930 hasitates on cruising speed (lambda problem)

[albinski]

hello Pelicans,

i have '86 930 US with lambda controlled mixture/emission. i am experiencing something that i call "square wheel syndrome" as it really feels like the wheels were not properly round

the problem occurs when cruising - it feels like something was literally holding up the car.

when pushing gas pedal to the floor all is okay, when driving hard all is good too.

the issue disappears when disconnecting o2 sensor but this is quick & dirty way to solve this.

any ideas where the problem may be?

o2 sensor is fine
seems the brain is fine
so is cycle valve / freq valve
TPS/TVS seems okay as well

thoughts welcome.

thanks and greetings from Poland

Krzysztof

[flightlead404]

Are you below 3000rpm when this is happening?

[albinski]

I would say - yes. Above 3k turbo comes in and then she goes like a goose
Speed relay you're thinking of?

[356911930]

Hi Krzysztof,

Have you checked for air leaks? Below 3,000 rpm if you have some air leaks after the mass air flow sensor you would have un-metered air and not enough fuel to compensate for the un-metered air. Under boost the vacuum turns to pressure, AFRs improve, and the engine runs more normal.

Rahl

[T77911S]

the air leak thing can really be tricky.
if the leak is on the pressure side then it can actually run RICH at steady speeds and on boost.
what happens is the turbo is still creating pressure but that pressure is going OUT the air leak, because the turbo is creating MORE pressure it needs MORE air. so it pulls in more air through the AFM making it run richer.
first hand experience with a BOV vented to atmosphere that was staying open at cruise.

how do you know the O2 is good.
if you unplug it and it runs better that could lead to the O2.

I would start with checking control pressures.
then unplug the O2 and check/set mixture.
connect O2 and if possible check duty cycle of freq valve.

before doing this you might do s tune up.
new plugs cap and rotor. set timing and check the advance

[albinski]

Thanks or all the comments.

Let me start from saying that the engine has gone thru top end rebuild, including fuel head and the the turbo.
Obvious ignition items are all new, so are injectors.
Fuel pressures are as per book.
Also Bov got new diaphgam as well as new spring

Hard for me to think of any leaks as we've gone thru pretty much everything already twice (if not more...)

Note the engine dies when bypass screw gets turned in completely.

Now, Why i consider o2 being okay?

1. It generates correct voltage
2. O2 sensor taken from other car (working properly) dod not change anything.
3. Afr at idle is at 14,7 when o2 is connected (goes down to 12.5 when disconnecting the sensor)

For me, It drives better on o2 disconnected simply because the system goes into open loop and there's no correction to mixture due to emission.

I was thinking of throttle switch 7° position that should provide some enrichment on throttle plate merely open but no idea whether it is a sensible lead and worth investigating at all.

Perhaps a drive with afr connected could offer some addn'l info to digest.

[flightlead404]

Closed loop mixture control via the lambda sensor is only on idle and just off idle. Has no effect when actually driving. I doubt this is anything to do with that circuit.

[pdx930]

Originally Posted by flightlead404

Closed loop mixture control via the lambda sensor is only on idle and just off idle. Has no effect when actually driving. I doubt this is anything to do with that circuit.

That is not what my wife and data shows. Below around 3000 rpm, the lambda is actively controlling mixture, and you can see it oscillating between rich and lean on my wideband O2 sensor. My 87 also feels like it is hesitating at steady state when driving and it correlates with the mixture going too lean 17:1-ish on the lean side of the oscillation. Very irritating.

I had a reputable shop go through it with a smoke machine looking for air leaks and changed injectors, O2 sensor, etc and found nothing. At the end of the day, I just unplugged the O2 sensor and adjusted the mixture with the wideband sensor and called it good. Drives sooo much better that way.

[albinski]

Originally Posted by pdx930

That is not what my wife and data shows. Below around 3000 rpm, the lambda is actively controlling mixture, and you can see it oscillating between rich and lean on my wideband O2 sensor. My 87 also feels like it is hesitating at steady state when driving and it correlates with the mixture going too lean 17:1-ish on the lean side of the oscillation. Very irritating.

I had a reputable shop go through it with a smoke machine looking for air leaks and changed injectors, O2 sensor, etc and found nothing. At the end of the day, I just unplugged the O2 sensor and adjusted the mixture with the wideband sensor and called it good. Drives sooo much better that way.

I am somewhat glad i am not the only one experiencing such thing; not saying i am happy pdx930 had this problem but being only one with a strange behaviour like this feels a bit like "hearing voices"

Now, i am yet to find out how afr changes across rpm ranges while driving; will report back here of course but my thinking here goes to 7° switch in tps, which should somehow offer a different mode of lambda system operation, something between 0° and 66° (wot in other words) where it gets to open loop state.

I have 86 factory service mamual that is describing lambda system operations but... it is all in German where i dont understand a word

There are also these temp senors - 15deg and 35 deg that somehow take part in the show but again, need to invest some time in translation to figure out how it all works together.

Back to solition applied by pdx930 - my mate who runs the shop is telling me the same thing for a longer while now but i am still a believer that it can be sorted out. Well, we shall see...
Any other ideas / oppinions or own experience cases welcome.


Thanks so far!

[pdx930]

One thing I can see in the wideband signal is that the frequency of the oscillation changes at idle compared to driving. Not sure whether that is driven by the 7 deg switch or rom or both.

My guess is that there is some analog circuit in the lambda unit that has an out of spec component that makes it overcontrol. I know they are designed to swing back and forth across the stochiometric point, but here it’s going too far.

[flightlead404]

Originally Posted by pdx930

That is not what my wife and data shows. Below around 3000 rpm, the lambda is actively controlling mixture, and you can see it oscillating between rich and lean on my wideband O2 sensor. My 87 also feels like it is hesitating at steady state when driving and it correlates with the mixture going too lean 17:1-ish on the lean side of the oscillation. Very irritating.

I had a reputable shop go through it with a smoke machine looking for air leaks and changed injectors, O2 sensor, etc and found nothing. At the end of the day, I just unplugged the O2 sensor and adjusted the mixture with the wideband sensor and called it good. Drives sooo much better that way.

I'll have to review the diags again but I think that its active below 3k and a certain throttle switch setting, so maybe at light cruise you're on and off it. have you checked the throttle switch?

[T77911S]

it goes open loop at either 3k or 3500, cant remember for sure. this is done in the speed relay.


12.5 at idle with it disconnected is very rich.
I don't deal with setting up lambada as mine is disconnected but I think you need to set the mixture per spec with the O2 not connected, with it this far off the lambda is at its limit to correct it back where it needs to be.

if you can check AFR while driving I would do that.


check you control pressures and check for power to the warm up reg since you are so rich at idle.

[flightlead404]

The manual has you disconnect the o2 sensor and set the mixture at idle to a specific %c0 setting. easy to find a conversion to AFR. This gets you within the range of control of the cis brain.

there are a lot of components that all need to work properly - or be ripped out

[mark houghton]

I'm one of those that said "the hell with it" and unplugged/tuned around the O2 sensor, taking lambda control completely out of the equation (well, except for the frequency valve that now runs steady open loop). These cars generally run better on the rich side vs. having the lambda constantly trying to push toward the lean side at the lower rpms.

[T77911S]

i agree, I would unplug it and remove as much electrical as I could, hey wait, I did that. even my FV is gone now.
but for those that want it original and even if my car was original I would keep it working.