RETed

My headlight switch has been giving me problems, so off to eBay to go fishing...

I ended up with a complete bezel + switches purchase off a 1986 model.
I do own an early 1987 Turbo II.

First, before I fired anything up, the turn signal switch plug is upside down.
I thought I was losing it when I tried to plug it in, but upon inspection, the two turn signal switches had plugs that inserted upwide down.
Weird.

I should've known it was just the start...
Once I had plugged everything in, the headlights were constantly on.
Retractor switch works okay.
Interior illumination dimmer does not work - everything stayed "dark" no matter how I adjusted this.
Okay, sounds like the headlight switch cluster is fudged...
Now, the best part...when I turn the motor off, the engine stays running.
WTF!
It eventually died like the fuel pump was cut-off, but the ignition was still firing.
Upon removal of the key, the car acted like it was still on.
When I unplugged the headlight cluster switch, then everything was killed.
So it looks like the ignition switch is shorted through the headlight cluster switch...ODD.
I haven't done a FSM wiring scan, but this just sounded really screwed.

Looks like I'll need to tear the switch apart, but I'd figures someone would find this really interesting.


-Ted

Icemark

The 89-91 use a different Turn signal switch than the 86-88, and of course the S5 light switch harness is also different. Internally the S4 and S5 light switches and turn signal switch are radically different.

But the pin out on the Light/defrost/headlight cleaner switch is the same regardless of series.

Sounds like you got a S5 light switch harness and are using it on a S4 turn signal switch.

RETed

I used my '87 harnesses.
All I did was swap out the bezel and switches.
So it can't be the harness...

I'll dig into it deeper later this week when I got the time.

Thanx Icemark!


-Ted

AbortRetryFail

Only slightly off topic:
My 86's headlight switch keeled over again last night. Re-soldering all the connections and cleaning the switch a few months ago only delayed the inevitable - so i ordered one off ebay and tore mine apart to rig up a temporary switch.

Upon inspection i found out what failed isn't actually the switch contacts but the circuit trace on the board. The trace heats up, melts the solder on the connector, burns the connector plastic fries harnesses etc.
The trace fried entirely on mine so i tried something potentially stupid and put a jumper wire straight to the switch and it worked again. It doesn't even get warm except from the illumination bulbs.

I know i'll probably get flamed for this idea or for the soldering job here but this switch has been soldered and de-soldered at least 5 times now.

Who knows, maybe this will help someone else get more time out of their switch until they can find a good working replacement.

Icemark

Actually this has been covered quite a few times.

The circuit board connection is just one spot.

Re-soldering it after it is burnt might temporarily solve the issue, but is nothing more than a band aid.

See the real problem is corrosion build up in the bulb sockets, which leads to increased resistance, which leads to overheating of the weakest connection. So the switch burning up there, really is only a symptom of a much bigger issue.

A external relay at the light switch on the parking light circuit will in most cases solve the problem of the light switch failing, but will not solve the issue of increased load due to corrosion at the sockets. If you clean the bulbs and sockets, use a good quality dia-electric or conductive grease to prevent further corrosion the problem almost never appears.

And due to the way the metal is damaged when it over heats, it is never a good idea to just re-solder a failed headlight switch. The re-soldered connection on a failed switch will never be as good or as strong as it should be. Also the changed metal conductivity of the switch and the plug will now result in the switch failing again and again, because that is the weakest spot.

So if you have a switch that burns, you need to replace the switch and the light switch harness, and I also recommend using a relay then on the parking light circuit (covered in the 2nd gen archive) just to pull the load off the switch all together.

AbortRetryFail

It sounds more lo me like Mazda used a fuse that could handle more current than the wires and switches it's supposed to protect...

Icemark

No, not at all.

The fuse is fine for the wires and the circuit. The connection over heats. It does not draw too much amperage.

The problem is when the owners don't maintain the bulb sockets. Hell most owners never even check the bulbs unless they have one burnt out.

Sideways7

iTrader: (7)

Di-electric grease FTW!
Seriously, they sell little individual packets of the stuff specifically for this at the checkout of most auto parts stores.

RETed

Icemark was right - the replacement was from a 1991 model.
I got them all mixed up. :P

When I did open up my 1987 headlight switch, I was greeted with LOTS of broken cold solder joints.
No signs of burn traces or any other "major" damage, so this was looking good.
Once I had redone all the solder joints, the cluster switch worked like a charm.
My headlight cluster switch is original, but it looked a few of the points were resoldered prior.
I guess you can't blame 20 year old electronics for failing like that.
The repeated twisting of the headlight switch (more so that flipping the wipers on) easily caused all the solder joints to fail.


-Ted

NZConvertible

Nice to see someone blaming time instead of Mazda for a change...

BLK FC3S

Damn mazda for not stopping time

RETed

I got to taking some pics and doing a small write-up on this...

http://fc3spro.com/TECH/PROBLEMS/EL/HL/hswitch.html

Enjoy!


-Ted

overkill-lt1

not to get off topic i have two working no molested headlight switches for sale if anybody needs one.

45.00 shipped anywhere in the usa! each