One more very easy test at the DME relay I recommend:

Remove the relay, measure voltage across pin #85b (-) and #87 (+) while turning key to 'START', do you have 12vdc? If not then the alarm module may be at fault. If you do not have 12vdc you then we need to debug further.

Regarding the power windows, this is a pretty good write up that explains the whole circuit:

911 Power Window Troubleshooting - Dorkiphus.net

Thanks - this reinforces what I think I've learned so far about the power windows circuit. Question:
Can I presume that this reference to "NR 2/60" just below the power window relay in the drawing indicates a wiring harness, i.e., this group of 7 wires wrapped in fabric or a common sheath?

http://forums.pelicanparts.com/uploa...1422567610.jpg

...and if so, does this image from a drawing that Gerry sent me reference a J-Connector fed with wires from that same bundle of 7?

http://forums.pelicanparts.com/uploa...1422567652.jpg

Learning bit by bit, wire by wire...

Thanks,
GK

My advice is pick one issue at a time and fix it, fix the windows or the fuel pump. Don't try to fix 6 things at once, you'll go crazy

After seeing all those wires cut and splices in different places, I recommend getting a beer or 5 and pondering the problem for a little while. I find after the 3rd or 4th beer I reach a moment of clarity.


GK,
The painting took longer than I thought. I'll try to get some pics tomorrow for reference.

I believe you are correct in your interpretation.

Did this test, and it was successful. I first tested fuse 3 (no power) then disconnected the DME relay and tested the hole for pin 30, and found 12 volts. When I jumpered pin-hole 30 to 87b on the DME relay connector, there was power at fuse 3, so we proved continuity from the DME relay connector to fuse 3 - I take that as good news.

I did this test as well - with the DME relay connector still disconnected, I put the black test lead into pin-hole 85b, and the red test lead into pin-hole 87. I turned the key to "start" and though the engine turned, there was only a momentary low voltage, maybe +/- 3 volts, measured across these pins. Since these pins connect to the DME, am I correct to conclude that the DME isn't providing the 12 volts needed to close the relay/power the FP during start? This would reinforce the notion that a previous Mr. Fixit jumpered the fuel pump for a reason.

You haven't steered me wrong yet, so my next step is to clear the duct work and blower out of the way so that I can get a good visual on the alarm system module and the harness leading to it. Apparently there's a "no cut" method of bypassing this module if I get to that point:
http://forums.pelicanparts.com/porsche-911-technical-forum/437580-no-cut-method-bypassing-factory-alarm.html - Any thoughts on this?

Didn't get a chance to chase any red wires down-toward or away-from the fuel pump - I might not get to that until after the weekend.

I hear that! Having said that, I did spend a few minutes re-routing a few of the wires near the fuel filler pipe to take some stress off some connections and make it easier to see a few other connections. This thing still baffles me:

http://forums.pelicanparts.com/uploa...1422587584.jpg

In this connector, the red/blue wires and red wires cross-connect with each other. This isn't just counter-intuitive, it looks exactly like the source of my problem, i.e., the red/blu wire being powered all the time, bypassing the power window relay. I don't see this connector on any drawing - the schematics indicate two separate connectors.

Thanks again to all for your guidance on this - I'm enjoying the hunt.

GK

Since the test for voltage across Relay pins 87 and 85b failed we now must test for the start signal on the DME harness pin #4. Here's the test:

Unplug the large 35pin harness from the DME Control Unit. Locate pin #4 and measure voltage between it and a good ground. You can actually measure across pin 4 and pin 5 since pin 5 is a ground pin.

I'll explain pin 4 since it's not been well covered before: pin 4 comes directly from the ignition switch 'START' signal pin, same signal that goes to the starter solenoid it is used by the DME to determine that the engine has started cranking and once the DME sees 12vdc on it it then grounds Pin 20 that goes to pin 85b of the DME relay. Once pin 85b at the relay grounds the Fuel Pump relay closes and the pump comes on. Once the engine catches the DME knows engine is running and the program code keeps pin 20 at ground. If the engine stalls DME instantly removes ground from pin 20 and the FP stops. It's important this function not be compromised because in an accident the pump will simply run forever and you have a huge safety issue with fuel possibly spraying all over the place even after the engine has stalled! I advice you get this fixed, not a good idea to simply always run the pump with key in 'RUN'.

So test pin 4 at the DME harness and report back.

By the way don't worry about the alarm module, after I reviewed the Alarm module in more detail I determined your fine. If the Alarm module was disconnected, alarm tripped or any wires cut the main DME relay would not come on at all. This is not your case, so for now ignore the alarm module as it's not the likely cause of your issues.

At a point your speed and reference sensors are going to come into this equation because they are what tell the DME the car is running

I'm glad I pulled the driver's seat! Access to these test points is a whole lot easier this way... Thanks for this guidance and the explanation. A primary reason for my pursuing this is the safety issue you describe here (plus, I just like things to work properly.)

To confirm: connect red test lead to pin 4, connect black test lead to pin 5 or other ground, turn key to start, look for 12vdc, correct? If I see 12vdc, it proves that the ignition switch start pin is working correctly and that it's connected properly to pin 4. I'll do this tonight and post results.

For good measure, this pin-out works for my '87 even though the post says it's for an '85 - correct?
http://forums.pelicanparts.com/porsche-911-technical-forum/397830-911-carrera-year-1985-ecu-pin-out-info.html#post3824163


FWIW - the PO had the DME replaced at some point, and I put a new DME relay in when I did my top end project. I have the old DME and relay on hand. For grins, I checked the part numbers on the old and new DME's and they match except for the last three digits - the old one ends in .14, the new one ends in .20; I Googled this and apparently either of them is proper for this car.

That's a relief! I wasn't really looking forward to dismantling the air handling system, so this is like a last-minute reprieve from the Governor!

Thanks again, much appreciated.
GK

- Correct, simply measure across pins 4 & 5 as you outlined.
- That pin out drawing from the 1985 car will work just fine.
- The DME ending in .20 is the better DME, it's a 88-89 28pin chip DME. the .14 is most likely the original one from 87 and may be a 24 pin chip but I'm not 100% sure on the .14

This J- Connector connects all four items together...a power distribution point. Ohm it out and verify that they are all connected. You will find this J-Connector on the schematic. I think it is the same one I show in my earlier photo to you where I explain what a J-Connector is.

Hi Mike -
First, thanks for showing me what they meant by J-connector, that was most helpful.

I had the same thought last night, so I checked this out with my ohmmeter, it is indeed a 4-way connector.

http://forums.pelicanparts.com/uploa...1422587584.jpg

It's hard to see in the pic, but the two wires entering the connector top left and right are red, and the two entering bottom left and right are red/blu. The drawings indicate two separate J-connectors: the red/blu j-connector is supposed to have blue tape, the red-wire j-connector is supposed to have red tape.

The red/blu wire is only supposed to be energized by fuse 1, which gets 12 volts from the PW relay. If red/blu is crossed with the always-on red wire (which it is) then the PW relay is bypassed and the windows will work all the time (which they do.)

I think I'm missing a j-connector and a PO thought it looked right to join these together in this one connector. Per the drawings, it looks like the j-connectors were used to provide a lead to the seat heaters, which I don't have. I may toss the j-connector altogether and connect red/blu to red/blu and red to red via some other means. If the windows and seats all work as designed, then I'm good. If not, I can reverse it and start over.

If I'm missing something (which is entirely possible) please let me know. I see in your j-connector pic that it looks like there are only two wires there, but it also looks like you were in the middle of a wiring project. Does yours j-connector have red and red/blu cross-connected like mine, or do you have two different J/C's like the drawings seem to indicate?
GK

I performed this test today exactly as you suggested. Indeed I saw 10.5 volts between DME connector pin 4 and ground with the key turned to start, also between pin 4 and pin 5 when the key is turned to start. I noted too that my battery reads 12.6 volts at rest, but 10.5 while cranking - is this the usual drop while cranking?

I also tested for voltage on the DME relay connector between pin-hole 86 and ground, and found 12vdc when the ignition is in the on position and 10.5 vdc when in the start position. According to another website, this rules out a problem with the anti-theft since pin 86 gets volts from anti-theft, so there'd be no current if there was an anti-theft issue.

For good measure, I repeated the earlier test where I removed the DME relay connector and looked for current between pin-holes 85b and 87 when key is turned to start. I got the same result as last time - the vm momentarily jumped to 3 or 4 volts.

I got under the car too, it looks like stock wiring going from the fuel pump and alarm horn up into the smugglers box, where the two harnesses disappear into a hole in the tub under the master cylinder. For next steps I'll be checking continuity to the FP from some of the suspicious wires up near the fuse panel, though I'm pretty confident that the FP wire is the one that I found spliced to the wire at the top of fuse 8 since the pump only runs when that wire is connected to the top of fuse 8.

If I've established that the wire on fuse 8 is a direct connection to the FP, would it be a mistake to simply hook it up to the bottom of the correct FP fuse, and see if the FP behaves properly and the car runs?

Thanks again for the continued assistance,
GK

You can move the fuel pump to bottom of fuse 3 and see what happens. You mentioned you have 2 DMEs? If so try both with the pump back on fuse 3.

And one more test:
Remove the relay and measure between 85b (-) and 30 (+) wit the key in 'START' do you have 12vdc (10.5vdc is also OK). Do this test with relay out. If this test passes then the coil pick circuit for the FP relay should be OK. I'd say move the FP back to fuse 3 see if she starts.

Your other points:
Alarm is fine don't bother with it for now.
And 10.5vdc at cranking is also acceptable.

The old DME looks pretty beat. I opened it up over the summer out of curiosity and there's spotty corrosion throughout. Still, I might try the swap at some point.

As far as your test suggestion, I'm going to follow it for sure. I looked through the drawings and want to insure that I understand what the goal is. Let me know if I'm learning something:
In start position, the ignition sends 12v to pin 21 on DME; DME in turn sends 12v to DME relay pin 87 and provides ground via DME pin 20, connected to DME relay pin 85b. This closes that relay, which in turn energizes the FP circuit via pin 87b. Bottom line, we're looking to prove out the functioning of the DME as it pertains to powering the fuel pump. Have I got that right?
Thanks again for the help and the education!
GK

I was going to post some pics of my fuse panel but I looked at it and I don't think it will shed much light on your problem.

It sounds like everything is checking out so why not determine the wire for your fuel pump and connect it back to the proper fuse and see what happens? It would appear that the worst you can do is connect it and go "well. s..t, that didn't work"

Hello Cab,
Yes that is indeed the next "test," put it back the way it was supposed to be and see if it works. If not, resume testing.

My winter project was supposed to be refreshing the rear suspension - if I can clear these two little gremlins I'll be back on track for that.

GK

See, if it would have been me I'd have found the fuel pump wire, connected it to the proper fuse and skipped all the other steps. I'd have started troubleshooting once the wire connected to the proper fuse failed. On the plus side, you've got some good step by step going on here. Since you have your seat out the next logical step once you've figured out if your FP works properly would seem to be opening the power seat connector and seeing which wire is hot and tracing that back up front to the fuse panel.

Good luck!

Happy Ending!
 

Today's progress report:
- Having firmly established that it was spliced to the fuel pump wire, and thereby a direct-connection to the fuel pump, I re-positioned the red wire from the top of fuse 8 to the bottom of fuse 3. Turned key to the ON position: fuel pump did not turn on... so far, so good. Turned key to start, the car lit right up and ran! The fuel pump is now operating when, and only when it's supposed to.

- Having studied numerous iterations of schematics, and having convinced myself that the red and red/blu wires were not supposed to cross in the "baffling" J-connector pictured above, I pulled the red/blue wires out of the J-connector and connected them directly to each other. Windows no longer worked with the car off and key out; Turn key on, windows work; Turn key off, they work until the door is opened. Mission Accomplished!

A battery drain was the initial impetus for this mission, so to insure that I achieved what I set out to do, I put the ammeter in series with the battery negative terminal and the car's ground cable - parasitic draw is down to 40-50 milliamps. (It wavers between the two once every second, so I'm assuming that's the clock ticking.) When the power window relay was continuously energized the draw was closer to 100 mA.

My conclusions are:
- The fuel pump was probably powered by a bypass circuit as the old DME and/or old DME relay was failing. I'll have to test the old relay to see if that theory holds water. If it does, I'll have to discard the old relay which I had kept as an emergency spare;

- There are signs that a wiring harness from another car was at some point grafted into the car. I suspect that the person who did the grafting made an error in joining the red and red/blu wires in the j-connector thereby crossing the circuits and creating the power-window-relay battery-drain.

Now all I've got to do is finalize the new/restored wiring configuration with some better connectors, shrink-wrap and sheathing. I'd also like to tidy up that nest of wiring around the fuel-tank filler neck.

So while the solutions to these problems amounted to little more than restoring the connections to their rightful places, I couldn't have done so without the help of the brother Pelicans who responded to this thread. Thanks to your wisdom and patience I resolved my issues and learned a good deal more about my car in the process. I'm grateful for your assistance. I'll be applying some of this new-found knowledge to solving a few more wiring-related riddles in that frunk...

Best Regards,
GK