Date: Sat, 03 Aug 1996 08:54:29 -0700 From: "David E. Tweten" <tweten@frihet.com> To: BRETT_GLASS@ccgate.infoworld.com Cc: pechter@shell.monmouth.com, hardware@FreeBSD.org Subject: Re: SCSI Drive recommendation Message-ID: <199608031554.IAA01070@ns.frihet.com>
next in thread | raw e-mail | index | archive | help
pechter@shell.monmouth.com said: >Any views on the DEC DSP5200S scsi drive. I'm considering one -- >since it meets my needs -- under $500 and around 2GB. To which I replied: >Mine has been running pretty much full time for two years. It is a lot >quieter and cooler to the touch than the Toshiba 3 1/2" drives on my >machine at work. To which BRETT_GLASS@ccgate.infoworld.com replied: >Mine is the hottest drive I own -- almost makes your finger sizzle. >And the case has a LOT of surface area, which means a lot of power >dissipation. Okay, curiosity killed the cat, but would it sizzle my fingers? The above degree of contradiction inspired me to doubt my memory, so I opened up the machine's tower case. It has been running continuously for weeks. My full tower case has an exhaust fan in the power supply in the top back corner. It has a slightly bigger intake fan for filtered air in the lower front corner. I like cases to be slightly pressurized with filtered air -- that keeps dust accumulation off the diskette, tape and CD mechanisms. The DSP5200S is mounted in the top front corner, behind the case's switches and LEDs and just above the stack of tape, CD and floppy drives. That's about as far away from either fan as it could get. At the top of the case, it's also in a location where heat might be expected to accumulate. The result? The DSP2500S is warm but not hot to the touch, and the part of the metal case that clears its top by less than a quarter inch is cold to the touch. Using the temperature probe that came with my DVM, I measured the drive's HDA case surface temperature at 94 degrees F, not even human body temperature. According to its specs, it should typically dissipate 17 watts spinning continuously with no seeks. Under continuous random seeks, that should rise to 23 watts. Spin-up consumption should be 57 watts. The coresponding maximums are 20 watts, 28 watts, and 66 watts, respectively. BRETT_GLASS@ccgate.infoworld.com said: >It has problems on Sun 3s, and some PC-compatible controllers will >give you infrequent but resounding core dumps and segmentation faults >under FreeBSD. Mine's been terminating the inside-the-case end of the SCSI bus served by my Adaptec 1742 for two and a half years. In that time, my FreeBSD 1.1.5.1, 2.1, and 2.1 Stable systems have never seg faulted or core dumped except when I've been fiddling with the software, or doing a system installation. The Adaptec must not be one of "some PC-compatible controllers." Lest this read like a ringing endorsement, rather than just a defense of the positive part of my previous neutral statement, I'll repeat. It's old (purchased on sale over 2 1/2 years ago), big (5.25 inch, full height), and slow (12.9 ms seek, 3600 rpm). So is the rest of my machine (4 1/2 years, full height tower, and 486DX33). So am I (none of your business). The lesson? Who knows. Maybe, "Your mileage may vary." -- David E. Tweten | PGP Key fingerprint: | tweten@frihet.com 12141 Atrium Drive | E9 59 E7 5C 6B 88 B8 90 | tweten@and.com Saratoga, CA 95070-3162 | 65 30 2A A4 A0 BC 49 AE | (408) 446-4131
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199608031554.IAA01070>