getting CEL for 3 solenoids. help

[rallysam]

Sometimes, I get these codes:
24, 35, 44 (Idle air control solenoid valve, canister purge solenoid valve, wastegate solenoid valve)

But here's the weird thing:
75% of the time, I get no codes and the car runs great (idles and boosts fine, suggesting those solenoids are ok).

But when I do get the codes (25% of the time), I get all three of those codes together. I NEVER get just one or two of the codes - it's all or nothing. All that suggests that the solenoids are fine and instead we are dealing with something systematic.

I replaced the ECU, and all the same codes come up. That suggests that the ECU is not the problem.

Resistance measured AT the solenoid:
Wastegate solenoid: 20.7 Ohms

Resistance measured inside the passenger compartment on the harness connector that normally connects to the ECU
Wastegate solenoid: 21.4 Ohms
IAC valve (close end): 10.2 Ohms
IAC valve (open end): 10.2 Ohms
Canister purge control: 43.3 Ohms

(to measure the resistance of each of those, I put one probe on each of those pins and the reference probe on pin 5 of the Select Monitor connector becuase that was the closest place I could find where the "common" was broken out)

Any ideas what could be wrong? It seems like the solenoids check out, the harness checks out, and the ECU was eliminated because I swapped and still had the same problem. I'm pretty desperate to get rid of this!!!!

[rallysam]

Bump.

Any ideas?
Do those resistence measurements look correct?

[vrg3]

The resistances do seem all more or less correct.

The manual says to start by fixing the problems with the lowest trouble code, clearing codes after every repair. The implication is that it's possible for problems that trigger lower-numbered codes to also trigger higher-numbered codes.

So maybe we should start by figuring out why you're getting code 24.

When the codes are active, what are your symptoms?

Have you tried seeing if jiggling the big connector on the ECU can make the problem go away?

[rallysam]

Thanks for the response, VRG! I was hoping you might come to my rescue. BTW - just hooked up your scan tool last night and it works great.

The manual says to start by fixing the problems with the lowest trouble code

The lowest trouble code is 24 - the idle air control. I guess I will start there.

When the codes are active, what are your symptoms?

Well when the codes are active, it drives in limp mode. Limp mode is bizarre behavior so that makes it hard for me to tell what is a symptom and what is a "feature" of limp mode. Pushing the throttle down more than maybe halfway causes fuel cut. It idles high too. Otherwise it gets me around fine even in limp mode and seems to be working ok.

Have you tried seeing if jiggling the big connector on the ECU can make the problem go away?

Not exactly. I did replace the ECU entirely and the first time I started up the car the new ECU threw the same three codes... but that doesn't exactly answer your question. I will try that the next time I get the codes. Coincidentally, all 3 of the "problem areas" do come into the ECU on the same connector, and they share a common branch of the power circuit.

VRG - Do you know what exactly does the monitoring circuit look for exactly from the IAC valve in terms of electrical properties? I'm an electrical engineer, so don't be afraid to hit me with the geek speak.

[vrg3]

Thanks for the response, VRG! I was hoping you might come to my rescue. BTW - just hooked up your scan tool last night and it works great.

No problem. Glad it's working for you.

Well when the codes are active, it drives in limp mode. Limp mode is bizarre behavior so that makes it hard for me to tell what is a symptom and what is a "feature" of limp mode.

There isn't exactly a "limp mode" per se. There are just failsafe actions the ECU takes when certain codes are active.

Pushing the throttle down more than maybe halfway causes fuel cut.

This is the failsafe action the ECU takes when it's throwing a trouble code for the wastegate control solenoid.

It idles high too.

Ah ha... This is what I was wondering about. This may give us a clue. As far as I can tell, even with the IAC trouble code being thrown, the ECU still tries to drive the valve.

The IAC valve has two solenoids in it. The valve's normal resting position is somewhere around halfway closed when it's warm (when it's cold it's open further). The "open" solenoid pulls the valve all the way open, and the "close" solenoid pulls it all the way closed.

If somehow there's a problem with the "close" side circuit but not the "open" side, you'd get a high idle. Maybe that's what's happening in your case. Maybe. It is what happened to Phil.

Otherwise it gets me around fine even in limp mode and seems to be working ok.

Yeah, that sounds reasonable. There isn't any failsafe action associated with the CPC solenoid trouble code.

Coincidentally, all 3 of the "problem areas" do come into the ECU on the same connector, and they share a common branch of the power circuit.

Heh. "Coincidentally."

But so does the pressure exchange solenoid.

And the IAC and CPC valves go through the SMJ, while the boost control and pressure exchange valves don't.

But, yeah, next time the problem happens, try popping the IAC valve's connector off and checking that the middle pin has +12v on it.

Do you know what exactly does the monitoring circuit look for exactly from the IAC valve in terms of electrical properties?

I don't know exactly. I do know that you can trick it by replacing each solenoid with a resistor (wired between the ECU pin and ignition-switched +12v). This resistor's value can actually be a good deal higher than the solenoid's resistance. Not too too much higher, though.

If I had to guess, I'd say it pulls a very small bias current through the solenoid (by grounding it through a high-value resistor) and looks for a voltage near -- but not quite at -- system voltage on the line.

[rallysam]

Ah ha... This is what I was wondering about. This may give us a clue. As far as I can tell, even with the IAC trouble code being thrown, the ECU still tries to drive the valve.

Went back and checked - Yeah it idles around 1k when the codes are throwing.

Tried jiggling the connectors on the ECU - didn't help.

But, yeah, next time the problem happens, try popping the IAC valve's connector off and checking that the middle pin has +12v on it.

OK, that'll be next.

If I had to guess, I'd say it pulls a very small bias current through the solenoid (by grounding it through a high-value resistor) and looks for a voltage near -- but not quite at -- system voltage on the line.

Hmm... it would be interesting to throw a pot on there and just dial around on it until the ECU wigs out (to determine what the "alarm thresholds" are).

[vrg3]

Also take note of how the idle changes when you unplug that connector.

Yeah, it'd be interesting to know what it's looking for, but probably not all that useful... Maybe after you figure it out I'll be proven wrong though.

[rallysam]

Hasn't been much progress recently because it hasn't thrown the codes in the past week.


In other news of my electrical woes, my car had trouble starting this morning. I noticed that a connector under the rear seat was getting hot as crap (pretty sure it's the wires leading to the fuel pump). I noticed that as a jiggled the connector, the fuel pump changed pitch. Bad sign! (Note that this is not the connector on the fuel pump itself, just some intermediate harness connector en-route to the fuel pump).

I guess I have to cut out that connector and rebuild the connection with butt connectors or something...

[vrg3]

Yup, that fuel pump connector does sometimes go bad, and the first indication is the heat from the high-resistance joint.