* shawn wilson <ag4ve...@gmail.com> [101226 12:28]: > I don't think that brand or manufacturing process are the issue here > unless you bought cheap asus boards
Cheap boards generally do not have solid capacitors exclusively. > > purchased -- Asus M3A78-T (AMD64) and Asus P5Q-EM (i386) > So, assuming a decent board (it doesn't sound like you have a problem > spending $200+ usd if you're replacing it because video is starting to fail > vs just putting another video card in) How can you trust a board (still in warranty) which has video problems, even if you use an external video card? > then, I wonder about outside factors. First, have you had the > machines plugged into a ups? UPS and surge arrestor. > Did you check your RAM before trashing the boards (and probably in > another computer that doesn't use shared RAM for graphics as I don't > know how memtest86 handles that)? Recall that the M3A78-T made three RMA trips to Asus. If the RAM is defective, Asus should have detected it. > Are you in a real humid or dry setting? Is it real hot all the time? Tropical mosquito swamp. Summer is hot and humid. Winter is cool and humid. There are two nice days a year; one is called "spring" and the other is called "autumn". > As for your issue with electrolytic caps (let me see if I can remember my > electronics here). They are more suited for higher voltages and can hold a > charge longer than the solid state variants. Personally, I like them better > because when they blow, its visually noticeable (mushroom head or > electrolyte all over the place). I am a graduate engineer with electronic expertise; you obviously do not understand electrolytic capacitors. Electrolytics are low-voltage capacitors; they tend to be leaky; they lose capacity with age; the aging process is accelerated by heat; they are subject to internal shorting. Even the best of electrolytics have a rated operating life of about five years. The heat generated by an internal short causes internal pressure to rise and may cause the case to burst. Electrolyte from a burst capacitor can ruin a motherboard. Electrolytic capacitors are widely used because they provide high capacity in a small volume at a relatively low price. Ten years or so ago, electrolytic failures gave every motherboard manufacturer much grief, because after only a three to six months of service, many of the electrolytics had decreased in capacitance to the point that the associated circuitry quit working. This problem was front-page news for months in professional electronic design journals. It is this problem which has lead to the use of so-called "solid capacitors" (there is no such thing as a "solid-STATE" capacitor) on motherboards, despite the higher cost. > Lastly, I've got stereo cross over circuits with those caps that > have been used for 10+ years. And as the electrolytics decrease in capacity, the crossover frequencies change. But the human ear becomes accustomed to slow changes. A frequency response curve made with a calibrated microphone likely would surprise you. > Point of all of this is, in most environments, I wouldn't really > dwell on the caps one way or the other. Buy what works, treat it > well and, in five years or so, you'll end up throwing away an old > motherboard with perfectly good caps. Not so. In five years, the typical electrolytic has only a small fraction of its nominal capacity, so that parameters (such as ripple and time constants) of the circuit of which the capacitor is a part are outside of specification. There may be as many as a hundred capacitors on a motherboard; many of the function as essential elements of the power supply circuit. > As for specific board recommendations, I can't really give you any as I > don't work that way. I either get whatever cheap dell I can get gold support > on and then replace it or I get proliant servers. If this is truly a desktop > system for you and nothing more, you might opt for the dell with gold > support (crap hardware with insurance :) ). Obviously reliability means nothing to you. You really should not speak concerning things of which you are ignorant and about which you are indifferent. All in all, you and a Dell appear to be made for one another. RLH -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101226132256.ga16...@rlharris.org