Multiple threads have addressed Trailvoy fan clutches and (apparent) failure modes - I'll try to describe the history and detail of the clutch as well as I can for it being 3:00 am...
The TB fan clutch is the first electronically controlled visco clutch to be produced. Like any other cooling clutch it works with silicone fluid that is transferred between an internal reservoir to a working area and is controlled by the opening of a valve. Different from earlier clutches that were heat activated (bimetal) the valve is opened by engaging a coil (that black unit sticking out of the front of the clutch) and creates a magnetic field that attracts the valve.
Initially the distance travelled by the opening valve was controlled by the coil (EV) itself. Rubber buttons were built into the valve to reduce impact noise. If you look at the older fan clutches you'll find shims under the EV unit - that is how valve lever travel was controlled. Valve lever travel is important in maintaining controllability of the clutch. Unfortunately, the first designers (Germany) of the clutch used buttons made of fluoro-silicone and after about 2 years of contact with the silicone fluid they dissolved an fell off.
Once they were floating free they would get trapped under the valve and keep it open - this led to numerous complaints of fan roar and transmissions not shifting. Now the clutch doesn't really impact the trans - but because it took a higher rpm to pump the fluid out of the working area, people thought they were stuck in first gear due to the noise they heard. The buttons got replaced with a Viton material and that particular problem was solved.
That wasn't the end of the bad news though, for a brief period there was another situation where the assembly equipment for the clutch caused some tearing of the threads where the EV mounts to the clutch - this introduced aluminum debris into the clutch and the effect was the same as the free floating rubber buttons - the valve got stuck open again. Ok, so that got fixed too - all EV units are now hand started rather than let the equipment force a cross thread.
The clutch needed work - general improvements for speed (high and low) based on requests from GM Engineering - however other groups in GM held back the release of the new design (US designers now) for almost 2 years... this means TB owners had to live with fan clutch issues for a while longer. Finally it gets released - the new valve uses a mechanical stop instead of the EV unit - this led to greater control of the clutch but, GM was fearful that the clutch would go too fast at low speeds (fuel economy) so they changed the control software that used to work at 100 pwm (pulse width modulation) to 80 pwm. They figured this would be a positive thing, what happened though was all of a sudden you could hear a ticking from the clutch when it was turning at low rpms. SO TAKE NOTE OF THIS: A ticking fan clutch is not going bad - what you are hearing is the control software opening the valve at a regular frequency - tick, tick, tick.
WTF... so why is it ticking? Because the clutch is being told to engage and most likely it is for the a/c head pressure. The inherent problem with visco clutches is the fluid used to drive the fan for speed. It is a visco-elastic fluid. It is chemically indentical to Silly Putty - if you pull it slow it stretches, if you pull it fast it snaps. This is what happens to the fluid during startup conditions - there is not enough fluid in the working area of the clutch to sustain the shear of the startup condtions. The primary input from the clutch goes from 0 rpm to ~1100 rpm in no time flat and the fluid shears. The problem becomes one of an underspeed fan - it isn't turning fast enough to pull air through the a/c condenser and the a/c cuts out to save it's own life. While the head pressure is building the software is telling the clutch to speed up by opening the valve at 80 pwm - there's that damn ticking again. The difficulty here is that at such low speeds there isn't much centripital pressure to encourage the fluid to move out of the internal reservoir - so the problem has the potential to continue beyond the time of safe a/c head pressure and may not resolve itself until you hit the road and ram air helps spin the fan up to speed.
This called for another solution. In the latest design of the clutch there is a feature that allows fluid to drain back into the clutch when it's in the off state (parked overnight kind of thing). Now when you start up there is enough fluid to spin the fan up and keep the a/c working before you get out on the road. This feature came online with the launch of the '06 TB.
Ok, so you'd think everything has been resolved - there have been the occasional leaky valves - and here comes the morning fan noise complaints again. This isn't as bad as it sounds - you can do a neutral rev disengagement of the clutch - meaning leave the truck in park and run the engine up to about 2500 rpm (you probably have a rev limiter that stops you there anyhow). The clutch will disengage (you'll hear the noise fade away) and you'll be on your way - or you can go to get a new clutch put in to solve the noise complaint. All clutches are now leak tested during assembly now to make sure this doesn't happen again.
BTW, if you want to go through the trouble - the leaky valve is a postional problem, meaning sometimes it will give you a noisy clutch and sometimes it won't. It depends on what postion the clutch stops in after shut off. If you pop the hood and examine the clutch (rotate it by hand) you'll find a 4.1 mm ball bearing stuck in the side of it (you have to look through the fan blades to see it). Put that ball bearing down - it is opposite the valve opening. What this does is put the valve up out of the fluid in the reservoir and it can't leak back. If you want to do this then make you life easier by marking the fan opposite the ball and just put the mark on top
Note to aftermarket clutch buyers - some of the aftermarket clutches (made in Korea, Dorman brand) are clones - meaning they weren't designed, but copied. The design they copied was the type of clutch made with rubber buttons in the valves. The have a slightly lower top speed and slightly higher disengaged speed - meaning you could lose some performance when towing and your fuel economy will drop an mpg or two.
Common codes that relate to fan clutches:
P0526 - lost connection to clutch - you can set one of these by unplugging the harness from the tether to check for ticking noises.
P0495 - fan overspeed - got excessive fan roar? That'll do it. Fan higher than requested speed by more than ~300 rpm for longer than 120 seconds.
P0483 - fan underspeed - set when the clutch is told to engage and it doesn't hit the target fan speed within ~300 rpm within 120 seconds.
Codes have to be set twice before it turns on the SES light.
Notes about the clutch: It is attached to the water pump with a 36 mm hex shaft (factory torqued to 90 Nm). The fan bolts are 13 mm (factory torqued to 26 Nm).
With the launch of th '07 TB there was new software, and a new issue - but this time it was on the software side. A software patch fixed the issue of the fan underspeed code being set in overspeed condtions and now the latest design of the clutch is also the most reliable - figures, it will go out of production with the launch of the '08 TB.