Air Conditioner Troubleshooting
From the Chrysler FAQ
From: firstname.lastname@example.org (Bohdan L Bodnar)
This article briefly describes the control system of
computer controlled air conditioning systems which are
typical of Chrysler vehicles which do not have automatic
temperature control. This system is also similar to
that of other manufacturers' products. The following
components are used:
1). An electromagnetic a/c clutch.
2). The fan relay (the relay is a
double-pole-single-throw type; 1/2 is used for fan
control whereas the other 1/2 is used in the relay
logic circuit used to control the a/c clutch).
3). The wide-open throttle (WOT) cutout relay.
4). The pressure cycling switch.
5). The switches used to select air conditioning:
defrost and a/c.
6). The fan switch.
The circuit shown below is representative:
------> to switched battery
_) a/c clutch and surge suppressor
--- 1/2 fan relay
-/- WOT cutout relay
|-----------> to ecm
-/- pressure cycling switch
a/c switch / / defrost switch
/ fan switch (closed if fan is on)
--- chassis ground
The point marked "to ecm" is drawn to a positive value
via a pull up resistor on the computer's board. The
pressure cycling switch is used to prevent ice formation
on the evaporator's core (this would obstruct air flow
and cause system overheating). The WOT cutout relay is
a normally closed one; it is energized (opened) by the
ecm if the ecm detects a WOT condition -- this reduces
engine load during high power demand. The fan relay is
energized by the computer when cooling is required, a/c
is required, and, under appropriate conditions, for
radiator demisting (to reduce steaming when the
vehicle's engine is on, but the vehicle is not moving).
The clutch is an electromagnetically controlled "brake"
which is attracted to the compressor when current is
supplied; the friction material ensures that, if the
clutch is working properly, the compressor will fully
engage. The surge suppressor on Chrysler products is
often a pair of zener diodes mounted back-to-back in the
(polarized) connector going to the clutch. Operation of
the system is as follows (assume that the car is not
1). If the fan is on, adequate pressure is present in
the system (pressure cycling switch is closed), and
either a/c or defrost is selected, the point "to
ecm" is drawn close to ground. The computer
interprets this as an air conditioning demand.
2). A timer (typically, 300 ms to 400 ms) is set.
3). If engine rpm is below a threshold (typically,
around 900 rpm), the rpm is raised to the threshold.
4). Once timer expires, the fan relay is energized. The
compressor is now on. The compressor will continue
to be on unless WOT is reached or the pressure
cycling switch opens. The process then repeats.
The following readings were taken using an OTC 500 meter
and OTC current measuring clamp; they are off of a 1986
- Clutch resistance: 5 ohms, as measured at the clutch
- Operating current: 2.4 amps (per Chrysler service
manual: Current in excess of 5
amps indicates a shorted clutch).
The current will vary with system
- voltage drop to battery's negative post from clutch:
< 500 mv.
A common failure is the fan switch's resistance
increasing. Indeed, these switches are known to melt
from overheating. On 1980s products, this switch is a
replaceable item (about $18 -- dealer only) -- the
entire pushbutton assembly does not require replacing.
Here are some problems I ran across in my Le Baron:
1). BAD FAN SWITCH The switch didn't overheat, but
caused an insufficient current to be delivered to
the clutch. This caused the clutch friction
material to overheat, which caused the clutch to
float on a layer of gas. The symptoms were (1)
gradual reduction in cold air delivery and (2)
smoking refrigerant oil (leaking from a 9 years old
compressor front seal).
2). BAD PRESSURE CYCLING SWITCH AND CONNECTIONS The
pressure switch had bad intermittent connections;
also the terminals in the external connector were
corroded. The symptom was the a/c cycling on and
off about three times per second (recall that the
computer's timer is set to something between 300 and
400 ms). Copious use of television tuner cleaner
and polishing the blade terminals on the switch
would temporarily eliminate the problem. The fix
consisted of replacing the switch (aftermarket
bought from a parts jobber -- about $30), replacing
the terminals in the connector (available in any
hardware store), and spraying the connectors' boot
with silicone lubricant (for a weatherproof seal).
The original terminals were merely crimped; I
crimped and then soldered the connections.
You will require the factory manuals (or equivalent
information) and a high impedance multimeter. If a
digital meter is used, it *must* have a bar graph
display since the digital display with the dual-slope
integrating analog to digital converter will react too
slowly to changes in the circuit. A min/max function is
also useful to have. The voltage drop from the clutch
to the battery's negative terminal MUST be low when full
operating current is seen -- about 700 mv MAXIMUM is ok.
In my case, after all repairs were made, the voltage was
between 400 and 500 mv. For one test, I disconnected
the clutch and replaced it with an unpowered test light
-- this allowed me to check voltage drops with the a/c
always on (pressure cycling switch would never open).
Although full current was never approached (the light
draws only about 300 ma), I did find bad connections.
Measuring individual voltage drops is the ONLY
reasonable approach. Once problems are found,
resistance measurements can be used to home in on target
components. For example, I found bad electrical
contacts by isolating the problem to the pressure
cycling switch, then turning off the engine,
disconnecting the WOT cutout relay connector, and
measuring resistance from the battery's negative post to
the WOT connector. Moving the terminals on pressure
cycling switch's connector showed a huge decrease in
resistance. Television tuner cleaner didn't reduce the
resistance too much, so the terminals were replaced.