10 mpg 3.2 - can’t find where the fuel is going

[ischmitz]

Close Dave (Loren) but the two comparator outputs from 139 are evaluated by the IO port P1.6 and P1.7 lines of the MCU. The two interrupt inputs of the MCU are used to decode engine speed and TDC signals from the custom flywheel amplifier.

From a hardware perspective there was a very small fraction of early Euro DME boxes that came w/o the components for the O2 sensor. My guess is Bosch quickly decided that making different hardware isn't worth the savings over the additional logistics required to track the different models.

But it's the code on the EPROM that decides whether inputs to P1.6 and P1.7 are used to modify the mixture or simply ignored. My bench-test fixture makes that very obvious when simulating full rich or full-lean O2 signals and then watching the injector duration changing at a given load/RPM/CHT point.

And the time constant of these mixture changes is on the order of 5 - 10 seconds and not momentarily. But hey, I am just a hobbyist and talk about data I collected over the last 20 years I've been doing this.

@Steve, please drop off your 911 and leave the pink slip in the glove box - preferably with a stock chip

Originally Posted by mysocal911

The S100 chip (339 quad comparator) two (of 4) outputs drive the 8051 two interrupts to either cause the fuel mixture to become lean or rich momentarily,
based on the O2 sensor's input (pin 24) voltage to the DME ECM. The way the S100 inputs are biased (resistors) can affect the Lambda, e.g. a value greater than 1.0.
Most likely, it's just the two interrupt subroutines of the EPROM which determine what the average Lambda will be.

Note: Remember some Euro chips run closed-loop too! Understanding digital/analog electronics and repairing a DME ECU isn't as simple as changing map cells
in an EPROM - "performance" tuning, as we've learned from naive posts of some.

[ant7]

Out of interest, this is another unit I acquired for parts a while ago, can anyone recognise this ECU by the chips fitted, my original assumption was/is that its a very early 3.2 unit, but...
Ant.

[ischmitz]

I can’t quite make out the EPROM part number but since it’s missing the components for the O2 sensor signal it’s got to be an early ROW box. There is an unpopulated area near the heat sink for the ignition final driver.

Can you read and post the Bosch number on the EPROM? I seem to see 1 267 355 027 and that is the earliest code I’ve ever seen for a ROW box. And 1 267 355 028 is the corresponding U.S. version.

[mysocal911]

Originally Posted by ant7

Out of interest, this is another unit I acquired for parts a while ago, can anyone recognise this ECU by the chips fitted, my original assumption was/is that its a very early 3.2 unit, but...
Ant.

It's has no O2 sensor monitoring capability;

[mysocal911]

Originally Posted by ischmitz

Close Dave (Loren) but the two comparator outputs from 139 are evaluated by the IO port P1.6 and P1.7 lines of the MCU. The two interrupt inputs of the MCU are used to decode engine speed and TDC signals from the custom flywheel amplifier.

Actually it's to start/stop a timer, from which the RPM is determined, and the number of teeth between each reference signal, thereby needing it only once.
My error in not having the wiring diagram in front of me for the actual pin numbers.

Originally Posted by ischmitz

From a hardware perspective there was a very small fraction of early Euro DME boxes that came w/o the components for the O2 sensor. My guess is Bosch quickly decided that making different hardware isn't worth the savings over the additional logistics required to track the different models.

But it's the code on the EPROM that decides whether inputs to P1.6 and P1.7 are used to modify the mixture or simply ignored. My bench-test fixture makes that very obvious when simulating full rich or full-lean O2 signals and then watching the injector duration changing at a given load/RPM/CHT point.

And the time constant of these mixture changes is on the order of 5 - 10 seconds and not momentarily. But hey, I am just a hobbyist and talk about data I collected over the last 20 years I've been doing this.

@Steve, please drop off your 911 and leave the pink slip in the glove box - preferably with a stock chip

The response time to the O2 sensor signal is about 1/2 what you've indicated, and has a less significant effect than does the AFM, CHT, & RPM on fuel mixture.
Too slow of a response time to the O2 signal will cause idle hunting/oscillation. When normal idling occurs, there's a small 25-50 RPM hunting.

[ant7]

Thanks, the number you quote from the chip I am mostly in agreement with, however; its not clear whether the third from last digit is a 0 or a 4, tried looking at it using various methods.
Ant.

Originally Posted by ischmitz

I can’t quite make out the EPROM part number but since it’s missing the components for the O2 sensor signal it’s got to be an early ROW box. There is an unpopulated area near the heat sink for the ignition final driver.

Can you read and post the Bosch number on the EPROM? I seem to see 1 267 355 027 and that is the earliest code I’ve ever seen for a ROW box. And 1 267 355 028 is the corresponding U.S. version.

[ant7]

I gathered that it had various components missing for possibly the 02 monitoring, hence the reason it was purchased here in the UK, we didn't use' or have much need for such methods of monitoring emissions at the time.
A bit more info on this unit; with it installed, my car starts fine, idles great, and will rev up easily until it reaches around 2,700rpm, then it will hesitate, but if you give it a bit more throttle it will go past that rpm range with a struggle, hence my initial hesitation in accepting it as an early 3.2 ECU.
Ant

Originally Posted by ischmitz

I can’t quite make out the EPROM part number but since it’s missing the components for the O2 sensor signal it’s got to be an early ROW box. There is an unpopulated area near the heat sink for the ignition final driver.

Can you read and post the Bosch number on the EPROM? I seem to see 1 267 355 027 and that is the earliest code I’ve ever seen for a ROW box. And 1 267 355 028 is the corresponding U.S. version.

[pmax]

Originally Posted by ant7

...some very useful info has been submitted here though' by those with extensive experience and knowledge of these old ECU's.
...

I will echo the same.

Learned a lot from what mysocal and ischmitz have posted here in this forum.
And that they participate regularly in this forum is much appreciated.

Anyway, enough of that, back to our OP's saga ...

[ant7]

Originally Posted by pmax

I will echo the same.

Learned a lot from what mysocal and ischmitz have posted here in this forum.
And that they participate regularly in this forum is much appreciated.

Anyway, enough of that, back to our OP's saga ...

[ischmitz]

Originally Posted by ant7

Thanks, the number you quote from the chip I am mostly in agreement with, however; its not clear whether the third from last digit is a 0 or a 4, tried looking at it using various methods.
Ant.

It's got to be a zero. There was no Bosch number with a 4 in that position. I also noticed the two metal-can transistors and the ribbon cable looks like the early paper version. So this is one of the really early boxes. Check the IC for 4-digit date codes. They can be like 3583 meaning it's the 35th week of 1983. I bet you you'll find stuff from 1983.

[mysocal911]

Originally Posted by ischmitz

It's got to be a zero. There was no Bosch number with a 4 in that position. I also noticed the two metal-can transistors and the ribbon cable looks like the early paper version. So this is one of the really early boxes. Check the IC for 4-digit date codes. They can be like 3583 meaning it's the 35th week of 1983. I bet you you'll find stuff from 1983.

A TO-5 type transistor & the logic board material (paper?) not reliable like fiberglass PCBs.

[ant7]

Understood,
Yes, paper ribbon between boards, chip set pick below;
Ant.

[ischmitz]

Looks like the ADC was made in week 15 of 1983 and the MCU in week 26 of 1983. So the box was probably made in the second half of 1983.

[ant7]

Thanks for the info
Ant.

Originally Posted by ischmitz

Looks like the ADC was made in week 15 of 1983 and the MCU in week 26 of 1983. So the box was probably made in the second half of 1983.

[Discseven]

ECU was found to have a number of issues. Systems Consulting corrected them. Unit is on the way back to me from CA.

In the meantime, I've posted Blackstone's analysis for the Motul synthetic oil that's in the case. Virgin oil was also tested to have a base line. It's perhaps mute at this moment to consider fuel in the oil but this analysis was started a while back for the interest of it and to check one more possibility. Seems worth it to share the results, be they as they are.

https://forums.pelicanparts.com/porsche-911-technical-forum/1179640-blackstone-oil-analysis-motul-synthetic-20w50-sloppy.html#post12489969
.

[Showdown]

What were the ECU issues and how did they affect the fueling/timing/etc...?

[LJ851]

Originally Posted by showdown

what were the ecu issues and how did they affect the fueling/timing/etc...?

+1

[Discseven]

Originally Posted by Showdown

What were the ECU issues and how did they affect the fueling/timing/etc...?

Julian... Systems Consulting did what they do, tested, then installed the unit in a 3.2 on hand at their place. I was advised the issues with it were corrected and it tested & ran well. An mpg test was NOT done there. That's my responsibility when I get the unit back. Of course the mpg must be known. Once I've done my runs, given the generous participation instigated by Dave---thank you Dave---and followed through on by SC, as a matter of respect, I owe it to Systems Consulting to discuss whatever is found with them first, then get the status of the ECU as it was found by them and details as to all that was needed to make it right. Post that info when I have it.
.

[GH85Carrera]

So the Blackstone report did show some fuel in the oil. At least there was no antifreeze!

It will be interesting to hear of the new fuel mileage with the repaired DME. Good Luck.

By the way, I recently drove my 911 for 450 miles on one tank of gas. It is possible (maybe) that I might have been running at 90 for the trip. 3,600 RPM, and 19.5 gallons used. 23.07 MPG. At 70 or 75 I can go 500 miles, but who cares. Interstate 70 in Kansas and Colorado is pretty boring and goes on forever.

[ischmitz]

+2

Let's understand what issues were found and what Systems Consulting did to correct them. Also why is Loren gaslighting everyone here posting under his new moniker mysocal911 peddling his business. I find it hard to believe that people vote with their dollars for such nonsense and lack of integrity....

Originally Posted by mysocal911

You need to talk to Systems Consulting (systemsc.com).

Originally Posted by mysocal911

<...>This is the Pelican Parts Forum. Let's post honest info to its members.

Originally Posted by mysocal911

Karl, I spoke to these guys. They'll send you a no-charge 302 U.S. Porsche OEM DME ECM chip; https://www.systemsc.com/

Originally Posted by mysocal911

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

Originally Posted by Discseven

Systems Consulting did what they do, tested, then installed the unit in a 3.2 on hand at their place. I was advised the issues with it were corrected and it tested & ran well.