I recently purchased a Passport Studio 320 ("Passport III") and had trouble with it freezing during heavy write activity over Firewire 800. I tested it extensively and found it always froze before being able to fill the drive when using FW800, but did not freeze when reading or using USB for read or write. I'm reasonably convinced it is a firmware bug, but suspect the firmware is not upgradeable. The freezing occurred with 10.5, 10.6, my 2008 iMac and 2003 PowerBook, with and without their drivers. The computer hangs until you disconnect the frozen drive (and 10.6 kernel panicked on me once when I did that).
I engaged in a lengthy exchange with WD tech support over this issue which was generally unsatisfactory until the end, when they offered to replace the drive with a new-generation drive (the one with an LCD screen), which I thoroughly tested and found it does not freeze. I had to push the support people before they offered the replacement, and it was a month after initial contact, but they did come through in the end and have retained me as a customer as a result.
I suspect that not all versions of this drive have the problem, because certainly some people have not experienced it, but I was not able to find enough information online about the issue to be sure. This is one reason I'm documenting it here. My drive was the WD3200MT (the P/N printed on the drive), with firmware 0x817 (use System Profiler to check this). Note that there are two part numbers listed in the WD store right now, WDMS3200TN and WDMT3200TN, at different prices -- does the MS version fix the problem perhaps? The MS original price is listed as $119, the MT $149, suggesting the MT is older. I got my drive at a pretty low price from Amazon.
On the topic of overheating, I concluded that the problem I was having was not thermal, based on monitoring the temperature of the case and noting that the freezing did not correlate with the amount of time it had been writing or the temperature (which didn't vary as much as expected). However, with the replacement, newer-generation drive, in most of my tests after it'd been writing for an hour or so the data rate dropped in half, and I am convinced that this is a thermal-protection feature. It effectively reduces the data rates you can use to write to the full disk, but when I improved cooling around the drive by propping it up and then inserted some short pauses in my test to "rest" the drive I was able to write continuously until the 500 GB was full without the data rate cut at an acceptable decimation of the write time. Adding the rests added 25 minutes to a 2.5 hour test, which is how long it took to fill the disk. This is still much faster than USB, which is the whole point! I measured 36-55 MiB/sec sustained write (inner to outer tracks), 37-68 MiB/sec read (not adjusted for rests). Compare to USB 2.0 that does not exceed 30 MiB/sec.
