DSMchip and clutch interface switch...updated

[belize1334]

The plastic tab on my clutch interface switch broke leaving me with no clutch input to the ECU so my car wouldn't start. I unplugged the switch thinking that it was a good idea anyway so that the crank wouldn't be loaded during cold-start. But then a curious thing happened.

With the switch unplugged the launch rpm became the driving rpm (couldn't rev above 4500 rpm at any speed) and the vehicle speed sensor stopped reading a value (my turbo timer reports vehicle speed from the ECU).

I was under the impression that with a 2-stage rev limit the clutch had no involvement. I tested it the other way and found that with the clutch switch plugged in but not depressed (clutch UP / car in neutral) I have no launch limit under any circumstances.

Can anyone else comment on their experiences in similar situations?

The chip is programmed for the selectable launch rpm but NOT the flat-shift mod.

Also, is there any reason that I couldn't just super-glue a dime over the hole in the tab for the clutch interface switch? Looking at the plastic piece that broke out of there I seriously doubt that I'm gonna be able to source another (junkyard) without it breaking when I remove it.

 

[Rausch]

The launh control, and even the NLTS should only be reading from the lower switch. The difference being whether or not the ECU sees that the vehicle is moving....

I remember this comiong up before, and I believe that it has somehting to do with teh wiring and the switch. If you are only seeing the launch RPM all the time, the ECU is not seeing the VSS reading for whatever reason, IIRC....I know Jeff (Keydiver) has precisely answered this exact question before....MAybe search a bit and see if you can find a similar post? They are on here somewhere...

 

[Rausch]

Quoting Keydiver:

If you don't care about having to push the clutch in to start the car you don't need a diode. You can just cut the wire shown in TMO's FAQ and connect a wire from the end still attached to the clutchswitch to pin #104 of the ECU. Remember to also cut the black ground wire loose from pin #104, or you will get the MAIN rev limiter all the time. Also, on a GVR-4 the wire is black with a red stripe, not green like a DSM.



Not exactly what you were looking for, but may help figure out what you need to do.

 

[belize1334]

Ok, so it's something to do with the switch interfering with the vss...which is consistent with what I was finding.

I think the simplest solution is to fix the switch. I'm just gonna glue a dime or something over the tab so that it depresses the button. Later I can wire in a bypass so that I can start it without the clutch but not unplug the switch...maybe wire it in parallel to the ebrake switch or something like that.

 

[belize1334]

ok...so the last known situation was that the plastic tab had broken which was keeping the switch from being depressed. I had two possibly configurations.

*I could leave it plugged in in which case everything operated normally but I had no launch control and I had to press the button in by hand to start the car.

*I could unplug the button in which case the car started fine w/ or w/o the clutch in but the vss didn't work and I had permanent launch control.

So, I came home and found a small (m6?) bolt which I put through the clutch arm in place of the plastic tab and tightened down with a small nut on the back. This made it so that the switch was operational and life could resume as normal...BUT WAIT.

Now the issue is that the button works normally (car won't start without it) but the vss DOES NOT WORK. The turbo timer reads 0km/hr all the time (it's always worked flawlessly till now), the idle is funky (info from the vss is used in the ECUs idle control algorhythms), and I have permanent launch control (not some other rev limiter because I can select the rpm during the normal startup routine).

So the question is this...how do I diagnose a problem with the vss? I know it has something to do with the button since all this started with the button. But the button works so how can this be? I'm WAY out of my depth on this. Where does the vss origninate/terminate? Maybe I can bypass everything else and see if that works.

 

[toybreaker]

Teh vss signal is used in a few different places.

It originates from a reed switch mounted to the back of the speedo.

The best way I've found to troubleshoot it is to check to see of the cruise works, or if the wipers cycle rate changes with an increase in road speed. (on the intermittent setting)

The engine idle rpm will also be ~slightly higher when coasting with the clutch in, and when you come to a complete stop...it'll slowly come down to the base idle.

The best test I've come up with is a test by substitution. Just swap in a known good instrument cluster, and see if the symptoms clear up.

edit Do you have access to a boneyard out there?

I *think* the vss is common to all Galants with cruise, and you could probably hoark it out of a ls instrument cluster, and get it to work in your vr4 speedo. /ubbthreads/images/graemlins/dunno.gif

I hate it when multiple gremlins pop up like that. /ubbthreads/images/graemlins/banghead.gif

Makes you wonder where to start troubleshooting!

 

The stutterbox code looks at the clutchswitch FIRST. Either glue the plastic plug back together, or get a similar rubber plug at Pep Boys to put in the hole. In the meantime, you would have to disconnect the clutchwire, and ground the wire going to the ECU to get your MAIN rev limiter back.

 

[belize1334]

The frustrating thing is that the clutch button seems to be working fine at this point. Granted the piece I have in there is a bolt instead of the original plastic tab, but it does the job perfectly. I press the clutch in and the car starts, no clutch no start. Could I have shorted something out by unplugging it?

Does the signal from the vss actually travel THROUGH the clutch switch or is it just an issue of the ECU looking at one before the other? Maybe the clutch switch is "partially" broken causing the ECU not to reference the vss? /ubbthreads/images/graemlins/dunno.gif

 

[belize1334]

I have determined that both clutch pedals are in perfect operating condition. I grounded the wire going from the switch to the ECU and all that accomplished was making it so the car wouldn't start. I started it and THEN grounded that wire and the vss still didn't work and I was still in launch control. I'm slowly coming to the conclusion that, unlikely as it seems, this is not a clutch interface related issue and that something entirely unrelated is going on. I'm going to disconnect the turbo-timer's vss line and see if maybe that's the problem (short /ubbthreads/images/graemlins/dunno.gif) and if that doesn't do it I'm going to the junk-yard to look for a new reed switch.

Could someone with a digital copy of the fsm post up an image of the wiring diagram for the vss?

 

The VSS signal has NOTHING to do with the clutchswitch, they never even go near each other. You have two SEPARATE issues. If in fact your chip does not have NLTS, and is only a 2-step, you shouldn't even have a clutchwire, as it won't make any difference.
It sounds to me like you need to replace the reed switch in the speedometer. I'd send you one, by I have been unable to find my supply since quitting my job and moving all my stuff home. /ubbthreads/images/graemlins/frown.gif

 

[Hertz]

What is a symptom of the reed switch going bad? I noticed on the logger my VSS flips between 0 and 61 or 63 at idle.

 

[belize1334]

Quoting keydiver:
The VSS signal has NOTHING to do with the clutchswitch, they never even go near each other. You have two SEPARATE issues. If in fact your chip does not have NLTS, and is only a 2-step, you shouldn't even have a clutchwire, as it won't make any difference.
It sounds to me like you need to replace the reed switch in the speedometer. I'd send you one, by I have been unable to find my supply since quitting my job and moving all my stuff home. /ubbthreads/images/graemlins/frown.gif



ok...then I was just imagining a connection based on a horrible coincidence. When I unplugged the clutch switch the first time so that I could start the car I then noticed the next day that I couldn't rev above the launch rpm. I plugged the switch back in and the problem went away so I assumed that the issues were related. Could it be that I pulled a wire loose while monkeying around under the dash? Is the vss wire easily accessable from the clutch area that I might have accidentally snagged it?

 

Quoting Hertz:
What is a symptom of the reed switch going bad? I noticed on the logger my VSS flips between 0 and 61 or 63 at idle.



I don't go by VSS, so I wouldn't know. I always log address C6 separately.

I don't *thing* the reed switch wire goes near the clutch, but then again it does connect to the ALDL connector, the ETACS, and the cruise computer, which are all down in that area. I've never heard of any of those causing this problem, but I guess anything is possible. Someone would have to pull out the manuals and see what route the wire takes after it leaves the speedometer cluster.

 

[belize1334]

Update:

I pulled the dash and swapped a new reed switch in from a 1G eclipse. It didn't help. I decided that it was possible that both switches were bad so I manually rotated the speedo input that is usually driven off of the speedo cable and watched the impedance across the terminals of the reed switch. As expected it changes between OL and 0.01 Ohms so the switch is fine. I then check the impedance of all the wires between the reed switch and ground as well as the reed switch and the signal to the ECU (witch I tracked to be the big center pin in the driver's side gauge plug. Again, perfect connectivity. I then check the impedance between the corresponding female plug and the VSS wire at the ECU and again, perfect connectivity.

So at this point I've determined that the reed switch works and that there are no shorts between the ECU, the switch, and GND. So, why would the ECU not be reeding the VSS?

In any event I'm stuck for the time being because it'll be a week or more before I have my logger running again so I can't actually log any of the addresses in the ECU. All I have to go on is that the idle is funny when pulling to a stop, the wipers don't auto adjust their speed, the turbo timer says 0km/hr at all times and I'm permanently in launch control mode.

 

Have you tried watching the voltage on the reed switch pin of the ECU with the key turned on, as you rotate the wheel in the speedo? It sounds like you may have a conflict, where something isn't allowing the voltage to swing up or down enough to register to the ECU. /ubbthreads/images/graemlins/dunno.gif

 

[belize1334]

That's a good idea. I'll have to have the instrument cluster in place to do that but I think I can pull the speedo cable at the transmission and turn it by hand. I'll try that and see what I can see.

 

[belize1334]

I pulled the C-61 plug from the ECU and watched the continuity between pin-18 and ground while turning the speedo cable by hand. The multimeter always read closed loop which I thought was odd since I head checked continuity in every sub-section of the circuit and determined the vss to be functional. So I pulled the instrument cluster again so that the vss wire is unplugged ON BOTH ENDS (ECU plug and cluster plug are both disconnected) and it still reads

 

Thats' what I was saying: that VSS wire gets split to several different computers and the ALDL connector, so a short at any one of those locations can kill the speed sensor. If you don't care about the cruise and variable speed wipers, you might just want to try cutting everything off and just running a direct wire from the reed switch to the ECU.
Either that, or start thumbing through the manual to see which connectors all get VSS signal, and start unplugging them one by one.

 

[belize1334]

I don't have cruise and I couldn't care less about variable speed wipers so I'm perfectly willing to start pulling those connections. The problem is that I don't have a wiring diagram for the vehicle. I used to have a Haynes manual but I don't remember it having good wiring. I'm hoping that I can find a digital copy of the full FSM or something like that with all the wiring and then start tracing stuff back.

Anyway, seems like diagnosis...short to ground somewhere...now I just have to find where.

Thanks for your input guys.

 

[FlyingEagle]

www.lilevo.com
www.lilevo.com
Everything you need is there, but there is a thread in the archives under my screen name and you can get more stuff than the manuals if needed.