8000 rpm/300 bhp 3.2 rebuild

[chris68]

Hi Graham,

I downloaded the XDI Manual Version 1.7, 2015 from Electromotive, and had a slow read through it. You've probably seen these many times already (apologies in advance) but here are a few points I noted for double checking ....

Page 9, 8-dip switch settings for 6 cyl dual plug should be;
UP, DOWN, DOWN, DOWN, DOWN, UP, DOWN, DOWN (switches 1 to 8 respectively)

Page 11, XDI should be separated from DFU by at least 6"

Page 12, Trigger wheel diameter should be at least 1/2" larger than pulley to which it is mounted
Note, Trigger wheel must be ferro-magnetic

Page 12, refer to table of max air gap and out-of-round specs for your size of trigger wheel

Page 13, crank sensor mounting bracket must be mounted directly to engine block, away from rotating
steel or magnetic pieces and should be non-ferrous

Page 13/14, Crank sensor must be aligned with the centre of the trailing edge of the 11th tooth after
the two missing teeth when engine is at TDC #1

Page 15/16, refer to Fig.6 and Table 5 wiring harness detail

Page 16, First step in wiring DFU's is to install the ground wire. DFU's are factory supplied with
ground wire pre-installed on a tapped, un-anodized hole. This wire (on every DFU) MUST be connected to
chassis/battery ground.

Page 18, refer to common engine setup for Porsche Dual-Plug 6-cylinder

Page 21, refer to Fig.10 for crank sensor connector pin-out (Red-pin A, Black-pin B, Bare-pin C)

Page 22, refer to Fig.11 for recommended installation of two DFU setup. Are you using relays on each
10A supply feed to DFU's?

Page 34, refer to 6 Cylinder Dual Plug Information Sheet

See you tomorrow with Rob and armed with my old scope to look at that variable reluctance crank sensor waveform.
It's a 60 tooth trigger wheel so we should see a sinewave at a frequency that corresponds to engine rpm.

Cheers,
Chris

[Gary71]

Sounds like you’ve got the crack team on this one Graham!

If it’s a dead ECU then I’m sure you will find out tomorrow.

[DarthMJH]

What a great community DDK is.

[911hillclimber]

Thank you Chris, I'll work my way down the list and give you my responses before tonight.
It is a 60 -2 pulley wheel, supposed to give amazing timing accuracy.
Darth:
Yes it is!
Chris and Rob got together to get me over the finish line (whatever that will be) before volunteering to come down tomorrow pm to see what is what.
Educated fresh eyes and all that should reveal operator error or product malfunction.

Whichever, i hope to be able to move on.

[RobFrost]

Chris and Rob got together to get me over the finish line (whatever that will be) before volunteering to come down tomorrow pm to see what is what.
Educated fresh eyes and all that should reveal operator error or product malfunction.

The finish line in this instance is the sound and smell of exploded petrol coming out of the exhaust.

[andytat]

Fantastic Graham.
In these situations a fresh pair (or two) of eyes is invaluable.
Good luck guys!

[911hillclimber]

John came over last week (ddk'r and 30+ year Porsche friend) and we say nothing.
4 more eyes?
Yes, a very good thing.

SEE COMMENT AT POINT 8!

I have gone through Chris' list.
1
DIL switches done. The XDi unit is not set-up for any engine, so you have to do this.
2
Distance fron XDi to DFU's (coil packs) is 36"
3
Trigger wheel. The wheel is a special supplied in the kit by Clewett. The pulley is alloy, tooth ring is steel, both same diameter.
4
Air gap has been tried and measured @ 35 thou, 10 thou, 48 thou. Max is 50 thou, currently 40. Size given by Clewett who said better large than tight.
5
Sensor must be on trailing edge of the 11th tooth. Again Clewett special parts, but alignment correct to trailing edge of #11
6
DFU grounds run direct to battery ground. Checked again, wire bells through.
7
Sensor pin out. The cable is shield, black, white. Cable supplied by Clewett from 23 pin to this sensor cable is shield, black red.
The connector is shield to shield, black to black, white to red.
We have tried changing this to shield together, black to red, and white to black, made no difference.
8
Relays to supply the DFU via 10 amp fuse? NO, no relays used.
12V to DFU's and ecu come off the toggle switch on the 'dash' of the car, the toggle is knocked OFF by a red cover switch, motorsport style.
Done this way so knocking off the toggle kills the engine ignion AND the fuel pump.

COULD IT BE THAT THE FACET FUEL IS TOO NOISEY? NEVER TRIED THE SYSTEM WITHOUT PUMP, WILL TRY THAT AFTER TYPING THIS!

Note!
pulled supply to pump, still the solid green light persists.

Note!
Is it critical to use relays?

[chris68]

Well done Graham. The relays would just be to take current off the ignition switch, or toggle switch in your case. Shouldn't be an issue provided the switch can handle the current and the voltage drop across the switch (when carrying current) isn't excessive.

I noticed that the XDI manual requires the crank sensor mounting bracket to be non-ferrous. Wonder how much this really matters? Guess it's to do with stray magnetic fields (that would concentrate in ferrous items) affecting sensor operation.

Best,
Chris

[911hillclimber]

I think the Electromotive bumf says the housing should be non ferrous if doing a DIY install.
Th Clewett part is nicely milled alum alloy anodized, just as the crank pulley is.

I made a sensor 'extension lead' this afternoon.

As a test, we could remove the crank pulley with it's toothed wheel and mount it on my small lathe. I've made an adapter this afternoon.
The wheel can then be spun in the lathe and the sensor out of the engine held close (not difficult).
This will give about 200 rpm and the sensor should respond.

The XDi could then be given power to it's yellow lead (disconnect the 2 coil pack power leads) from a battery and the lathe started.
Thus, the XDi would have no noise around it (coils dead), the sensor would be busy (if good) and the solid green light should alternate green/red etc very quickly!

I have 6 plugs out of the engine atm so the engine turns over quickly.
iirc the XDi needs 90 rpm to pick the sensor signal up minimum.

90 is a lot on a lardy 911 engine starting imho.

[chris68]

Sounds good. You should see voltage fluctuations from the sensor using a multimeter, just by moving something ferrous near the sensor face.

[911hillclimber]

It was a pleasure meeting Chris and Rob today and really appreciate the 5 hours we spent going through everything and more.
Big effort, learnt a lot and have good details to send to Clewett and Electromotive for when the XDi box goes back to America.

We did not get it going.
Positives:
The sensor is good, especially when we reduced the air gap to 10/8 thou, the sinusoidal wave was crisp and 4 volts.
The air gap made a significant difference.

The status of the light is unchanged though twice and briefly it did in fact alternate for maybe 2 or 3 seconds, but then reverted back to 'solid green'.
Rob and Chris could not see anything I'd done was the problem.

Photographs were taken of the traces to send to Electromotive

We have got to the conclusion the XDi is suspect.

Negatives?
Have to pay to ship to USA and back
Loose 4 weeks
Cancel events and rolling road.

[Nine One One]

Can you not borrow/ share/hire a car for your first event?

[911hillclimber]

I can go hillcliming no problem, simply use my Boxster (tempting).
To score in the Championship you have to enter a class and in that class you are awarded a point or more, but those points cannot be carried over if you swop to another class.

TBH if I can't race the Lola then I don't fancy risking the Boxster or even the 911.

The first Championship race is at Prescott 23/24 April.
Before the meeting my plan was to get the car to the rolling road on the 17 March and then to the test day at Loton Park on the 2 April, then to Prescott in a car i have some experience of.

The RR and Loton look unlikely.
no the XDi has to go back for a month.

Not the best situation in a car with about 560 bhp/tonne.

[Mike911scrs]

HI 911hillclimber

Give me a ring, (that's a telephone, I know you don't like using them) as I have some parts you can borrow untill the XDI unit comes back from the USA

regards mike

[chris68]

Hi Graham,

Really enjoyed meeting up yesterday despite the disappointment of not making much progress. Thanks to Mrs 911hillclimber too for keeping us fed .

At least we verified clean waveforms from the VR speed sensor all the way to the XDI input connector. We also proved that supply voltage dip wasn't an issue, using a separate battery (grounded to chassis) to power the XDI. The coils were un-powered eliminating the major noise source and cables were well segregated and screened where appropriate. I really can't think what else it could be and the obvious thing to check now has to be the XDI ecu. Reckon there must be a dry joint or faulty input circuit that's preventing it from seeing the speed sensor.

Very kind offer from Mike. We concluded yesterday that we need to try another ecu to make any progress.

Cheers,
Chris