Posted: Mon Aug 12, 2002 2:32 pm
Hi folks, I just registered after seeing that you really dig deep into this specific topic, and Enermax have specifically mentioned in the Forum - because my fanless Enermax PS died a week ago, after 7 months of average use.
<BR>
<BR>Background: I built a (almost) completely fanless PC out of a scrap case last December. The system consisted of an Athlon XP1700 on an Epox 8KHA+, with a Radeon 7500, driven by said Enermax supply. The Palomino is cooled by a large industrial heat sink with 0.3K/W, the Radeon has a Zalman heat spreader instead of the fan, and the WD600AB hard disk lives in a Molex SilentDrive case. Targeted goal was to real-time-encode MPEG-4 full-screen without any moving parts (excluding the hard drive spindle). Everything was monitored by a DigitalDoc V 8-point temperature-sensing and fan-controlling unit. The lessons learned with this setup warrant another story.
<BR>
<BR>Anyway, the Enermax (IIRC 330W) was selected for its enlarged heat sinks and reasonable price and reputation. The exterior fan was removed, the interor fan stayed in place. It was vertically mounted in free air to allow a convection stream to flow. Two sensors from the DigiDoc sampled the heatsink temperatures. During operation, the temperatures at the heat sinks stayed in a safe range, about 65 degrees C for the auxiliary rail, and a maximum of 82 degrees C for the main rail during full load (Linux kernel compile loop).
<BR>
<BR>I considered this to be safe, as typical MOSFET junctions limits are well beyond 100 degrees. The condensers are specified to 85 degrees, so they are within limits as well. However, last week, the power supply wouldn't power on, and a testbed setup caused actual lightning and smoke to come from a coil wrapped ferrite. I have not looked into it in detail yet, but it looks like some auxiliary circuit in the Enermax PSs can not reliably stand the heat of permanent fanless running. It does NOT look like the main power rails are blown or anything, but I am not the analog expert to make the final judgement.
<BR>
<BR>Right now, I run the machine on a cheap replacement PSU and the noise already annoys me (my main computer is a "Pismo" Powerbook with an IBM TravelStar 40CS FDB drive, and I'm demanding that the bleeding PC is *more* silent...).
<BR>
<BR>Has any of you a clue exactly what part might have blown in the Enermax?! Or a source for a good solid-state PSU that will reliably output enough to drive the Palomino?!
<BR>
<BR>Rich
<BR>
[addsig]
<BR>
<BR>Background: I built a (almost) completely fanless PC out of a scrap case last December. The system consisted of an Athlon XP1700 on an Epox 8KHA+, with a Radeon 7500, driven by said Enermax supply. The Palomino is cooled by a large industrial heat sink with 0.3K/W, the Radeon has a Zalman heat spreader instead of the fan, and the WD600AB hard disk lives in a Molex SilentDrive case. Targeted goal was to real-time-encode MPEG-4 full-screen without any moving parts (excluding the hard drive spindle). Everything was monitored by a DigitalDoc V 8-point temperature-sensing and fan-controlling unit. The lessons learned with this setup warrant another story.
<BR>
<BR>Anyway, the Enermax (IIRC 330W) was selected for its enlarged heat sinks and reasonable price and reputation. The exterior fan was removed, the interor fan stayed in place. It was vertically mounted in free air to allow a convection stream to flow. Two sensors from the DigiDoc sampled the heatsink temperatures. During operation, the temperatures at the heat sinks stayed in a safe range, about 65 degrees C for the auxiliary rail, and a maximum of 82 degrees C for the main rail during full load (Linux kernel compile loop).
<BR>
<BR>I considered this to be safe, as typical MOSFET junctions limits are well beyond 100 degrees. The condensers are specified to 85 degrees, so they are within limits as well. However, last week, the power supply wouldn't power on, and a testbed setup caused actual lightning and smoke to come from a coil wrapped ferrite. I have not looked into it in detail yet, but it looks like some auxiliary circuit in the Enermax PSs can not reliably stand the heat of permanent fanless running. It does NOT look like the main power rails are blown or anything, but I am not the analog expert to make the final judgement.
<BR>
<BR>Right now, I run the machine on a cheap replacement PSU and the noise already annoys me (my main computer is a "Pismo" Powerbook with an IBM TravelStar 40CS FDB drive, and I'm demanding that the bleeding PC is *more* silent...).
<BR>
<BR>Has any of you a clue exactly what part might have blown in the Enermax?! Or a source for a good solid-state PSU that will reliably output enough to drive the Palomino?!
<BR>
<BR>Rich
<BR>
[addsig]