Loose connection?

Hi Maury,

That's a disconnected oxygen sensor. The green wire with the male end is from the K-Jetronic/lambda computer and plugs into the other female end from the oxygen sensor (black circled connector). Someone's stuck a wire connector over the exposed end to prevent it from grounding out. The other 2-wire plug you see is the oxygen sensor heater.

I believe this is a not-uncommon method of removing the lambda feedback and subseqently "tuning" the fuel injection mixture to pass emissions.

Best,

Carl

 

Hi Maury,

Do you know how long the car has been in Louisiana? Possible it came from somewhere else?

I would reconnect and see how the car runs. The lamda "closed-loop" doesn't begin until the car is warm, and then you will see a bit of "cycling" of the idle speed as the mixture is tuned around stochiometric. Disconnecting the oxygen sensor doesn't lead to any permanent damage (unless mixture is far off). There's always a possibility it was disconnected for a "reason" (aforementioned tuning, non-functioning sensor) or the mixture was changed afterwards, so you'll have to see how much a a change it effects. Removing the oxygen sensor allows the system to run "open-loop" and essentially converts into a basic K-Jetronic system.

Here's a useful link to explain the oxygen sensor:


http://www.volvoclub.org.uk/tech/fuel_injection/k_jetronic4.shtml
Best,

Carl

 

Carl, you never cease to amaze...Heck you should open a shop for those poor souls like me... I read that article and felt like the caveman on the Geico commercial....ahhhh what?....lol..lol
Hope you & your car(s) are well!

Mike

 

It's not quite that simple, disconnection of the O2 sensor will fail to give a signal to the control unit that controls the frequency valve. the FV operates the duty cycle, the duty cycle controls the rich/lean mixture to achieve stoichiometric.

grounding the lambda will result in a 95% duty cycle and a rich condition, leaving it open will only give a 60% cycle, so at WOT you run the risk of going lean. yours is not grounded. also the thermoswitch sends a signal to the control unit to cycle at 50% once engine temp is up, with no feedback from the O2 sensor the car will run rough and rich/lean.

in order to go back to k-basic the control unit and frequency valve would also need to be removed. if you still have cats then I'd be very concerned about fire from a rich mixture. otherwise the fixed 50~60% cycle isn't doing you any favors.

a mistake alot of shops make with CIS is assuming it operates like the other Bosch FI, it doesn't. the CIS is what I'd call glorified carbs, it's a hydraulic fuel system.

if you'd like to set it right, give Larry F a call 251-929-3771, he's the CIS guru.

K-basic has no frequency valve or duty cycle it's just basic fuel to air metering.

 

hey scott,

I don't really follow this. Carl just said that a k-lambda system with the 02 sensor disconnected behaved like a k-basic.

If the O2 sensor is disconnected, then the frequency valve operates at a fixed duty cycle (whether it's 50% or 60%, and whether the ecu looks at other inputs to switch between them I don't know), but one way to interpret what ya said is it runs at 60% until warm, and then 50% after.

the assumption is that the base mixture is set for stoichiometric when the fvalve is operating at 50% so the frequency valve has roughly an equal range of compensation on both sides. If you force the fvalve to run at a fixed 50% cycle by disconnecting the o2 sensor, then the fuel mixture == the base mixture and carl's comment that it acts like k-basic seems correct.

if higher fvalve duty cycle equates to richer mixture, and you unplug the freq valve, then the mixture would go lean and you'd need to tweak the mixture adjust screw to bring it back. Then you'd have a true k-basic setup.

Where's the rich mixture/cat fire concern coming from?

do a search for people's results. The conclusion I got was it made no significant difference in performance (assuming your cat isn't plugged up), but gained you a bunch of stink. If you really want to minimize any effect of the cat, replace it with one of the modern high-flow ones, but I doubt you'd notice a difference.

The o2 sensor on the 3x8 cars that have them is upstream of the cat, so the system would still function. However, the point of the lambda system is to optimize the exhaust gases for the cat to do it's work. If you don't have a cat, then you disable the lambda and usually run rich enough to kill squirrels and roadside vegetation in your wake.

 

the KE and K-basic are similar but operate differently. the FV will continue to adjust the control pressure even with the O2 disconnected, the K-basics lack of the FV means that it's control pressure is only coming from the WUR, there is no interplay.

the FV when receiving its inputs will keep a 14.7:1 ratio, at a fixed 50% duty cycle there is no range and control pressure is now fixed, the K-basic does not operate with a duty cycle so there is no fixed percentage.

Lambda-sensor signals are processed with the other inputs to determine pressure-actuator current. that is the potentiometer current based on plate movement. with a warm engine and no lambda feedback the actuator current will have no change, ie WOT. at part throttle acceleration where you need enrichment the lack of a signal will not enrich the mixture and a brief lean condition will occur.
also the KE uses a fuel cut off for coasting based upon switching the actuator current, with out a lambda signal I don't know how this is affected it may or may not operate, if it does not operate then raw fuel gets dumped into the exhaust.

the mixture adjustment screw is not the answer to 'bypass' the FV, yes it will adjust the CO but only to a point, it's a broad range adjustment. the K-basic also is set up the same way. the only way to have the k-basic is to swap out to the k-basic.

a lack of stoichiometric control and unknown sensor interplay with a no signal O2.

any car can operate with cats, the danger is the non lambda systems. with no feedback if the mixture runs rich the raw fuel will ignite in the cats and catch fire. if you're running rich enough to tear your eyes with cats, a fire is not far behind.

bottom line, if you keep the cats keep the O2 functioning as it will allow the Ke system to function as designed. disconnecting the O2 sensor in of itself won't gain you any performance, only headaches when the car runs 'funny'

just my 2cents

 

hey scott,

granted KE and K-basic are significantly different...but he has a 328. It has K-lamba, not KE.

I guess it's really something in between K-lambda and KE, since the injection ECU is connected to more stuff, but there's no potentiometer-style TPS or plate position/airflow sensor, and the method of fuel control is still varying the chambers in the metering head.

did ferrari ever publish what exactly the injection ECU is doing? I think the assumption has been that the frequency valve is only responding to the O2 sensor output, but the ecu does connect to other things like the oil temp switch, water temp switch, WOT switch, and exhaust ecu, and not all of them are just flipping the system between open and closed loop operation.

which if those connections are inputs and which are outputs is not completely clear, but the manual either states or implies that:

1] if the oil temp switch or WOT switch is closed, the engine wants enrichment and the fv may be set to 60% fixed
2] otherwise, if the water temp switch says it's hot, then closed loop operation is enabled and the fv will fluctuate with O2 output.
3] bosch docs imply that if the O2 output is out of range, then the fv will be set to 50% fixed assuming no other sensor wants it someplace else.
4] the exhaust ecu connection I don't think has been explained, but it would seem reasonable that the ecu would want to lean out the mixture. The closed throttle switch also connects to the same ecu pin through a diode.

my understanding is that the correct way to set the base fuel mixture is to warm up the car into closed loop mode, then disconnect the O2 sensor. That puts the fv at 50% fixed, and you adjust the mixture screw as close to stoichiometric as you can. Then plug the O2 sensor back in and let the closed loop operation do the final tweaks.

in any case, I see your point that just disconnecting the O2 sensor isn't exactly like k-basic, since the ecu is still seeing other inputs. Whether it's responding to them I don't know, but most of them would result in a too-lean situation rather than too-rich, so a cat fire is not likely to result.

I'd also agree that if you have cats, there's no reason to disconnect the O2 sensor. I suppose if you did you could enrich the mixture a bit for a small performance boost and not have the system lean it back out, and if you don't overdo it the cat won't torch, but doesn't seem worth it.

If disconnecting the O2 sensor fixes a running problem, the either the sensor is bad, the air injection system is letting in air when it shouldn't be, or the ecu has a problem.

I finally got a scope, and one of these days I'm going to stick it on the fv and see what's going on, unless someone can find a doc that details what the ferrari injection ecu is doing (rather than what a generic bosch system does).

 

woops I meant K-lambda not KE, KE was on the testarossa. so the pressure actuator info is not applicable, sorry about that. teach me to reply at midnight. the lower chamber pressure is controlled by the FV in lambda whereas it's system pressure in the k-basic. so in order to get a lambda to operate like basic the lower chamber would need to read system pressure, which it can't.

but you're right, there are alot of sensor inputs and no known info on how they interact. even the bosch ECU info isn't readily available. I'd say one way to check is to measure the dwell time while connecting/disconnecting the various sensors to see if a change is effected.

do you have the bosch book with the lambda wiring diagram? I can make you a copy if you want it.