2003 Chevy Trailblazer 4.2L
The complaint was a noisy cooling fan and an MIL (malfunction indicator light). There are times when the cooling fan is noisy and it is just a normal condition. The fan is electrically controlled by the engine computer and there are times the fan just needs to be on "high". But, when the fan seems to be noisier than usual and the MIL is on it indicates a problem that needs a closer look.
I pulled engine trouble codes with the scan tool and got P0172, P0128, P1481 and P0493. Sometimes you can get more codes than you want. The trick is to try and hone in on what will be most productive to test. P0172 is a rich code. The fuel mixture was too rich at some point and set a code. Under most conditions a rich condition diagnosis can be difficult and many variables to test. Lets go for an overview before trying to get too involved with a single code. A P0128 is an indication that the engine temperature took too long to get warm from a cold start. P1481 sets if the engine computer cannot see an rpm signal from the cooling fan. P0493 indicates the computer was trying to slow the fan rpm but sees rpms that indicate it is turning faster than commanded.
The last two codes are interesting. If the computer sees no rpm signal from the fan it will command the fan to 100% duty cycle. The code P0493 could come from a fan that is stuck on.
A fan running at full duty cycle on a warm up would set a P0128 and an engine that runs too long cold could possibly set a P0172. On the other hand, a P0172 or a P0128 could not set the fan codes so lets concentrate on the fan codes. Now which one? The P01481 or the P0493? We'll let the car decide that by determining which problem is present at this time.
Starting the car and looking at the scan tool pid for fan rpm I saw that at times 0 rpm were indicated, at times there were high rpms indicated but the desired rpm, or the rpm commanded by the engine computer was never being met. The fan itself was operating steadily, nothing erratic at all.
Lets take a look at the wiring diagram. In the graphing capture above, most of the time the computer is seeing a fan rpm greater than desired. The computer "sees" rpm based on the signal
from the cooling fan hall effect sensor on the dark blue wire at pin D of the fan harness connector. From the spikes in the graph you can see a big jump in rpm and five dropouts to zero so it wouldn't be hard to imagine that there are conditions present at times for both fan codes.
If a scope was connected during the event above you would expect to see the fan speed signal as a 5 volt square wave with frequency indicating the rpm. Because the fan rpm is faster than the computer wants you would expect to see no voltage on the white wire on pin B.
What we know so far is that there is an intermittent problem with the fan speed signal (spikes and dropouts) and that the fan clutch is not reacting as it should to the computer commands. Nothing yet condemns a particular part beyond any doubt but it does have the fan clutch becoming a prime suspect. Since we have a P1481, then at times the fan speed signal must drop out for much longer than we've seen so far. At least 11 seconds of no signal on two consecutive trips. The customer says the noise is worse cold. Since the fan would be loudest at 100% duty cycle and since 100% duty cycle is commanded when there is no fan speed signal it sounds like we might see the conditions for the P1481 on a cold start. Normally you wouldn't expect the fan clutch to engage cold. After letting the car cool down I tried retesting from scratch. The scan data on the restart was not the same at all and the fan, it was very loud.
The scan shows no fan speed rpm even though the fan is turning at a very high rpm. If I put a scope on the fan speed signal at pin D I would expect the square wave signal is missing. If the signal is missing then the computer sees no fan rpm and should have full system voltage on pin B, trying to command the fan to turn faster.
A test of voltage between pin E and C showed that I had a solid ground and reference voltage. The testing confirms a bad cooling fan. Kenneth Hayes
AKA Deranger
