not sure what the correct name of that thing is. That "prong" is thinner that the rest of the inj prongs. Idaho Bill tested all these Bosch injectors including the replacement in #4... and approved them all. To move forward, I went with it.

Originally Posted by scarceller

The temps on the brake rotors seem to indicate a possible issue, how old are the brake hoses? A collapsed brake hose will cause a rotor to drag because it can't allow the brake fluid back out of the caliper. Keep that in mind.

Charted rotor temp numbers. Will be taking RR apart this morning. And looking at all the soft lines.



No soft line receipt found in my records last night. Sometimes parts come from friends---trades---in which case there's no paper trail.
.