Was my car born with an O2 sensor?(now missing...)

[Rookie1mbit]

Hi all, new member here.
I have a 1989/90 VW Passat Variant GT 2.0 16V with a 9A in it.
White label in the back says (100kw) GKAT 9A, (etc.), so guess it's the correct engine for the car.

Now, my problem is fuel economy, which is 23.5mpg, no matter what I do.
This is according to actual volume "fed" to the tank and odometer reading, NOT the weird MFA...

I have a good deal of experience jetting both EFI and carburetted car/MC engines with my Innovate LM-2, and have done so with this car too.
It runs stoich at part throttle(14.7'ish), and about 13:1 at WOT.
Compression is measured to 16 bars(232psi) on all 4 cyls.

Ignition is at the mark ( the line, not 0) at idle when I strobe it, and is not more bouncy than any other stuff I've strobed.

Now, I found a double connector alongside a single, male, yellowish connector at the passenger side, in the engine compartment, at the firewall.
They are NOT connected to anything, but there seems to be 15'ish volts between the poles in the double connector, and 500mV'ish + in the yellow thing.

There's no cat nor lambda in the car....not even an M18x1.5 bung.

Won't the ecu go into limp mode, if it lacks the lambda, and switch to a very non-economic fuel- and ignitionmap?
Of course, I have compensated for that fuelmap, by adjusting the co-screw and the electrical valve thingy on the side of the metering head, but won't it still run a less than optimal ignitionmap? (thinking the reason for bad economy here)

Car is very touchy on that CO screw and idle, I can adjust those like every week.....

I've hooked the LM-2 to the 4-gas tube that sits on the ex manifold with a special adapter I made, cause I can't reach the exhaust under the car, when it sits in the driveway....

Plugs are tan



So, is the ECU running in limp mode, making the car use too much gas?

What do I do, install a lambda eliminator, or install a new lambda in the exhaust, and recalibrate?


Regards, Rookie, Denmark.

[WaterWheels]

1989/90 VW Passat Variant GT 2.0 16V with a 9A in it.
White label in the back says (100kw) GKAT. . .

my problem is fuel economy, which is 23.5mpg . . .

There's no cat nor lambda in the car....not even an M18x1.5 bung.

So, is the ECU running in limp mode, making the car use too much gas?

Even if the sticker didn't indicate that the vehicle has a Katalysator I would think that year would. That "GKAT" stands for gereglter katalysator which means nothing more then "regulated catalytic converter. Before closed systems cam into mass-production they were listed as UKAT (ungegelater) due to the lack of feed-back to cntrol the mixture. So yes, your car was born with both an O2 sensor and catalytic converter.

23.5mpg is really not that bad for a 2 liter engine in a some what heavy car. It works out to 10 liters per 100 kilimeters if you use that system, not great but not all that bad. I would guess it will get a little better if you install the missing parts. It is really not a "limp home" mode but rather the system reverts to just the basic fuel map, the O2 sensor provides fine tuning to this map which is what you are missing right now. The connectors you found do sound like the ones that were used for the O2 sensor but I would check a wiring diagram before connecting to avoid any electrical damage.

The CAT itself has the bung for the sensor in the front section, so installing a CAT will provide the place for the sensor. I would not expect to gain much more than one half to one liter per 100 as that would be a good fuel consumption for a 2.0L 16v of that year (around 24-26mpg).

[Rookie1mbit]

OK, I see what you mean, regarding economy etc., it might not work wonders.

Worst of all problems is the ever changing idle mixture, which is making the car idle fine one day, and hunt like crazy the next.
Guess installing the narrowband lambda will eliminate this?

Last measurement says around 900kms on 75ish liters of fuel, which is 12km/l or 28mpg.

The Innovate LM-2 also has the capability of reading OBDII....I know this car doesn't have OBD, but I read somewhere, that you can buy an adapter from 2x2 to OBDII jack? would be nice to be able to read the ECU, and clear eventual alarms.

Last, just WHAT is that round, very deep hole in the lower, left corner of the clear glass in the dashboard, like to the left of the buttons to adjust the clock with?
Looks like there are a bunch of poles at the bottom of this deep hole....?


Regards.Rooks.

[WaterWheels]

Last measurement says around 900kms on 75ish liters of fuel, which is 12km/l or 28mpg.

Last, just WHAT is that round, very deep hole in the lower, left corner of the clear glass in the dashboard, like to the left of the buttons to adjust the clock with?

OK, something is wrong with your math or figures. If you traveled 900km and used 75 liters of gasoline that would be 8,3 L/100 or about 34mpg, 28mpg is 10,9 L/100.

I don't have a Passat dash to look at what you mean but it sounds like the "trip-o-meter", which I'm sure has a stupid correct name. You should have two sets of readings in the gauge, total km traveled and one you can reset whenever you like for whatever you like (most time it only goes up to 100km with a tenth reading on the right end). There should be a black pin kind of button to push to reset the trip meter comming from that hole. It is threaded and yours may have come un-screwed or was removed for some reason.

[Rookie1mbit]

900kms/75L fuel is 12 km per liter, which is 8.33L/100km, correct.

Nono, not the resetable trip-meter, it's dead in the left, lower CORNER of the screen.
It has a black plug, when you remove it, there's a like 1" deep hole, but with 7 contact surfaces at the bottom, looks like the surfaces are situated on the print.

Tried flashing the codes yesterday, but no flashing at all.....isn't the LED supposed to have a resistor on one of the legs, to keep it from frying?

Regards.Rooks.

[WaterWheels]

900kms/75L fuel is 12 km per liter, which is 8.33L/100km, correct.

. . . it's dead in the left, lower CORNER of the screen.
It has a black plug, when you remove it, there's a like 1" deep hole, but with 7 contact surfaces at the bottom. . .

.....isn't the LED supposed to have a resistor on one of the legs, to keep it from frying?

"900kms/75L fuel is 12 km per liter, which is 8.33L/100km, correct." Yes correct but it is not 28 mpg but 34 mpg. That's the reason for my response because 28 mpg would be 10,9 L/100 which is what was wrong above, or I just read it wrong.

Yea, looked at ETKA to see if I could figure it out and all I can tell is that vehicles with the multi function indicator gauge have that plug. So I would assume you have a multi function indicator gauge cluster, right? Just what the connection does I can't tell you, ask the local dealer I guess or someone with the same dash.

If you are testing 12v then yes an LED needs to have a resistor in the wiring. Is this a self made test light or a purchased one? If you purchased it then I would think a resistor was placed in the test light wiring.

[Rookie1mbit]

Right, I guess it's the usual confusion between standards, here in DK, we never ever use the l/100km thing, but they do i Germany. We only use km/l here.

I installed a sond today, drove it until warm, then on the freeway, and then some town-driving.

Low throttle driving has improved very much after the sond was installed, but, once it passes 90C water temp, the idle hunts like crazy.

Do I need to adjust the CO at idle again now, and do I need to clear eventual errors from the ECU?

It's a homemade test light, I'll install a new LED along with a resistor, and try reading out again.

Regards,Rooks.

[Rookie1mbit]

Right, only codes showing now are these 4:
4341, 4343, 4431 and 4443.
Reading threads elsewhere, I thought these were error codes, but they are NOT.
According to a VW manual I found, it's a manual activation procedure of the DPR, purge valve, IACV and CSV.
Maybe these can be turned off by the ECU when "real" errors occur, and need to be turned on afterwards manually.
I hear the IACV clicking, once I activate it with the jumper

Also, today I tried measuring the amps to the DPR, but no luck : 0mA.
Must have been because the DPR was still turned off...

So, now I just turned the 4 components back on, but haven't gone for a ride, it's like 00:57 now.
But, idle control works, it started out at like 800rpms, but slowly raised it to 1100rpms.

More tomorrow

Regards.Rooks.

[Rookie1mbit]

Forgot, here's the activation- and clearing procedure from that manual:

Erasing
Ensure ignition switched OFF.
Connect LED test lamp between black data link connector
(DLC) terminal 2 LI [A] and blue data link connector (DLC)
terminal 2 LI [C].
Bridge black data link connector (DLC) terminal 1 [A] and
brown/white/yellow data link connector (DLC) terminal 1
[B] with a switched lead - contacts normally open.
Access trouble codes. Switch ignition OFF.
Operate switch. Switch ignition ON after 4-6 seconds.
Release switch after 4-6 seconds. Check that LED flashes.
Differential pressure regulator activated. Trouble code
displayed: 4341.
Trouble code identification
Operate switch for 4-6 seconds to activate each of the
following components:
Evaporative emission (EVAP) canister purge valve. Trouble
code displayed: 4343.
ldle air control (IAC) valve. Trouble code displayed: 4431.
Cold start injector. Trouble code displayed: 4443.
After activation of last component operate switch for
4-6 seconds to display flash code 0000 (long flashes).
Operate switch for 4-6 seconds to erase fault memory.
Repeat checking procedure to ensure no data remains in
ECM fault memory.

[ziddey]

It shouldn't be hunting without lambda feedback I imagine. Have you tested your CTS? You can try shunting the plug to the CTS (read: you're supposed to use a resistor to simulate a hot CTS, but I've seen no damage to my ECUs shunting the plug with a paper clip)

This is cis motronic? The idle screw should be all the way closed for your application.

So you are getting 0ma through the dpr no matter what? Or only after it warms up?

[ellocolindo]

The passat 16v 9a engine code has cis-e motronic. The o2 sensor has to be connected to get about 1100 kms to a gas tank on the highway at a steady cruise control speed of 100 kms/h. I get that on a rebuilt engine with 60000 kms on it since rebuilt.

In city i get 900 kms to a tank.

The plug you ask about on the dash; it has to do with recalibrating the cluster from factory.
Only the factory had the equipment for that. Not even the dealers.

[Rookie1mbit]

Yeah(0mAh when warm or cold), but now I think it was during the deactivation of the DPR, as described above.
I have to check again.

But, my lambda is not installed correctly at the moment, it's hooked up to the 4-gas testing tube at the ex.manifold, and I suspect it's not getting warm enough that way.

I'll get a bung welded in next week.

Also, I just checked the potmeter at the flapper; 0.2V at idle position, then quickly rising to 2V when I lift the flapper, but then no more....it's supposed to rise to 7-8V.

ECU codes say "DPR reached lean limit -5mA" and "potmeter not working".

I found a new generic potmeter at ebay, whatever generic means.....gonna try that one.
Looks exactly the same part as the VW pot., and at $45 or so, it's worth a try.
It's listed under the same VW part number as my OEM one has.

Guess the ECU is trying to lean out the mix, but somehow can't due to reaching it's lean limit.
I'll clear the codes and try again, once the lambda has been resituated.

Oh, also, car hasn't got a WOT switch, only the fuel cut-off switch at idle pos.

If I turn the idle screw all the way in, the engine will stall.

Regards.Rooks.

[Rookie1mbit]

I currently get like 950kms on a tankful, mixed driving, but mostly countryside.

So, might not be much to gain on economy, but I'd like the car to at least drive properly.
Reacts very poorly on throttle openings, and WOT is very dull.

By the way, I have a complete spare 9A in the garage, but the potmeter on that one is also shot....

Regards.Rooks.

[Rookie1mbit]

Oh yeah, just to add to the misery, ABS light is flashing out codes.
First, it flashed 1241, so I changed the right front speed sensor.
Then it flashed 1233, so I'll change the left one too, and see if it stops flashing codes at me
ABS works fine though, but the light also flashes occasionally when driving on rough roads, or turning hard.

Regards.Rooks.

[Rookie1mbit]

Just hit me....what's the + terminal for feed voltage supposed to measure on the flapper potmeter?
Mine says just below 5V.

If that's correct, then how will it ever put out 7-8V on the signal???

Regards.Rooks.

[ellocolindo]

get yourself a bentley for the car. all these questions you are asking will be answered by it and you will gain a general understanding of the complete fuel injection system in the car...

[Rookie1mbit]

"Compression is measured to 16 bars(232psi) on all 4 cyls."....isn't this extremely high?

Plugs are still tan (so prolly no carbon build-up inside chambers), head has not been skimmed excessively (compared it to the extra head I have).
Engine knocks(pings) 1-2 times when starting, then the knock sensors take care of it.
Tried driving like 2 miles without the knock sensors screwed onto the block, but not disconnected.
Engine knocks at the LEAST bit of throttle opening, no matter what rpms.
Virtually impossible to drive the car without it knocking like crazy!

Dyno says 103-ish HP at engine.

My thought is, that it, for some reason, retards the ignition so much, that it loses both HP and mileage.

Any thoughts?

[ziddey]

CIS Motronic "limp" mode will not affect ignition timing. It is not so much limp mode so much as constant open loop.

Doesn't the LM-2 have a narrow band simulator? If so, you could always use that to make CIS-M go into closed loop.

I'm assuming your 14.7 and 13 AFRs are as seen with the LM-2? If so, you've done a decent job at tuning the mixture. You say that ignition timing is at the line and not 0, which should be 6*BTDC, which is right as well.

Is it actually a knocking you are experiencing, or more of a miss? It'd still point to a mixture issue in that event. Have you checked that you are getting even spray across all four injectors? I know you said you pulled all the plugs, so I'm imagining they're all pretty even. Getting good spark across all 4? Any change in behavior if you further enrichen it a bit? Or lean it out?

The plugs you mentioned in your first post sound like the lambda wires. The set of two that you see 15v across sounds like the o2 heater plug. The other one that you're measuring around 500mv would be the lambda wire (I'm assuming it's a thick green wire with a brown plug?).

[WaterWheels]

"Compression is measured to 16 bars(232psi) on all 4 cyls."....isn't this extremely high?

Engine knocks(pings) 1-2 times when starting . . . Engine knocks at the LEAST bit of throttle opening, no matter what rpms. Virtually impossible to drive the car without it knocking like crazy!

Not going to say it is but it sounds very much like a problem I had some years back when I played with 16v engines. Strange as it may sound, the timing chain between the two camshafts was my problem. It took me a long time to discover this and it really drove me crazy as everything seemed fine but as soon as I hit the throttle, showed most in first and second gear, I could hear a knocking sound. Even in fith gear it would do this but just not as bad. This was not really the same sound as a ping you get with ignition problems, but more of a light metalic knock. The chain when worn enough starts to slap around as the engine speed changes and about 20 Euro for a new chain and my time to install it cured the knocking. I would say it is worth the time to pop the valve cover and look to see if your chain has some slack in it.

16bar does seem a little high but I wouldn't give it any thought.

[Rookie1mbit]

OK, the AFR values are from the LM-2, and yes, it can simulate narrowband.
But I'll just weld in an extra bung for the LM-2, so no simulation necessary

I checked the slack in the cam chain, and it's about 1/8 of a tooth on the timing belt.
Same slack as in the spare head I have lying around.

I'm sure it's knocking, but that went away, as I discovered, that the timing at idle was 12 degress, rather than 6....

I'll have to check the AFR again, as I have only mapped it with a makeshift bung and a hose off the stainless tube in the manifold.
I suspect it might have been showing too lean results overall, due to reversing to fresh air.
Car runs much better cold, where it might "eat" the extra gas better than when hot.

Does anyone know what HP figures these engines generally put out when dyno'ed?
I use Dynomaster, an app for android, and for serious work, a friend runs a Maha 4x4 dyno with Eddy brakes.

[Rookie1mbit]

Right, got the thing dyno'ed....129HP on the engine, 35HP train loss, that must be ok.
Now, the ignition acts weird.
At no resistance, its at 16 BTDC@3000rpms and 20 deg@4000rpms.....

I changed the knock sensors, no difference in timing at all.
If I unhook the sensors, still the same ign. advance.....it should be around 30 deg BTDC@2300rpms, I think?

ECU only has lambda error, as the signal from it is 0V...but does that affect ignition?

What else controls ignition? isn't it just rpm signal and knock sensors, or does that damned potmeter at the flapper affect it too?