> Looks like 2 of my 3 603s are in need of a (Never fails) KLIMA
> relay. The one I need is a 001 545 80 05 for a 124 87 TD wagon.
They fail just fine. And sometimes resoldering brings them back.
I repaired one, but its problem was an internal relay coil that
opened up. Enclosed are my notes on the subject, Alex Chamberlain
donated the relay towards the list's fund of car repair knowledge.
-- Jim
Conts: (1) Klima from 603.960 (SDL).
001 545 80 05
D-Klima Kickdown
05 8958 40 12V
(1) Note indicating gift to knowledge bank.
Notes: Not for repair, just FYI general list information. I expect
that a resoldering job may cure it. The condition of the pins
was good, not particularly corroded. I then opened it up. It
is built on a single-sided PCB with two relays, two National
Semiconductor 4-bit COP processors (COP411L-MNG/N & COP425LN),
an LM2903 dual comparitor (date codes on the IC's from 84-86),
four signal transistors, eleven diodes, and one power
transistor. Plus the usual crop of passive components,
notably five aluminum electrolytic capacitors. The two relays
looked good, but the soldering on the back of one of them was
fried. The board is even blackened at one pin, but the cooked
pin doesn't have a trace going anywhere else, what's up with
that? Fried drippy solder is in several other places, most of
which aren't high power at all. Did this stuff migrate from
the hot spot(s)? Did a non-solderer attack the board?
Regardless, it appears that this thing might possibly be
brought back to life by a good resoldering, there's plenty of
solder joints that appear crystallized. If, of course, either
masked-ROM COP is dead it's all over, but the rest of the
components look to be generic. A fried relay contact might
also be difficult to deal with, the relays are smallish and
would be hard to substitute. Of note is that one relay is
DPST (NC), and both contacts are wired in parallel. It's
fine. The other is DPDT, both NO contacts in parallel, but
with one NC contact missing, and the burned spot is the lone
NC contact, yet there's no trace from it to anywhere. If both
NC contacts were there and in parallel would this unit have
failed? By spreading the current across two relay contacts
the heat is spread out, and the intertie copper also serves as
a heat sink. The pristine surfaces of the relay contacts
indicate that they weren't overloaded per se, but that the
poor heat dissipation ruined the contact soldering. But what
heat could there be on a lone pin that doesn't go anywhere? I
pulled the (identical) Klima from our SDL, and its solder
looks a whole lot better, though I notice a bit of 'coldness'
starting on some of the relay pins. The fried relay pin
indeed doesn't appear to go anywhere, so how did it get fried?
The _only_ sign of heat is at that one non-connected pin.
It's a mystery. Perhaps I should try to repair it, I've got
a car that it'll fit perfectly in for testing.
Notes: Alex did try resoldering, and claims no skill at all. OK,
weird blobs explained. Denies burning the spot in the board,
so that's a mystery. I'll try repairing this one and swapping
it into the SDL. If it works, I'll add this to my service
list. New Klimas are >$100, so it ought to be cost-effective.
(And will help prolong the supply of NOS spares.)
Notes: August 2 2006: 8:00 PM: Resoldered. Went easily, took a little
less than half an hour. Board is not varnished, so that nasty
paint stripper step is unnecessary. Will set repair rate at
$20, if Klima tests out. Burned mark on board is explainable
as a pad that lifted off during soldering, I had one do that
(it was for uninstalled components) and it left a black mark
where the copper had been. As the original burn was under a
pad that went nowhere, it was easy to overheat and burn/lift
off during soldering. Just like the one I did myself. Will
test Klima at some future convenient time.
Notes: August 3 2006: Klima doesn't work. Back to the drawing board.
Notes: August 4 2006: 9:30 AM: Put the sick
apparently-not-yet-repaired Klima on the bench. (Also brought
in the good one from the car.) Began tracing the schematic.
Measured the tachometer input on the car, the Fluke says it's
about 1800 Hz at 3V at idle. The Klima seems to react at
about 800 Hz, as soon as that frequency is applied on the
bench the kickdown relay activates, allowing the kickdown
switch to work. The good Klima will activate the compressor
clutch relay for a few seconds when the input pin (10) is
grounded, it then cuts it off because the compressor
tachometer input is not running. The bad one doesn't react to
the input pin. The COP processors seem to get about a 376 kHz
clock, they _are_ running: there's plenty of wobbly signals on
their pins. The small COP425 gets a direct copy of the engine
tachometer signal, after being signal conditioned, so for
there to be any reaction I deduce that at least it is working.
I got to probing around, and I found that the transistor drive
to the compressor relay _was_ getting a correct signal, so I
measured the relay's coil resistance. 500 ohms. The kickdown
relay has a 70 ohm coil. Freakish, the repaired electronics
work fine, but the relay itself is defective. You don't see
that too often. Maybe it got cooked? I removed it and then
when I measured the resistance it had gone up to nearly an
open circuit. Finding a replacement relay may be difficult,
the junkbox relays are all much larger than that one.
It appears that the resoldering job was successful, but the
patient died anyway. At the U-Pull today I grabbed a small
japanese relay that might be small enough, when removed from
its shell, to fit. A suggestion that I find a replacement
relay from Mouser (http://www.mouser.com) is a good one, but
they don't actually have an exact replacement. They _do_ have
a few that might be approximately the same size as the
original, which should be easier to use than the cut-down
japanese relay.
Notes: August 5 2006: 9:30 AM: The shell-less U-Pull relay is
_just_ barely small enough to fit in the place occupied by the
dead one, but it's very tight. I cut it down enough to jam in
there and wired it in place, that was very tedious work. Not
a cost-effective maneuver for a commercial effort. Perhaps
one of the inexpensive relays from Mouser would be a better
fit. The Klima appears to work on the bench. Later in the
day I ran some errands, and the repaired Klima worked perfectly.
Memo: "Noticed you were looking for KLIMA failure modes. My '89 560SEL
unit had a cracked solder connection on the board at the
compressor clutch output pin that caused intermittent shutdown
of the A/C. Reflowing the solder resolved the issue."
_______________________________________
http://www.okiebenz.com
For new parts see official list sponsor: http://www.buymbparts.com/
For used parts email [EMAIL PROTECTED]
To Unsubscribe or change delivery options go to:
http://okiebenz.com/mailman/listinfo/mercedes_okiebenz.com
