[maddios]

Hey guys, for the past week or so I've been getting an intermittent red cog which shows up for about a second or 2 :eek and goes away on its own. I pulled the codes and got error 5506, which seems like it's a bad wiring issue to the clutch valve. Now the question is this, could it be the valve itself that's gone bad or likely a bad plug?

Seeing as how the valve is hard to get to without removing the gearbox I'd like to know if i need to order something or just get in there and rewire/clean it up.

Further, where is the valve itself located? is it on the side next to the clutch slave or someplace else? Any further advice is definitely appreciated.

PS: I've attached the output from inpa where it mentions that it's an open circuit fault. Interesting to note that it looks like my clutch valve draws too much current, max rated 800mA but mine draws 833 in that image. Odd.

 

[maddios]

Ok, nevermind on where or what the valve is. I just found it in Rheingold (Replacing clutch valve), seems pretty clear/simple (or as simple as can be with an smg).

What are the chances the valve itself is dead? Especially intermittently. I'm betting on a bad/cracked plug or broken wire.

Guess I'll drop the cover this weekend and try to fish out that connector and see if i spot anything obvious, hopefully it's something simple.

Otherwise I'll just have to replace all the whole wiring loom since I won't be able to trust it not to fail when i least expect it.

 

[maddios]

Looked through all the plugs/wiring today, whatever i could see, tested continuity, ohms and even ran Rheingold diagnostics on it. All checks out, grrrr.

Rheingold even has a useful test in it where the SMG ECU sends a 1A signal through the lines and measures it itself, so you can bump all the connectors/wires to see if any resistance increases, nope, mine read right around 1A, oscillating from 980 to 1050ma.

The only thing I found a little shady is that the connector on the hydraulic valve was very easy to remove, i'm used to having to apply a little force to pull them off, this one slid off on its own almost as soon as i squeezed the clip. So I wonder if moisture got in there.

Either way, for now I just cleaned it well with contact cleaner and sprayed some silicone dielectric in there. I'll re-run the clutch adaptation tomorrow and hopefully that'll be that.

Also replaced the input shaft sensor while I was there, since that thing was causing me some grief too.

 

[maddios]

No dice, it's actually worse now. which is good since the worse it gets the easier it'll be to debug.

It's logging that clutch valve error (5506) as an "active" error now and i can barely drive the car. Funny that the clutch works fine to start/stop in 1st but won't let me shift. And starts/stops are perfectly smooth, odd.

Going to do some more debugging this weekend, i might remove the valve and try to actuate it on a bench, make sure it doesn't have a dropout position. Also, clean the ecu contacts.

Otherwise I guess I'm at a loss, I wonder if I should try to get a used solenoid from ebay and see if it'll work, though mine is within spec on ohms (3.3)

 

[maddios]

So I solved the issue :jump:, the main connector on the side of the gearbox (one above the radiator) had an intermittent connection fault. All tests from Rheingold showed everything to be ok, but it still kept throwing the codes. Checking ohms through the system was OK, and it was even able to teach in the clutch curves and such but would throw errors during and after about the actual solenoid. I don't have a set of connector test leads so I couldn't check the ohms through the connector under load but I'm pretty sure that was the issue, at .5V the connector was fine, but at load there was too much resistance through it and especially with heat it'd just go way out.

I took all of the connections apart and sprayed them down well with a contact cleaner (not an electric parts cleaner) and then lubed with silicone spray and reassembled. Been driving for a week now no issues, before i couldn't even drive it off the ramps without doing a reset and it'd throw errors almost immediately.

But here's the kicker, before this error came up I used to have intermittent hard shifts, and had a really hard time adapting the clutch. It was very temperature sensitive, so if i adapted it when cold it was really rough when hot, and vice versa.

Now driving for 2-3 days it's silky smooth shifts all day long, love it. You can still feel the tap from the gears engaging but the clutch goes in like butter.

I'd recommend everyone put this cleaning in as a regular maintenance whenever you go in to change the gearbox oil to clean up these contacts too, they're pretty easy to get to while you're already there and the result is night and day. It's never driven this smooth since I bought over 2 years ago.

 

[maddios]

it's plug #8 in this picture:

It's directly above the gearbox oil radiator. Which you have to remove, but while you're changing the oil you have to remove it anyhow so might as well clean that plug.

You don't need to do any mechanical cleaning, just open it up let both sides hang downward and spray contact cleaner in there, let it sit for a couple minutes, spray it again then let it dry completely (about 5-10 minutes for the female side to fully dry) then spray a light coat of silicone spray lube in there, let excess drip for a few seconds and reassemble. done.

For me this was night a day, I was getting similar simptons to others before, slightly rough shifts, especially at low speed the clutch seemed to actuate slightly rough. There was some discussion on the forum about this and the fix at the time was to replace the clutch bearing and the guide. Seems like in my case it was all in the plug, and it makes sense actually if you think about how the valve is controlled.

The control signal is a PWM with a 5.6V amplitude, and by varying the wave they generate varying analog Voltages, but the problem is that it's at 1.5A, so any resistance in the plugs on the way will drop the voltage significantly and the test procedure doesn't really test the plugs correctly because they're tested unloaded and a slightly corroded plug will show almost 0 ohms but will read a significant voltage drop when loaded.

 

[maddios]

PS, I'm not recommending dielectric grease here for a reason, it's too thick and I've had issues with it actually lifting the pins off the surface or if you put too much it won't even let the plug slide flush, so just a liquid silicone spray is best.

 

[maddios]

lmao, i just realized that on the left side there's what looks like a chevy logo. wtf. hahaha

 

[maddios]

So looks like the error came back after about a month. Same exact symptoms as before.

Tried cleaning it well again, no dice.

At this point I'm kinda lost, running through diagnostics says ecu is ok and so is the valve. What else can it be?

Can't believe i'm the first one to ever come across this error, at least seems like that from searching the forums.

 

[maddios]

Was just examining the error codes i posted on page 1 and noticed something odd, not sure if it's the issue but looks like the solenoid valve is being energized but the clutch position is unchanged. I wonder if that's the issue that the ecu is seeing.

But if it is, why wouldn't it just say the plcd is bad?

In the following 2 exceprts you can see the mA increases but clutch position is unchanged.

1. Umweltsatz
solenoid valve clutch actual current Raw value 82.00 mA
actual clutch position 655.00 Ink
Duty-Cycle clutch 4.02 %
2. Umweltsatz
solenoid valve clutch actual current Raw value 833.00 mA
actual clutch position 655.00 Ink
Duty-Cycle clutch 39.12 %

Wonder if this could be caused by the throwout bearing binding up. But then doesn't the SMG ecu throw other codes when it jams up? strange

 

[maddios]

The actual problem was trying to solve with the 5506 was solved by replacing the SMG ECU. I bought a random one on ebay, popped it in, and ran all of the 4 calibrations in ISTA/D. Been running great for the past 2 weeks since I replaced it.

 

[maddios]

The software version is older than the one I had. All of my car's modules (other than the SMG now) are latest firmware versions, verified with ISTA/P.

I'll probably flash the latest euro code on the SMG ecu soon, been lazy to setup/test my new power supply. My old one died while stored, for no apparent reason.

Btw, old SMG ecu did cause some errors to be stored in the DSC and DME ecu shadow memory. But that may have been due to it being in fault mode which actually sometimes would throw the DME into limp mode.