Karl at this point i would call Steve Wong, and send him your ecu unit ..
Ivan

Not really! The chip just has different fuel maps for a different application/country.

Since you've tested your vehicle with a good DME ECM & EPROM, send it here;

Why take chances!

I just have to chime in here. First -congrats to Karl for finally figuring it all out.You probably will have a really good running car as a result of all the checking and testing. I have been following this thread and appreciate all the tips and suggestions that came up. And kudos to Karl for his persistance and thoroughness.
I suggest sending the box to Ingo or Sal, not some storefront outfit. Maybe Steve Wong could suggest a chip for your configuration. And if you want to borrow my ecu back again while yous is at the doctors-just let me know.

Bully and huzzah!

The suggestion was darn near the only thing left to test - you tested and fixed everything else, sometimes more than once!

One of my earlier posts! :)
Ant.

Sent from my iPhone using Tapatalk

Seems a few of us were on the right track early on, sometimes we over think things, that often lead us down rabbit holes we don't need to go down! :)
Glad to hear you finally got a positive result.
Ant.

Sending this wonky ECU to Sal, Steve or Ingo, the right move indeed if having anything like I've faced. However, you probably recall this ECU was not the ECU in my car prior to the rebuild. More to the point, it was not my decision to replace it as was done. Was a Porsche-friend of mine who was responsible to get the engine back in the car and sort out whatever things needed sorting so when I picked up the car from his place, it was running in primo form. In the process of completing things, he didn't like the way the engine was running so had the Lucas injectors cleaned and the ECU "tested"... tested as I knew it early on. What in fact happened, and I learned this later on, was the "testing" place swapped my ECU for the wonky one now in the car. Intentionally? That's a rabbit hole.

My Porsche-friend knows the saga of this fuel situation from the very beginning, so little discussion was needed with him. I only had to tell him of the just-completed mpg test with a substitute ECU. He's going back to where the wonky ECU came from and getting it replaced with a properly working unit. We know "proper" is possible thanks to Dean's ECU proving this engine with:

964 cams, 40/40 grind, 1.45 timing
SSIs
and Dansk 2-in-1-out muffler

...is capable of "reasonable city mpg." Didn't make 18 but came very close. Will admit, I seriously questioned whether it was one or all of these mods that were the cause. Then having to sort out which one or combo was to blame.

Whether my friend and/or the ECU source pulls through as would be respectful has yet to be seen. If not, I'll be dealing with wonkyECU one way or another. I will only call this journey "done" with a properly working ECU installed and acceptable mpg proven. Sooner than later would be quite nice.

While I have yet to resolve this matter, "the puzzle" has been solved---at least on the surface of it. Those of you who added to the journey... Impressive of you. Thank you.

...I agree... interesting air-fuel questions remain.
.

Hey, since the Pelican Parts Forum provided most all the support (2 long threads), shouldn't Pelican Parts be your source?


http://forums.pelicanparts.com/uploa...1750197296.png

Dean... Appreciate the comments. More so the further borrow offer---crazy kind of you! Will have to see how the situation pans out with the shop where this wonky unit came from.



Dave... Indeed, PPF is a great place! And price for Pelican's rebuilt ECU is competitive. As it stands, I believe the resource where this wonky ECU came from should replace it with a proper working unit. If that fails to happen, Pelican's ECU is a possibility among others. Thanks for the heads up on this. And for your input from the very beginning.
.

A lingering wonder is how a 3.2 can efficiently (as proven by spark plug color) process nearly twice the air-fuel for city driving with one ECU (mine) and then get nearly 18 mpg city with another ECU (Dean's.) Look at the city wideband for each of the ECUs (post #173), can you tell there are are two different ECUs here? I can't. Warm Idle and Holding 3,000 rpm do show Dean's ECU running the engine leaner. The Julian-theory is proven---"there is no missing fuel." To support this, if fuel is making it past the O2 sensor with my ECU installed and given the 10-12 mpg city scenario as it exists, a raw fuel odor would be evident at the tail pipe. This does not exist.

Fuel in the oil? While this is significantly mute at this point, it's still interesting. I have yet to receive the oil analysis results from Blackstone Labs. This should arrive this week. If nothing else, that info will possibly say something about my rebuild and perhaps speak for Motul's synthetic crank oil. Post results when they arrive.

There obviously must be some difference in my ECU compared to Dean's. Assuming the chip's genuine in my ECU, what else could it be? With the engine's stellar performance to redline in every gear, there is the possibility this ECU was intentionally modified. Perhaps for racing. Were this mpg matter a glitch in the ECU, am inclined to think the performance would NOT be as crisp & clean as it is throughout the full acceleration range. In other words, am supposing a glitched ECU would NOT perform so well. Were I capable of reading the chip's fuel map, that would happen. Am not.

To Julian's question: Does a stock 3.2's timing play well with 964 cams? I strobe'd my pulley and timing does advance. Post# 168. But I'm not sure if the advancing degrees are correct for the engine/cams. Pulley has no advancing degree marks. I suspect with the Motronic system, advance marks were uncalled for. Or, this 3.2 has an odd pulley. Does anyone have specs for pulley degrees per rpm for a 3.2? If so I'll calculate the advancing degrees from the strobe photos.
.

What seems odd to me is that there's no unburnt fuel with either ECU yet there is a pretty big MPG delta.

So with both ECUs the fuel map and timing map are resulting in full burn but one is consuming a lot more fuel. So that means more fuel injected and perhaps different timing that would still burn all the fuel but not provide a marked power increase...(Karl hasn't mentioned if there was a different feeling with the ECUs but I suspect not) I don't know that much about cams and timing (still learning) so I don't have an answer, just questions.

I would guess the answer lies in these questions:

What's the stock 3.2 cam fuel and timing map
What's the 964 cam fuel and timing map
How and where do they differ (at low RPM and high load; city driving) and would those differences result in a loss of MPG?

Or, maybe the old ECU was misreading the O2 sensor? I know with standalone ECUs you have to calibrate the O2 sensor and if it's off your readings could be off... Not sure if that's possible with a stock ECU...

Lots to read here, very thorough. IIRC, euro ecus like yours don't use O2 sensor input. Am assuming the ecu lent to you does? Is it possible that accounts for the difference you're experiencing? Could it be that the cumulative difference over half a tank doesn't show obvious or significant improvement at the resolution your afr scope uses?

Just a thought.

Someone like Sal could pull the maps from both EEPROM chips and compare — if you REALLY wanted to figure it out.

Since there's no obvious evidence that the extra fuel was going anywhere but the combustion process, it's entirely possible a terrible map — no decel fuel cut; earlier enrichment at partial throttle; longer idle/warm-up enrichment; etc — contributed to bad city MPG without an obvious seat-of-the-pants improvement or impact to AFRs.

964 timing map is apples to oranges since 964 is twin plug vs 3.2 single plug.

Karl, was there a difference in the oil temperature with the replacement ecu ?

The chip has the fuel and timing maps in it so a solution could be as simple as a chip change in the “wonky” ecu.

The thread has reverted to guessing again! Yes, a simple OEM U.S. chip (32K #302 24 pin EPROM, $225) change will solve the problem, a 10 minute effort.

Karl, just catching up to this thread--congrats on solving the problem! Lots of persistence and critical analysis on your part.

Back to speculation: Like others have suggested, the ignition timing in your ECU in the low RPM range may be retarded. Even with correct AFRs, if the timing is retarded, combustion will not provide as much pressure down on the piston, resulting in poor fuel mileage. And, after all the testing and troubleshooting you have done, low RPM timing seems to be about the only variable left that you haven't measured. But, the heat energy has to go somewhere, so that would be into the engine or out the exhaust (or both).

Someone else asked if the oil temp was different between the ECU's, which addresses the first path. The other question is: are the exhaust temperatures different?

At this point, you probably want to just get the correct ECU for your car from the rebuilder/tester, and then drive off into the sunset! However, if the intellectual challenge keeps tickling your brain, consider measuring exhaust temps too.

Congrats again on wrestling this problem to the ground and "pinning" it (bad pun).

Right!

Julian... Good questions on stock 3.2 and 964. Someone must have this data.

I did not WOT with Dean's ECU installed. In retrospect, I should have to make that comparison. Priority was determining mpg. And I promised to send Dean's unit back Monday so there was a sense of urgency working. Absolutely no engine performance distinction between the two ECUs for: start up, cold and/or warm idle, city driving. Wideband logs tell a different story. My "warm idle" and "3000 rpm held steady in 4th gear" both showed the engine running slightly richer than with Dean's ECU. The "City drive" logs I cannot see a difference between the two. Logs in Post #173.

Regarding the ECU misreading the O2 sensor, the wideband sensors installed now are NOT connected to the ECU. At the outset of this journey, the topic of whether or not USA 3.2s require the O2 sensor to be connected was discussed. Word was this connection is NOT necessary. I was reluctant about this and so tested city mpg with the narrowband (at that time) O2 disconnected. City mpg was exactly the same whether this O2 sensor was or was not connected to the ECU. After that, the wideband sensors, Innovate controllers, and LogWorks software to record streaming ARF/Lambda data was installed.


Dale... see the note above concerning the O2 and ECU.



Silver... Would be interesting to know what the maps hold. But... Between parts & tools, and what is yet to come, this flight has not come with an economy seat. Add the time invested and this rabbit hole has run deep. Would be nice to pay less per mile sooner than later and now get there as intelligently as possible. I'm all for finding more answers to what's going on here---they will come---but not up for adding more $s to the kitty to find them.


LJ... Is an interesting note concerning 964 config and oil temps. In the front fender, I have a Mocal with dual fans that ON/OFF with a switch in the dash. I look at dash gauge and when temp reads 210 (or above,) fan is switched ON. Below 210... OFF. Because of this setup & procedure, I have no idea if there was any oil temp distinction between the ECUs.



Dave... Is the chip below the one you refer to that should be replaced:

http://forums.pelicanparts.com/uploa...1750421825.jpg



Pete... Thanks. You know the journey here. Yes, there are interesting unanswered questions. Also yes on "sunset." :)

Exhaust down pipe temps have been measured with my ECU, not with Dean's. I do see now I should have done more tests while Dean's ECU was installed. Depending on the original resource for the wonky ECU providing a replacement that works properly, or my fixing the unit I have, or getting a new ECU and keeping "wonky," there may or may not be the opportunity to comparison-test further. If wonky departs, obviously further exploration comes to an end.

"Heat" being a clear point of interest, I will shoot the down pipe temps again/now to have a current (as well as an earlier) temp profile on file. And coordinate this temp reading with leaving the oil to heat up as it wants over a timed run. When the "corrected ECU"---whatever that may be---is installed, I'll shoot pipe temps again for comparison. Post results for this downstream.

If indeed it is the chip, and I believe Dave knows this stuff inside & out, then it must be an inappropriate fuel/timing map in the chip that resides in the wonky ECU. So, the wonky ECU could be "good" ...for the right engine, not for a 3.2.



Blackstone's "Oil Report" arrived. Looking at that shortly. Post their docs later today.
.