2.1 mv doing my head in

[Aidan]

ok doing my head in too now

you say it runs ok with the grey wire disconnected ?

which grey wire ?

[axeman]

hoew did or do you go about setting the throtle switch? as this is something that is still on my list, from my understanding there should be s slight click at full throtle and when it gos back closed (tick over)

thanks neil

[tencentlife]

Hey all,

I just registered here because it seems that so far no one is aware of a very common problem that you see with lambda on both Digifant and Digijet FI's, the problem being inadvertent grounding of the O2 signal circuit. It has to be isolated from ground all the way from the sensor to the ECU. Inadvertent grounding, however tiny, along the signal circuit will make the ECU go pig-rich. You know it when you see it, this is the only situation I've ever seen that will make the engine run so rich as to produce black smoke. It sounds like that's what is going on here, the telltale being that it began right after doing work on the exhaust and O2 sensor.

To save you having to wade thru a bunch of other irrelevant dreck (and to save me a lot of work), I'm copying here one of the entries I've made on the Samba describing how to check this out, hopefully it will help:

Check it exactly like this:

First, the quick and dirty test (this is alluded to on p.118 of the Protraining manual, it is also in Bentley):

There's a quicky functional test for grounding of the O2 signal circuit with the engine running and warmed up, everything connected as normal including the O2 sensor (this test also works with the o2 sensor disconnected). Under those conditions, being sure the engine has bbeen idling for at least two minutes to be sure it is in closed-loop mode, you take a jumper from ground and connect it to the O2 signal connection (or simply ground the free end of the coax signal wire with the sensor disconnected). Whenever you ground the signal wire, the engine should noticeably alter speed and run stink-rich. If there is no change, then there is already shorting to ground somewhere of the O2 signal. If it does change to a very rich idle, and back to normal when you unground, then functionally all is fine with the signal circuit. This does not confirm anything about the quality of the O2 sensor output, only that the signal is reaching the ECU cleanly.



Now the slow version:

Disconnect the single-wire signal lead of the O2 sensor from the green coax. You will find the connection hanging between the left cylinder head and the left wall of the engine bay.

(If your van has a three-wire heated sensor, you can leave the two-pin heater plug for the sensor connected; it has nothing to do with the signal circuit).

Leave the disconnected end of the green coax hanging free. Make sure it won't touch ground inadvertently.

Ignition off! Disconnect the multi-pin connector from the ECU.

Put VOM in ohms mode, lowest-range if not auto-ranging.

Connect one VOM probe to the free end of the coax center wire.

Touch the other probe to ground. You should only ever see infinite ohms (open circuit). This verifies that the signal wire is not inadvertently shorting to ground. With the probe still attached, move the coax wire around a bit and watch that there is never any continuity to ground.

Leave the probe connected at the signal wire of the coax. Now touch the other probe to the O2 signal pin at the ECU harness plug. You should see zero ohms. This confirms that you have a continuous signal circuit.

Pull the small rubber boot back from the double-male connector that should be on the free end of the coax. Now take the first probe from the signal wire at the end of the coax, and touch it instead to the coax sheathing braid. Touch the other probe to ground. You should see zero ohms. This confirms that the noise-suppressing sheathing is properly grounded.

That is the complete, isolated test of the signal circuit and noise-suppression.


Don't concern yourself with checking the resistance of the O2 sensor to ground (irrelevant), the resistance of the O2 signal circuit thru the ECU (also irrelevant). Confine your tests to the signal coax in isolation.

The area to inspect very closely is the free end of the coax. Many times the ground fault is due to the fine wires of the sheathing braid fraying and coming into contact with the signal core wire or connector. Usually this happens when the connection gets manipulated while replacing the sensor, or any time work is being done in that area. Trim back the sheathing and make sure not a single braid wire can reach any part of the center wire.

If the fault isn't there, sometimes the coax has been mashed or pinched somewhere, causing internal damage where the braid is forced thru the center wire's insulation. You should do a careful visual check of the whole coax run looking for damage.



Don't be tempted, by the way, to ground the free end of the coax's sheathing braid. It is already grounded in the wiring harness, and to cancel noise a sheath like that can only be grounded in one location.



Check that out, sounds like your problem, and if it isn't, best of luck.

[jed the spread]

Lee put an AAZ in and see your electrical woes ebb away

jed

[kevtherev]

Great post Tencentlife..

[jason k]

lee, ghia is flogging a complete mv inj system on brickyard.......

[Simon Baxter]

TDI, TDI, TDI, TDI, TDI, TDI, TDI.... Need I go on?

[SyncroSwede]

I love the character of my WBX and wouldn't change it for a diesel even if it blew up tomorrow!

Thanks to tencent for the input. I have read a good few of his tech help posts on 'the samba' and here's someone who knows his WBXs (and builts sweeet looking 2.2l versions too).

If anyone else has issues then you could do worse than looking over there & searching his posting history.

[..lee..]

tencentlife CHEERS that is indeed a valuable post, many thanks. not had any time to spend on it today and wont till the weekend as i`m needing peace to concentrate on it. faulty wiring seems to be favorite atm. i`ve measured just about everything in the engine bay "THANKS SYNCROSIMON" and all values seem good so my next set of tests will be at the ecu hoping to see a bad signal there confirming a wiring fault. my first port of call will be the signal wire from the oxy sensor, as tencentlife suggests. fingers crossed its that. sounds very likley as both times its played up has been directly after contact with oxy sensor.

aiden the gray wire in the engine bay i belive is to disconnect the auto mapping of ignition and possibly fuel so that the base level timing can be set properly. " well i think thats what its meant to do." there are two yellow connectors next to each other on mine,grey wire and the other green wire one has a 7 volt feed the other a 5 volt feed cant remember which way around it is now but the grey one disconnected results in a perfect idle conditions. it wont rev cleanly in this state which i think is down to the mapping being disconnected i belive. if i connect the grey wire quickly it has normal throttle responce however shortly after the co rises dramaticly and whether i rev the engine or let it continue to idle it goes to run very very badly." not drivable ". disconnect the grey wire-yellow connector and perfect idle conditions return. i suspect something is influencing the ecu when the auto mapping is re connected and tencentlife`s suggestion seems spot on.atm.

any other suggestions from anyone very welcome and thanks to all so far. really annoying thing is that i only want the engine to keep me going for a little while as i allready have a tdi set up on the workshop floor waiting for the loom and stuff to be sorted.

cheers lee.

[Aidan]

Lee
thanks for this, I'm still confused as to the grey wire please can you do a piccie of the plugs and wires concerned.
The green wire single connector is the shared hall sender feed to the idle stab unit, which you disconnect when setting idle speed and co and checking idle stabilizer; the grey wire I don't know which one you mean other than the white grey that doesn't connect to anything connector T1k which I think is for air con or aux alternator connection for load detection for idle stab
regards
Aidan

[tencentlife]

Thanks to everyone for the warm and friendly welcome, and I hope the post helps. I guess you all don't see as many lambda FI's over there, but since '81 or so there's practically nothing sold into the US market without it. The signal shorting thing has turned up a lot, mostly due to aging wiring and careless handling of the connector during periodic sensor replacements. You really have to watch that connection at the end of the coax. In my case, and I think this experience is fairly common, at some time I had cut off and replaced the female spade at the coax end because the wire strands had fatigued, and by shortening the core wire I left the connector too close to the sheathing braids. When I slid the little rubber cover back over the connection the braids were able to touch the connector. The symptom didn't appear immediately but after some movement one of the braid wires came into contact, and boy do you know it when it does; gouts of black smoke, and about 5mpg. You could practically watch the fuel gauge falling driving down the road. I started running continuity checks from the ECU connector, and it was only by accident that I happened to see the signal circuit go to ground. Just trimming back the sheathing braid far enough that it couldn't touch the center wire was all it took to fix. Some other people have found their coax smashed or cut somewhere along its length; that would be especially vexing because it could be so hard to find, and the problem could be very intermittent with movement of the engine making the wire flex.

So Lee, good luck with that, maybe you get lucky and it's something that simple.

Nice forum you have here, I stick my nose in once in awhile, probably visit more often since I have a project Syncro now, I'm aways curious to see what others do with theirs.

Cheers, all!

[..lee..]

will put pics up asap.

[..lee..]

these are the green and grey`s i`m on about. you can see the grey disconnected and in this state it starts and idles fine hot or cold. although like this it wont rev cleanly.





and this is wherte they are in the engine bay. i`m praying to the heavens its going to be the signal wire faulty. i`ll be on it first thing tomorrow morning.

lee.

[tencentlife]

You have different wiring on the MV over there. I see it uses the same idle speed control valve, and we have just a single white wire with yellow connector coming out of the loom at that same location, which is the rpm signal for the idle speed controller; you're supposed to disconnect it to set up the base idle speed but I've never seen it make any difference. As for the green wires, I have no idea, but a small green wire is usually #1 coil tach signal (same as rpm). Time to dig into the diagrams for you.

And what is that thing just below those wires with the 2-pin connector on it and what looks like a knob on top? It's funny to see how different the same engine is built for a different market. At least i can see your O2 signal coax over there on the left, I'd have thought I'd lost my mind if that wasn't there.

[Aidan]

ok i'm getting there slowly
i've not seen that wiring set up before but it makes sense the way they've done it.
the green wire is the hall sender feed, from the sender and the side with the two wires is the hall and ecu side goes to pin 18 on the ecu, disconnection has the same result as on my earlier version where there's just single green wire either side, disconnects hall sender fom idle stabiliser conrol unit.
the grey is i think temp 2 sender signal ,if so the twin wire side may be ecu and will go to terminal 10 and the other side to pin 13/T on the idle stabiliser control unit, or maybe twin will be to ecu and idle stab, I'm not sure, easy enough to continuity test.
you need to know which wire goes into the connectors, could be the single is the ecu wire which would make this a quick way of disconnecting temp2 sender for ignition timing, and no burn the hand/plug on the manifold which makes most sense.
I can't be sure because this split connection isn't shown in last wiring diagram for 1990 onwards in Bentley, though the twinned connection for the green is shown there whereas it was elsewhere in earlier loom, no problem for later modification; anyway you should be able to trace the wiring to confirm this or not