| Gently pry the cover from around the
vent openings. There are no fasteners, just tabs / prongs that hold it in place.
|
 |
| Insert the 5 mm Allen wrench into the hex
bolt behind the vent.
|
 |
| Unscrew the bolt until it disengages from the
air flap. It is a left-hand thread.
|
 |
| Carefully pull the vent out. There are two
tabs on the vent that clip into the top of the console. You may need to
disengage them using a small flat blade screwdriver. Be careful not to
bend them too much, as they can break if old.
|
 |
| Let the vent rest on the drive shaft tunnel.
Inside you will see the air volume control flap, and the
left and right heater cores. There is a separating wall between the left
and right heater cores. There is a large common evaporator directly
beneath the heater cores which is not visible.
|
 |
| Loosen the two small Philips screws holding
the rear AC module control to the console.
BE CAREFUL to NOT let the screws drop down into the
console. You may never find them. I suggest you stuff a rag into the
console to catch them if they fall.
|
 |
| Pull the module out of the console.
|
 |
| Inside you will see the two heater core temp
sensors.
|
 |
| Carefully push up on the end of the temp
sensor. Give it a little jiggle to help it out of it's socket.
It will come straight up and out.
|
 |
| The complete temp sensor.
|
 |
| Carefully pry the locking tabs of the
connector away from the temp sensor, and pull the connector off the
sensor.
|
 |
| The sensor is a thermistor assembly. These
are resistors that change value depending on ambient temperature.
The temp sensor should provide the following resistances
at these temps:
Degrees C
k Ohms
10
19.0 - 21.2
20
11.9 - 13.2
30
7.7 - 8.4
45
4.2 - 4.6
There was nothing wrong with the left temp sensor. To
double-check, I pulled the known-to-be-good right temp sensor out and
took a resistance reading from it. Both temp sensors were very similar in
values.
This was not the best news, as it implied that there
could really be a short or open in the left rear duovalve.
|
 |
| Reverse the disassembly steps to put it all
back together, but stop just before you're ready to put the vent back in.
Straighten a large paperclip and place it through the
hole in the plastic piece as shown.
Reassemble the vent to the console. The paperclip must
pass through the bottom of the air vent. The paperclip aligns the female
thread so that the 5 mm screw can be installed. Without this, it's just
about impossible to get the 5 mm screw started in the female thread.
Make sure all wires are tucked up inside the console and
out of the way. If you leave the wires as they are shown in this photo,
they will interfere with the vent.
Make sure you do not cross-thread the 5 mm screw into
the plastic. The plastic will strip easily if the screw is not lined up
correctly.
|
 |
| You will need to adjust the depth of the 5 mm
screw to ensure the vent door opens and closes properly.
Now that this is reassembled, it's time to dig out the
rear duovalve assembly.
|
 |
| Keep this in mind.
I thought I needed to remove the fender well to get to the rear duovalve
assembly. I'm not sure this is necessary. Try removing the panel directly
behind the front wheel under the car, and the trim panel behind the front
wheel on the side of the car. You may be able to reach the duovalves
without having to remove the fender well.
The photos show the fender well removed.
The panels are held on with either 8 mm or 10 mm hex
head bolts.
|
 |
| The rear duovalve assembly is tucked behind
the middle plastic panel.
I had a heck of a time getting in there even to
disconnect the electrical connectors. I couldn't do it. I saw that access
would be very simple if I removed the middle trim panel from the fender.
|
 |
| Remove the front 5 fasteners holding the
lower body trim to the car.
Open the passenger door, and remove the two bolts
securing the middle plastic panel to the front fender.
There are three pop-in-place fasteners at the top of the
middle trim. Carefully pop the top of the middle trim out of the fender.
There is a hook feature on the trim that grabs the
fender near the wheel opening. Gently persuade that free, and the panel
should come out.
|
 |
| While holding the lower body trim down out of the way,
carefully pull the middle panel out of the slots in the lower trim.
|
 |
| Now the duovalve assembly is easily reached.
|
 |
| The left rear duovalve is circled.
|
 |
| Lots of corrosion apparent. I was hoping that
this was the cause of the problem.
|
 |
| There is a nylon tie holding the rubber hoses
together. This needs to be cut off.
There are three rubber hoses attached to the valve
assembly. One for each valve, and one for the pump (the long part).
Mark the hoses so you know which one went where. Take a
flat jaw vice grip pliers or a C clamp, and gently squeeze the hose flat
so that no coolant can pour out of it once it's removed from the valve
assembly. Clamp where circled in the photo.
Have an old spark plug ready to stuff into the hose to
plug it.
|
 |
| I bought two 3/8" and one 1/2" pipe
plugs to cap off the three hoses temporarily.
You will now be able to pull the valve assembly out of
the car.
The car is drive-able in this condition.
|
 |
| The area behind the trim panel was full of
road dirt. I took the opportunity to clean it out. There is a drain hole,
circled, that was plugged up. Make sure you open it up.
|
 |
| After cleaning...
|
 |
| The metal lines for the valve assembly were a
bit rusty as well. I cleaned off the rust as much as possible, and painted
the lines with a rust inhibiting paint.
I also touched up any rusty areas on the frame and other
areas.
Before...
|
|
| ...and after.
|
 |
| The assembly is out and on the work bench.
There was quite a bit of corrosion around the unit.
|
 |
| The next step was to verify if there was an
open circuit in the left valve. I measured the right solenoid and saw that
the coil had continuity as expected.
|
 |
| The left side did indeed show an open
circuit. My problem was the left rear duovalve after all.
|
 |
| There are eight
T10 fasteners holding the pump and the two bobbin housings to the plastic
base of the assembly.
I broke two T10 tips getting 3 fasteners out. The other
5 fasteners were completely seized in the plastic. This surprised me. I
didn't think it would be so difficult to get them out.
I removed the pump from the base, and with this removed,
I was able to grind the heads of the remaining fasteners off with a Dremel
tool (hand grinder).
|
 |
| Carefully pry the bobbin housings from the
base using a sharpened spackling knife. Work slowly and carefully. The
connector part will come off with the housings.
|
 |
| Pull the two bobbin housings and connector
away from the base.
|
 |
| A view of the chambers and passages inside
the base. There was some residue and particulates in the passageways.
|
 |
| These two pointy brass parts are the valves.
You can see the rubber gasket that is supposed to seal the coils from the
coolant.
Again, lots of serious corrosion present.
|
 |
| Remove the valves by pulling them straight up
and out.
Remove the rubber gasket as well.
|
 |
| Pull off the intermediate plate. This will
expose the coil bobbins.
More corrosion. It is clear that coolant got past the
seals and gaskets and into the bobbin area.
|
 |
| Pull the plungers straight up and out.
|
 |
| Another photo of the bobbin area. Major
corrosion in this area.
|
 |
| The next step was to get the connector piece
off of the bobbin housings. This was tricky. It was not clear how this was
attached to the housings.
As you can see by the X-acto knife blade, there is a
soft potting compound in the connector cavities.
|
 |
| The connector piece is made of plastic. There
is a molded tab that mates with tabs on the bobbin housings. By carefully
and gently wiggling and moving the pieces around, I was able to disengage
the connector piece from the bobbin housings.
The magnet wire from the coils must be cut from the
connector piece to separate the pieces from one another.
|
 |
| Now the coil and bobbin can be pulled out of
the bobbin housing.
|
 |
| Inspection of the left bobbin coil showed
that one of the two wires was broken off. This was the cause of the open
circuit. Notice the corrosion on the left wire as well.
|
 |
| Some corrosion and moisture present inside
the housing.
|
 |
| Remove the O rings from
the intermediate plate, and from the posts inside the bobbin housings.
|
 |
| I then blasted the metal parts with aluminum
oxide and glass bead mixture to remove the rust.
I also blasted the metal bracket that the assembly sits
on, but remove the three rubber pieces from it first.
The insides of the bobbin housings have a rubber insert
in them. I did not try to remove them, as they looked like they were in
good shape. They did trap some of the blasting grit though, so I spent a
lot of time making sure I got that grit cleaned out completely.
|
 |
| I painted the metal parts with a few coats of
clear corrosion resistant enamel, and let them dry thoroughly.
Update 7/230/06...
A fellow MB friend, Dave C., recommends
filling any remaining pits with JB Weld (or similar), and filing the
surface smooth after it cures. This will recreate a flat surface. It is
possible that coolant could leak past gaskets due to poor sealing over
pitted areas. Thanks for the comment Dave!
|
 |
| The plastic base was the next problem. Most
of the screws holding the components to it had sheared off in the base. I
carefully drilled out the broken screws using a 0.040" carbide drill bit, and
worked my way up to a #30 bit.
I did not plan on using the same type of fastener. I was
planning on using 4-40 stainless steel screws, washers, and nuts to
sandwich all the parts together. This turned out to be good, because
drilling the old screws out proved to be very challenging.
The bit tips tended to walk off center of the broken
screws, which opened the
holes in the base. This wasn't going to be a problem because of the
different fasteners I was going to use.
Once drilled, I cleaned the base thoroughly to remove
all metal chips and residue.
|
 |
| The next step was to determine what size
magnet wire was used on the bobbin, and how many turns of wire there was.
I carefully unwound the end wire loop from the bobbin,
and began counting as I unwound the wire.
There are 1,150 turns of 28 gage AWS wire, wound CCW
looking at the top of the bobbin.
|
 |
| Here is the culprit. The start wire of the
winding corroded completely through at the start slot. My guess is that
the HML insulation on the wire was scratched through when the coil was
wound. Later, as the coil came into contact with coolant, it began to
corrode and eventually broke.
Although there are O rings sealing the top and bottom of
the bobbins, these components were wet when I disassembled them. It may
not be possible to keep them 100% dry in service.
You can see traces of coolant on the bobbin.
|
 |
| The next step is to pick out all the old
potting compound from the connector assembly. This is a bit time
consuming. Just pick at it slowly with tweezers and a small needle nose
pliers. It will eventually all come out.
|
 |
| I wound new coils using 1,150 turns of AWS 28
gage HML magnet wire. This gave a DCR (resistance) of 17.7 ohms.
This worried me because the DCR of
the original coils was 15.4 ohms. I was concerned that this difference in
resistance would cause faulty operation in the system. I thought perhaps
the valves were proportionately modulated during operation. It turns out
that these valves operate either fully opened or fully closed. Temperature is
controlled by duty cycle, or how long the valve stays open.
The small difference in DCR did not affect operation.
|
 |
| After the coil was wound, I wrapped it with
several wraps of adhesive Kapton tape.
|
 |
| The finished coil.
|
 |
| I bought new O rings for the bobbin housings.
They are Neoprene 70 compound, which is good for coolant. Size is -016 ( a
standard in O rings).
|
 |
| Assemble the coils into the bobbin housings.
|
 |
| Note proper orientation of bobbin inside
housing.
|
 |
| Both bobbins in their housings, and the
connector attached.
|
 |
| Route the coil wires as shown. Poor quality
photo, sorry! See the schematic below for clarification.
|
 |
| This is a schematic of how the coils need to
be wired up.
|
|
| Prepare a small square of sand paper as
shown. This will be used to gently strip the insulation off of the magnet
wire. The insulation must be removed to solder the wire to the connector
terminals.
|
 |
| Carefully sand the insulation off of the coil
wires using the sandpaper.
Gently squeeze the sandpaper on the wire, and pull the
sandpaper towards the end of the wire. You must hold the left side of the
wire (left as pictured) wire to prevent it from being stressed or broken.
|
 |
 |
| Wrap the coil wires around the terminals as
shown.
|
 |
| Solder the wires in place.
|
 |
| Place the O rings onto the intermediate
plate.
|
 |
| Place the plungers into the bobbins.
|
 |
| Assemble the intermediate plate onto the
housings.
|
 |
| I temporarily clamped the plate in place to
do some functional testing. I wanted to verify that the valves were
working before I continued assembling.
I hooked up 12 volts to the center common terminal.
|
 |
| I then grounded the left and right terminals
to verify proper valve action. Note that unenergized, the valves are
retracted or open. Applying voltage extends or closes them.
Both valves work as they should. We can continue
assembling.
|
 |
 |
| Assemble the rubber gasket, and the valves.
|
 |
| Assemble the bobbin / connector assembly onto
the plastic base. I used 1" long 4-40 stainless steel screws,
washers, lock washers, and nuts to hold the bobbin housings onto the base.
|
 |
| Next, reassemble the pump to the plastic
base. I used three 1.5" long screws for the pump, as they needed to
be a bit longer due to the thickness of the pump mounts.
Note proper orientation.
|
 |
| Time to test for functionality again. I
repeated my earlier tests, but this time tried to blow through the port
under each valve to see if it was opening and closing. Both valves worked
fine.
|
 |
| A tip, I keep an old battery charged up just
for these kinds of tests. The core value of the battery is $8.00. It's
worth more to me to have it on hand to power up 12 VDC devices. I throw it
on the charger every now and then to keep it alive.
|
 |
| Now we need to seal up the electrical
connections. I used room temp vulcanizing adhesive sealant. I think any
kind of water-proof caulking would work though.
|
 |
 |
| Assemble the rubber pieces to the bracket.
|
 |
| Place the duovalve assembly onto the mounting
bracket.
Notice that I also cleaned the corrosion off the
connector pins.
|
 |
| Make sure the post on the bottom of the
plastic base mates with the rubber grommet properly.
|
 |
| Using pliers, gently pull the two rubber
mounts up through the holes in the bracket. Make sure they are securing
the duovalve assembly to the bracket securely.
|
 |
| Reinstall the assembly into the car. Don't
forget to clamp the hoses with a locking pliers to prevent coolant from
pouring out.
|
 |
| Start the car, set the rear AC
controls for high heat, and let the engine warm up thoroughly. Watch for
any leaks on the assembly and on the hoses.
I had to let the car run a good 15 - 20 minutes while
playing with the rear AC controls until the air bubbles and pockets were
purged from the lines. It took some time for the system to work as
expected.
I did hear the valves opening and closing as they
should. Once you feel heat and cold the way it should be, you have
verified that the system is properly repaired and functioning. Make
another visual inspect for any leaks prior to assembling all the body
panels.
Good Job! The Rear AC is working again.
|
