High idle on an LS3

Hopefully I'll find a GM guru in here.

We've got a 2009 Miata with a crate LS3 and controller from GMPP. The intake is an unmodified LS7 part. It's got a map that came out of a car that was tuned to use this MAF diameter. We've run this setup in cars before without a problem.

But the car's got an idle/closed throttle problem. Basically, it's acting just like a car with a sticky throttle. You never go into closed throttle on decel, and the idle usually sits about 1400 rpm. Once in a while, it'll drop down to 800 or so, but there's no rhyme or reason to it and there's always a hang.

It looks to be coming straight from the PCM. We can see the throttle plate position sitting at 22%. When it drops to 18%, all is right with the world. But I can't figure out why the PCM wants the higher number.

Any idea what might be messing with the PCM? The car's got a couple of thousand miles on it. There are no air leaks that I can find in the intake system.

I have zero experience with LS engines so disregard if it sound stupid. But this sounds like a malfunctioning TPS. Did you visually match the throttle plate position with what the PCM is telling you?

Drive by wire right? Have you tried a different pedal?

Drive by wire, yes. It's the pedal that came with the GMPP controller kit. I don't have a spare, but I believe the signals coming off the pedal are appropriate. I'll plug in to confirm. It is acting as if it never sees 0% on the pedal.

The PCM is driving the throttle plate. GMPP diagnosed it as a bad throttle body, we swapped in a new one and no change.

Well, lookee here. You ask a few more questions...

Bill told me the pedal signal was good. But when I asked him again, he admitted that the reader wasn't giving him good info. I just plugged in my own OBD-II reader and got a minimum TP reading of 20%. Seems awfully high for a closed throttle. I'll compare it to some of the other drive by wire cars at the shop tomorrow and see.

Good luck. Sometimes even new parts are bad

Yes, like the clutch on the AC compressor. Guess what we're changing tomorrow?

Keith Tanner wrote: Well, lookee here. You ask a few more questions... Bill told me the pedal signal was good. But when I asked him again, he admitted that the reader wasn't giving him good info. I just plugged in my own OBD-II reader and got a minimum TP reading of 20%. Seems awfully high for a closed throttle. I'll compare it to some of the other drive by wire cars at the shop tomorrow and see.

My (Dodge) fly by wire reads 10-14% throttle at idle, depending on temp and load. Reading up on obd2 scanners, this is norm, but 20 seems high. It's not an actual 10% opening, it's simply how the obd2 readers pick up fly by wire sensors.

Both my scan gauge and torque read false tps readings. The factory display (hidden menu) reads spot on.

I would imagine a gm scanner will pick up an out of spec sender, or if something else is requesting the high idle due to suspected load.

The future is scary.

Maybe an out of spec temp sender? The engine's under higher load when cold plus the PCM bumps the idle up cold as it tries to clear out the cold start fuel quick, so if it thinks the engine's cold but the temp reading is within parameters it won't set a code... yeah the future is scary. [Luddite] Gimme a cable throttle any day. [/Luddite]

Keith, why don't you ever call me. you know i'm your go-to lsx guy. i personally run kicking and screaming away from drive by wire setups. but the guys at the shop got it down. give me a call on monday

Johnny@Mtiracing

Monday's too late, it'll be in the hands of a magazine by then. And I don't think I've ever talked to you about LSx engines, Johnny. I have to fix this today. Luckily, I've got a couple of good running cars to compare readings to.

ECT looks as expected, so it's not a temp sensor.

DBW is not scary, it's just a different kind of technology. We can talk about the sticky mechanical throttles on my Locost later if you'd like.

Even if it is DBW it still has to have some kind of sensor that tells the PCM the throttle position. That sensor may be faulty.

We used to have a lot of the "high idle" problems on Zetec engine Focuses. Some times it was a faulty TPS but if some one was trying to modify the setup they usually had the wrong TPS (auto vs. manual) or screwed it on in the wrong position when installing an aftermarket TB. An aftermarket TB for use on all the engines had 3 sets of holes to fasten the TPS to and you had to use the right one (auto vs manual and different years) for the car.

More info: I checked against another LS3. TP% at idle was identical, but idle was considerably slower. Interestingly, this problem car will occasionally decide to idle properly, and you see the TP% drop, bringing down the rpm. This points to something the PCM is deciding...but why?

The throttle body (which incorporates the throttle position sensor) was replaced with another new one. No change in behavior.

I smoked the engine again. No apparent leaks.

Are you sure the pedal isn't at fault?

Can you load HP Tuners or EFI Live on the ECU and command a lower idle?

Can you command a lower idle with a Tech II?

If you can't command a lower idle even maxing out the idle control, probably a vacuum leak post throttle body. If the commanded idle works but its still wonky, look at the throttle body and sensors, also make sure the car is going all the way into closed loop

*This is based on a production based PCM. I have no idea how the GMPP PCM is configured, if its an off-the-shelf tune or if its specific, not OBDII, etc

Throttle body or pedal stop screw set wrong (I assume it has such an animal)? Other than that I got nothin'.

The idle speed is set for 750, I think. I might go plug in and dink around with HPTuners and see if I can find anything, but we haven't been able to yet.

I've looked hard for a vacuum leak. No luck. The fact that occasionally the PCM decides to drop the TP% to 18% and we get an 850 rpm idle would back that up as well.

These cars make some sort of click underhood when you move the pedal off idle position. I tried three of them, and the problem car did the same thing.

Out in the car right now. Accel position 1 is sitting at 18%, regardless of whether we're in high idle or low idle. Accel position 2 is at 10%, same deal. Car is in closed loop.

Unfortunately, I'm having trouble pulling a map off this thing so I can't check the settings, only view the PIDs with VCM Scanner. Argh.

Call Steve

Did y'all get it resolved?

Nope, not yet. I've spent quite a bit of time with datalogs. Interestingly, I also loaded the original GM-supplied tune back in the car with no change. Which points back to something physical, not tuning. Damned if I can figure it out yet, but I will.

Damaged wiring to the DBW pedal maybe? You could do a continuity test with each end of the cable, then check each pin with the cable unplugged at each end to see if any wires are grounded.

unk577 wrote: Call Steve

Old man knows alot

Don't they provide you with a flow chart/logic diagram? Something to show you the things that can influence the idle...

RossD wrote: Don't they provide you with a flow chart/logic diagram? Something to show you the things that can influence the idle...

On the same vein, have you measured the vacuum level with a vacuum gauge and compared to a good running car? I'm wondering if the MAP is wrong.