|Zalman VF3000A VGA Cooler|
|Reviews - Featured Reviews: Cooling|
|Written by Bruce Normann|
|Friday, 10 September 2010|
Page 3 of 8
Closer Look: ZALMAN VF3000A
Let's get to the heart of the matter, right now. THIS is the view that you want to see, admit it! I know the top view is attractive and all, but this one tells you what you need to know. Take a look at those five 6mm heatpipes running through the sizeable copper block with the mirror finish. Then look at how much surface area there is for the heatpipes to transfer heat to the aluminum fins. Can you think of a better way of extracting a couple hundred watts of heat from 334 square millimeters of silicon? Yeah, I know - liquid cooling. I've been tempted once or twice, but I always balk at the expense, complexity, and risk. Air cooling is still at the sweet spot of cost v. performance, and the nice thing is, it's a big fat sweet spot with room for a variety of products.
The VF3000A is aimed at the high end of the graphics card spectrum, the HD 58xx series from ATI. While it is currently not the biggest, baddest, hottest GPU around, it was for almost a year until the Fermi introduction in mid 2010. Like most products aimed at PC enthusiasts, these Radeon GPUs get overclocked a lot, and some people found out the hard way where the thermal limits were. The HD 5870 reference cards in particular had a nasty pattern of overheating the VRM section, and there was no real good way of monitoring the temps of those devices, so it happened without warning in many cases. The Zalman VF3000A is specifically targeted to work with reference design cards and to alleviate the VRM heat problem. This has some potential drawbacks for the second generation cards as we'll see later, but I'm sure this product is going to make a lot of reference card owners happy.
The VF3000A is a tall product that sits a good distance above the ATI board. The cooler's heatpipes and fins are responsible for removing heat from the GPU, that's their only task. There are separate heatsinks for the memory and MOSFETs on the board, and they sit in the airflow from the two 92mm fans at either end of the shroud. In order to allow for adequate component clearance, and perhaps some contingency for other products, the fins don't extend down all the way to the board surface. While this is good for cooling all the peripheral board devices, the downside is that the VF3000 is larger than a true dual-slot cooler. It takes up about two and one half slots, once installed on the graphics card. I'll get into detailed measurements in the Installation section a little later, but for now, just think about whether this is going to affect you. It will not be an issue at all for some, and it will be a deal breaker for others.
The fan shroud is a little "over stylized" in my opinion, I could do without the REO Speedwagon wings and all the extra slots and relief cuts in the aluminum. But hey, I'm a Bauhaus guy, and my opinions on industrial design aren't always welcome. Still, as regular readers know, I'm a fan of everything that is anodized blue, red and grey, so it gets extra points for the anodized finish. The shroud also does a good job holding everything together in a rigid unit and provides for mounting of the two fans on a very open plastic subframe.
The fan wiring is consolidated into one lead with a 3-wire electrical connection, and it exits at the rear of the card, furthest away from the I/O plate. It has a standard 3-pin fan connector and will support fan speed monitoring with its tachometer output on the white wire. Most video cards use a smaller connector of slightly different design for their internal fan headers, so there is little chance that you will be able to plug the VF3000 cooler into the graphics board, and retain the use of the on-board fan controller. Some enterprising soul could probably re-terminate the connector and get it to work, but I'll leave that exercise to someone better suited to it than I. There is 280mm (11") of cable provided, which was plenty to reach at least one of the fan headers on my motherboard. The wiring is a little more complicated when the FAN MATE 2 fan controller is used, but it generally improves the cable reach, rather than hindering it.
The fans feature clear blades which are enhanced by blue LED lighting when the VF3000A is operating. The intensity of the LEDs increases and decreases with the voltage supplied to the fan connector, there is no separate wiring for them. The fan blades have ridges molded into them for some aerodynamic purpose. I'm not up enough on my fluid dynamics to know exactly what they do, but since they're on the back side of the blade, I'm guessing they reduce noise. They break up the airflow in a similar way as the notches you see on the edge of some fan blades; other suppliers use dimples for the same reason. Zalman promotes a reduction in "Switching Noise" for these fans, which is usually most evident when running non-PWM at reduced voltages. Since these are standard DC fans and their speed can only be controlled by varying the DC voltage, they are susceptible to this phenomenon. In my testing I never heard any noise that sounded like it was related to this issue; with other fans it sometimes sounds like clicking, at other times like low growling. I ran the fans at all different speeds, but mostly at the lowest and highest, using the supplied controller. Zalman specifies the speed range as 1300-2500 RPM, but I measured 1500-2700 RPM using the monitor function built into the motherboard.
We've loooked at a few details on the Zalman VF3000A already, but click on over to the next section and we'll break it down even further. There are some details you don't want to miss!