Keyless entry in 90 S4 - problem
 

I'm having some issues with wiring a keyless entry into a 90 S4. I'm following Ed Scherer's wiring diagram, shown below. For the lock signal he uses pin 7 on the 14 pin connector to central locking unit in the center console near the radio.

When I click the keyless remote lock button it works, but the window motors activate for just a fraction of a second. I understand the locks on MY90+ have a function to raise the windows and close the sunroof if open by holding a door lock in the locked position. That's not what's happening with the remote. The window motors pulse and move the window a fraction of an inch when using the remote regardless of window position. The remote sends short pulses, so the window does this whether I hold the button down or just flick it.

How do I get this to operate correctly?

Also, is Ed's hatch release wiring optimal?

http://www.ed.scherer.name/Porsche_9...90%20928S4.gif

Bill-
If I read your post correctly, you want to enable the functionality whereby the windows go up. Right?

I think this is a funtion of the Keyless unit you're using. On mine I had to program it so that if I hold the button it sends a continuous signal to close the windows rather than a pulse. Check your owner's manual


Oh wait... (haven't used this one for a while...) :rtfm:

My keyless entry system could be configured for short or long pulses to ground. Can you check and see if you can change that setting?

Bill:

Mine does that slight window close thing, too. Based on what keyless entry unit you're using, you might be able to shorten the duration of the signal to lock. On mine, I think that how long the button is pushed on the remote is irrelevant; the Omega REC-8 always generates a fixed-duration lock pulse (of two selectable durations; I've got mine set to the short pulse, IIRC).

I forgot to mention this in my original post because it's really rare for me to leave my windows open, so it actually took me quite awhile to notice that it was even doing that.

My REC-8 manual(s) are at home and I'm at work, so I can't look right now and I can't find a REC-8 manual online.

But, there's a REC-11 manual online. For that model, I see that the lock pulse time is configurable at 0.8 seconds (default) or 3 seconds. That sounds familiar, and it's probably the same on the REC-8.

I'm guessing that something like 0.5 seconds (or maybe even a little less) would be more ideal.

It probably wouldn't be that hard to add another pulse circuit that would shorten the pulse; I'm not sure it would be worth it (at least for me).

Ed & Andrew. This is a Commando unit, and it can be programmed for short or long pulses. I'll try the long pulse as default.

The only reason I ask about this is that I do NOT see this happen when I use the key in the lock. The windows don't move unless I hold the key. So, the remote short pulse is doing something the key does not do. Puzzling.

Naw... mine does it too. Don't be so paranoid.

YOUR CAR WILL BLOW UP! YOUR CAR WILL BLOW UP! YOUR CAR WILL BLOW UP! YOUR CAR WILL BLOW UP! YOUR CAR WILL BLOW UP!

Bill - its a grouding issue - clean the console ground wherever that it - or supply the KE from a different ground.

The hatch release is quite independant from the rest - but it won't actually even work as shown unless the rear motor is rewired to a permanent ground (at least not without opening a door first...).

I would do it differently but rewiring the motor ground is an OK way to go...IMO

Oh and Andrew - I have this bad feeling your cars going to blow up or something - be careful!... don't know what is is that makes me feel this way... must be subliminal or something...

Alan

Alan:

I will move the ground!

I do want to include the rear hatch realease function. I'll follow one of these wiring diagrams just for that section, all courtesy of you! The last one includes an interlock I requested.

My GTS does the same if the window is open - just moves a tiny bit when the button is pressed.
Never been a big deal, however if you find the fix I'm in.
Roger

The ground made no difference. Puzzling. What is the remote lock doing that the key doesn't do? It is a very minor thing. I just wasn't expecting it. Here's the relevant 90S4 wiring diagrams if anyone wants to have a shot at it. The wire into which I tapped is one of the wires on pin 7, indicated in red. The second image shows where those wires lead.

http://norcal928.org/90S4centrallocks01.JPG

http://norcal928.org/90S4centrallocks02.JPG

I don't think it does anything differently. It's the same as moving the key in the lock to lock it for exactly 0.8 seconds (same as the Omega REC-8 pulse width, or 0.9 seconds for your Commando, I think). I can reproduce that behavior using my key; I just went out to the garage and tried it just to make sure.

I think it's the first time that lock cylinder has seen a key in at least a year. :)

We just need to get that pulse width shortened to a half second or so and that should do it, I'd think.

Yeah, you're right. I forgot about that modification, which I made at connector T9 (below the hatch release motor): move the brown wire from pin 1 to pin 2 on the connector, as shown below.

http://www.ed.scherer.name/Porsche_9...90%20928S4.gif

This does compromise safety, as I believe you (and/or others) have pointed out before, as it allows the hatch to be opened while underway. I'll confess that this was an expedient solution, not the best solution.

(BTW, the above fragment of the wiring diagrams is at location M10/N10 on sheet 2 of the 1990 wiring diagrams).

I think whats happening here is that the low pulse on the lock lines is getting coupled to the switch inputs to the window controller. It seemed likely this was via ground noise - but maybe its just direct capacitive coupling. The switch inputs to the window controller just float high normally (internal resistor pull-up) when a switch is presssed these are pulled low (one wire for up, one wire for down).

If these see a capacitively coupled dip on a line they will operate the motor as if you very very briefly tapped the switch. I don't think the pulse length will make any difference because I think its only leading edge transition related (not pulse duration related since the pulses should minimally be long enough to see more window movement that you note...

Solutions path might include adding a capacitor to +12 power on the switch inputs (probably just the down end). A few ufarads should make a difference... or you could try to slow down the output transition (to couple less noise) via a series resistor on the lock outputs. I like the capacitor idea a little better - its more predictable.

Initially you could just try this at the plug connector for the windows in the console - pretty easy access...

Alan

Alan, don't you think that it's more likely that this is just what I mentioned in #12?

I can reproduce this behavior with the key, although since I'm just a human and I wasn't using any test equipment, I'm not sure about exact timing.

When using the key, there's a slight delay before it does the "window/sunroof close" routine. I'm guessing that the delay is shorter than the 0.8 or 0.9 seconds pulse width that these keyless entry systems provide.