Lizard928

Hello Gents,

I am working on a 90 S4 auto (ex Japanese car).

Car came in for timing belt/WP (inner bearing failed on WP).
Installed new Stage II w/ tri-flow camshafts into car at same time with new springs, and lifters. Cams set to +2 adv.
Top end remained untouched. Other deferred maintenance is needed.

Car ran perfect for nearly a week with the customer, then it started to run rough at all RPMs. I went and inspected, the ign leads were arcing (had told to replace awhile ago), so they got replaced along with the caps.

The car still ran rough, replaced the MAF with a recently rebuilt one. Car ran much better. Customer asked to adjust the idle up more. I used the JDS idle controller in STII and attempted to raise it. Car would studder and not return to a nice idle after higher RPMs. This was if just revved, or driving.
Returned to the stock idle controller, and car ran much better.
Doing a high RPM run in the parking lot there was no stuttering on decel like before. Burnt chips with a couple modifications, installed into his brains, and off he went.
He let me know the car ran famously the entire way home.
Following morning car would barely idle and had NO power, however once warm car apparently ran perfect.

I brought the car back to my shop and have found that the MAF signal is 190-200 cold, with Inj PW being high 4-low 5mS. Effectively running rich enough to cause rich misfire. I tested my car and it never gets above 3 mS when cold.

I have now tried another MAF, inspected all Temp II wires, MAF wires, all are intact with no coating degradation.
When the vehicle is warm, the idle stumble is still present though with normal driving you simply dont notice it as you arent returning to idle vigorously from higher RPMs.
When the car is idling the MAF signal sits around 135-145, and the injector PWs sit between 2.9-3.1mS. I have tried 3 sets of brains (two with PEMS, one rebuilt, one not), all function the same. Plugging these brains with the same code/settings etc. into my car and my car runs flawlessly. I can adjust the fuel maps up/down and the injector PWs go up and down according to the changes I make.
Taking these same brains back to the car in question and I can pull the fuel map down to -127 and up to 40 and the Inj PW remain at 2.9-3.1mS (car idles the same, and responds the same regardless). This doesnt matter if the O2 us used or not. (cat vs/ no cat, connected or not).
Even pulling the injector sizing from 19# up to 30# has zero effect on the PW.

I have put in new injectors, no change.
I have now cleaned all the grounds no change.
On my car I see the Inj PW are on mine 2.4mS, with a MAF signal of high 90s at idle.
On this car they are 130-140 MAF signal and 2.9-3.1 mS.
Plugs are new WR7DC+
Battery connections (- and +), are all cleaned,

I am at this time a little stumped as to what could be causing the injector PW to not be able to change.

Any suggestions are eagerly welcome!

sendarius

Take this with due deference to my knowledge - which verges on zero.

Are you sure that the MAF is connected? That certainly sounds like a REAL case of "limp-home mode".

IIUC, the LH sees no MAF, so goes to a fixed pulse-width vs RPM map.

Bill Ball

Steve did well for supposedly verging on zero knowledge.

You've shown SOMETHING is wrong with the MAF signal. Does it run the same with the MAF unplugged? Can you run Sharktuner with the MAF disconnected and look at the fuel map readings and see if they are similar.

If so, then the harness and plug need another going over.

No Cats or O2 sensor I assume. Also, can you verify that the Hall sender is working. If not, it really shouldn't have much effect, as you know (-6 degrees advance at higher RPMs), but I've seen cars that were almost undriveable when the harness was shorted.

John Speake (or Jim Corenman), where are you?

the flyin' scotsman

Colin...........Ive worked on a few exJapanese 928s where the intake rubber components were well worn through from great many heat soaks the car is exposed to while stuck in Tokyo traffic. One was so bad the insulation for the CPS was flaking off.

Might be worth a look.

Lizard928

Steve,

I'm, sure the MAF is connected. Tried two separately.

Bill,

Haven't tried with MAF unplugged, will test first thing in AM.
It does have an O2 sensor, and cats (required for emissions when brought into country).
I used the STII to force cat mode to ensure use of the sensor and ignoring the coding plug.

Malcolm,

I've worked on more ex-japanese 928s than most people I think. And I certainly understand what you are referring to about the underhood idling heat issue. However this car's harness (LH/EZK at least), is absolutely mint. The wires are nice and flexible, none of the boots are torn, you can still roll the boot on the MAF connector up and off to see the wires without it breaking in half!

the flyin' scotsman

did the car display any of these symptoms prior to having work done?

I recall some other component that's only fitted to Japanese models that forces 4 cylinder mode


found it......ignition circuit monitoring relay

On very rare occaisions the GTS goes into 4 cyl mode........I disconnect the battery

chitown928s4

My recent limp mode turned out to be the ignition modules in by the drivers headlight...check for corrosion there are ground wires under there too.

GregBBRD

Individually check each wire going from the brains to the engine for continuity and then check for wires crossing over to another wire. Especially MAF wires.

Steve could be correct.

Connected but not connected?

sendarius

I am sure that someone, somewhere is tearing their hair out

The IMS-induced 4-cylinder mode is NOT limp-home mode - somebody here better described it as "I've prevented a cat fire, but stop now, or I'll break the drive shaft mode"

Colin: I don't mean to harp. The MAF is plugged in, but is it connected?
Your AM test with it un-plugged will be revealing.

Edit: I see GB beat me to it - that's what I was trying to say.

Lizard928

Malcolm,

The Ignition monitoring relay would engage if you tried to drive when cold and running terrible.
The Jap cars also had two temp probes after the cat which would illuminate light on the dash to tell you that the cats were too hot. Does not interfere with the running operation.

Greg,

Thank you, I will be checking all MAF wires for continuity and ensure that there are no short circuits first thing tomorrow.

Chitown,

That grounding point has been cleaned, however the final stages by the drivers headlight are actually grounded to a single spot under the 14 pin connector/jump post on the passenger side of the engine bay.

GregBBRD

And stick your favorite bore scope or mini camera down under the intake and make sure that one of the hoses from the idle stabilizer hasn't split along the seam, like they seem to "love" to do.

Sometimes tough to see the "back side" and you just have to break down and pull the manifold.

How does a large amount of air not running through the MAF factor into your problem?

Lizard928

I've had the split ISV hoses before. I've never had trouble with diagnosing them.

I cannot see a split ISV creating the problem, but post MAF connector dissection, I am likely going to pull the intake to inspect them all. The one I'm more concerned of being split is from the filler neck to the Y going in between the MAF and TB.

John Speake

I wouldn't use the ST2 programmable idle controller until you were quite sure other issues weren't still present.

Jim wil be producing some user notes for this system in due course, it's ot officially released yet.

jcorenman

Colin, Sorry I missed this last night, we are still struggling with internet problems here.

The most basic symptom seems to be the high MAF reading, and the (relatively) unchanging injector PW-- I would focus on that, the rest seems like distraction.

Inspection is good but as was said above, not adequate to determine that the signals are actually connected. ST shows temp-II values for on both LH and EZK pages, are those readings both correct within 5-10 deg? What about MAF volts? That seems to be a prime suspect, your MAF signal is high and unchanging-- but what about ST's MAF-volts display? It should show around 2.6-2.8 volts at idle, and vary with throttle. If that is not right, then that is the primary issue.

It could also be a problem on the ground side, every electrical signal measurement needs a ground reference as the basis for the measurement. The signal-ground for the LH is the engine block, at the two ground points at the back of the engine under the airbox. Dive down there and check/clean those. Then clean and check the main ground strap under the engine, passenger side. Even a slight resistance in the main engine ground-strap can result in a large voltage drop between block and chassis grounds, caused by alternator current.

As John mentioned, ST's new idle controller cannot fix a poorly-running motor, and won't do any better than the stock idle for a relatively stock motor. It's purpose is for extensively-modified engines that won't happily idle with the stock idle control.

Our phones might be working today, and I'm around all day if you want to give that a try. Or PM a number and a time and I can call you, that might work better.

Cheers, Jim

FredR

Colin,

Given your knowledge/abilities and what you have tried it seems you have pretty much eliminated the obvious so somehow we are looking for the "not so obvious". You know that MAF and LH/EZ problems are not in the equation and the car runs OK when warm off tickover- or does it?

The injectors are not getting a signal that you would expect to see. Are you seeing a fixed output from the LH no matter what throttle position or just at tickover? We know that the MAF should not be in limp home mode but presumably the LH has a diagnostic ability to sense a faulty MAF signal and thus [when hot] generate a pulse width suitable for supporting a low load/lowish rpm running?

There is no direct link from the MAF to the injectors so logic suggests the LH is reading a fault from the MAF even though you think one does not exist because the MAF works in a different car.

If I am reading the above correctly logic suggests the MAF signal is being corrupted and the LH is applying a limp home signal. Thus check the MAF signal wires through to the LH or wherever they go to interim and test for continuity and resistance to earth/ no continuity between adjacent wires.

Doubtless you have this all sussed by now but just in case.

Bull**** Baffles Brains- every time!

Good luck

Fred