Fix for A/C air distribution

[MEL1523@ACS.TAMU.EDU]

Lots of people said they have this problem, so I thought I'd post it to the 
whole list.

In my case, the symptoms were as follows:
     - In hot weather mostly, the a/c would be coming out the center vents 
like normal, but then switch to both the defrost and footwell outlets.  It 
seemed to do this worse when the weather was warm, and seemed to do it when 
I hit large bumps, but could do it at any time.  Putting it on low didn't 
help.

In the course of looking for the problem, I found that the vacuum servo for 
the fresh/recirc door had broken off its mounts and the spring that pulls it 
in the other direction when the servo is off had corroded and fallen off.  I 
wired it back onto its mounts and made sure the vacuum line was firmly on 
and not leaking.  This was not the problem, but you might want to check this 
because, you need the door on recirc in the summer for maximum cold air, at 
least here in central Texas.

The problem is in the a/c programmer.  For those of you who have not seen 
it, it is a little black box mounted up in front of the glove box on the 
right side.  It can be removed with only one nut.  There is a bundle of 
vacuum tubes coing out of it and one push/pull cable.  There is also the 
electrical conntecor.  

The problem is particularly hard to find because it is a bad solder joint on 
the board.  Remove the cardboard that covers up the circuit board and you 
will see 4 vacuum solenoids soldered onto the board.  In my case, the 
culprit was the center one of the three that are lined up in a row.  It 
works like this, 12 V comes in and goes through a diode.  This drops the 
voltage about 1 V.  The 11 V then goes to one pole of each of the 4 
solenoids. You can follow the solder river on the back of the board and see 
this.  The other pole of each solenoid goes over to an individual pin on 
this big chip with "DELCO" written on it.  I guess this chip grounds the 
lead, and so there is 1 V drop there too.  This leaves about 10 V (in my 
case) to activate the solenoids.  I first picked up on the problem when I 
got about 6 V on the solenoid that wasn't doing its job.  

The thing to remember is that electromagnets take more pull to engage, then 
to hold in.  So, if you get 6 V and the valve is on, that might not be 
enough to pull it in reliably, even if it is enough to hold it in.

If you check for continuity, you'll get a good signal, but when there is 
current flowing, I got 4 V across what was supposed to be a solder joint.  
There should obviously be 0 (or very little).  The bad one was the side that 
is the common power circuit to all the valves.

My fix was to run a jumper wire to all the other terminals that were on that 
common solder river.  I did this for all of them preemptively.  The jumper 
basically runs in parallel to the bad joint, and then I got 10 V reliably 
across all of them when they were on.  It has been working fine ever since, 
and I have been running it enough that it should have screwed up by now.

If your bad joint turns out to be the side that goes over to the chip, you 
could run a jumper right to the pin on the chip, or remelt the solder where 
the solenoid is mounted.  Don't solder to the chip unless you know how to 
solder properly!  If you heat up the pin on the chip too long and too hot, 
you could mess up the internal connections within the chip.

A good first try would be to simply remelt all of the connections at the 
solenoids and see if that helps.  You might even put a little extra solder 
on each one for good measure.

Good luck!

 - Mitch Loescher