Rob Edwards

For those playing at home, by TB service Cardinal means 'throttle body', namely the replacement of the HK1012RS bearings on the throttle shaft.

Don't let Clarkson near the car.

worf928

Prior to seeing the 'smoking gun' on this thread I had already chased and eliminated all known causes of high idle to my satisfaction. A final check of pressurizing the intake and listening/spraying for leaks confirmed for me that there was no false air. The final straw was hooking up my 'locally manufactured Porsche special tool' that showed the A/F mixture super-lean. Then I plugged in my spare LH.

Since my spare LH cause this car to idle at the spec I do not plan to chase the idle further.

No emission test in Massachusetts. So, no data available. But based upon my O2 gizmo results, it would have failure on NOx (and probably also HC since the LH would have been dumping fuel to try to balance the 02.)

CardinalCar

Ok -- if you're talking to John and/or Louie, I would be very interested to know if there are any parameters specific to a refurbished LH which could adversely effect idle behavior. It seems like a highly improbable failure mode but you have now doubled my sample size for the investigation so I'd prefer not to let the idea drop.

Unfortunately I have no means to interrogate the car at that detail on my own, but your description of pursuing idle leaks is entirely comparable to the efforts Rob underwent, with similarly fruitless results.

If the LH itself is causing the high idle behavior in your car and in mine, I wonder whether something that's part of the rebuild protocol is the cause. A new resistor, a change in logic flow somehow, who knows what, but I've previously commented to Rob that the rise in idle speed as the car warms up feels a lot more like a controlled function than a more primitive response to an air leak somewhere.

worf928

I've replaced ... a dozen or so ... LH's with rebuilds from both John and Rich and not seen anything like this..

For small values of 'controlled' perhaps...

I did a lot of black-box testing of the idle control loop when I worked on the tuning of the twin screw. It - the control loop in OE form - is exceedingly stupid and 'unadaptive' outside of a very narrow window of operation.

CardinalCar

Fair enough -- do you have any idea what about that window you specifically changed by swapping in the second LH which cured your high idle?

I've asked John and Louie about this as well this evening and haven't ruled out coincidence obviously but if you know how to isolate some or all of the relevant parameters I might be able to improve the quality of my own testing.

John Speake

You may be interested to know that the Sharktuner now has a fully programmable robust alternative idle control loop. This has proved very useful with strokers + hotter cams which the stock loop couldn't cope with.

As far as the 3v on the O2 line is concerned, I've only seen one before (in over 1,000 LH rebuilt) and that was many years ago. It cerainly affected emmissions but I don't recall a side effect of high idle...

worf928

After exchanges with Louie this morning, I'm reminded that this LH I'm working with has Autothority chips. So, that is a BIG variable. Had I known/remembered that this one had them I would have swapped them out (after getting permission from John to break his seal) because they usually (in my experience) have weird, inconsistent, idle 'gremlins.'

The main thing about the OE idle loop is that it has a limited range of adjustment. I'm sure John knows more on this subject though.

I've never had a 'bad' rebuild.

John Speake

Most of the Autothority chips I've seen here, whether in an S4 or Gt LH ECU have the idle speed set to the GT's 775rpm.

CardinalCar

At last. Victory. After six attempts.

It is very clear that John and Louie's efforts were instrumental in getting the car to pass. The original rebuilt ECU (from seven years ago) did indeed develop a secondary fault in some aspect of the oxygen sensor control circuit. The car still ran fine, but it was very rich, and once we replaced the ECU with another rebuilt unit, the O2 loop began to work again as expected. So thanks to John for standing by his product long after the original warranty expired, and thanks to Louie for helping with the logistics within the USA so I didn't have to deal with trans-Atlantic shipping.

However I do think that during the rich running (roughly 1000-1500 miles worth at best guess), the cat suffered a bit, and now really only works when it's quite hot. Under real-world conditions this is fine, as it will certainly get hot enough to light off and be fine. But if the car sits turned off waiting for its turn before getting into the test bay, it's unlikely to be hot enough.

I also discovered Virginia has two different types of tests. The two-speed idle test is not done on a dyno. It's an option for cars with mechanical issues that prevent it from being strapped onto the rollers. The limits for that test are defined in terms of only HC ppm and NO%, whereas the dyno test measures those factors plus NO ppm.

HOWEVER the very odd thing is the different limits themselves between the two test types:

Two-speed idle test
HC ppm @ idle: 125
HC ppm @ 2500rpm: 125
CO% @ idle: 1.00
CO% @ 2500rpm: 1.00

Dyno test
HC ppm @ 15mph: 55
HC ppm @ 25mph: 60
CO% @ 15mph: 0.22
CO% @ 25mph: 0.32
NO ppm @ 15mph: 720
NO ppm @ 25mph: 700

For comparison, here are California's limits for their own dyno test:

HC ppm @ 15mph: 108
HC ppm @ 25mph: 83
CO% @ 15mph: 0.71
CO% @ 25mph: 0.58
NO ppm @ 15mph: 767
NO ppm @ 25mph: 706

That's right. California's limits are *less aggressive* than Virginia's on a dyno-based emissions test at 15 & 25mph.

I have put a call into the Virginia Department of Environmental Quality (which oversees the emissions test program) to see if I can find a real human who might be able to offer insight into why the HC and CO% limits are so much tougher here than on the west coast.

At any rate it's moot as the car has now passed and is good for two years, which takes it beyond the 25-year window for future testing. But it's still a very odd set of circumstances I'd like to understand better.

Thanks to all who provided support and assistance during this saga!

fraggle

I'd suggest having a BG intake clean, it may help bring everything back to a better state. The guys oveer at Piedmont Tire have a lot of experience doing it on the 928 after Frenzy two years ago, I think they did 20+ cars!

fraggle

Of course, the clean won't help the cat, but it won't hurt.

CardinalCar

I'm planning an intake refresh anyway in order to address the TPS WOT switch you noted with the scanner. I'll take the time to refurbish the throttle body butterfly and bearings while I'm in there.

So if the runners and other bits are grungy they'll be thoroughly cleaned at the time. I suspect they're not too bad since Rob had the intake apart in March anyway, but you never know.

fraggle

The BG clean will also clean valves and the combustion chamber, which may have some buildup due to the rich running.

John Speake

Good to see a good result at last :-)

AO

I am happy to hear that you got it running again.

THis is very interesting. I have had to run open loop for a very long time because the oxygen sensor would always go pig rich with it enabled. I wonder...

I have another brain that is un-rebuilt and still unctions. Never thought about trying that one in place.

John, do you have any further info on this problem? I know of at least two other guys that have had similar issues.