Date sent: Wed, 3 Feb 1999 00:38:25 -0700 (MST) Send reply to: classiccmp(a)u.washington.edu From: Jim Strickland <jim(a)calico.litterbox.com> To: "Discussion re-collecting of classic computers" <classiccmp(a)u.washington.edu> Subject: Re: Need some O.T. help Snip! About meltdowns...wow! Where do you got those PSU and what brand says on those PSU's? I think you might have a bum reseller using substandard ATX PSU's? I have worked with ATX PSU to lower level so I could learn how to troubleshoot them down the road. Reasons this is really impossible to burn up several by holding that button in: 1. Softswitch is plugged to the motherboard (that removes any chances of burning up PSU!) 2. Turning on and keeping running the ATX PSU requires the power enable (it's held high themselves.) held low to ground to keep running. Simple job of a right AWG copper wire jammed between that power enable and the ground wire next over. That's is controlled by the motherboard itself when PSU is plugged into motherboard. 3. Motherboard, CPU and ram is enough load as is unless that PSU is poor make. Oh, insure yours is ATX 2.01 spec with .8mA to 1A on standby 5V and fan that blows out. That is where has to put a new rule along with one hand behind back, if needs to work on NLX and ATX machines, unplug that power cord! But this is good for anything else anyway. Jason D.

Well, I fried two power supplies by getting the connector for my 3.5 inch drive off by one pin which apparently grounds the 12 volt circuit. If you hold the button down for more than about 3 seconds, the wire melts and you get a flash of light and copious quantities of smoke out of hte power supply. For reference, the fuse in my fried PS was intact. I had another one that stopped responding to the power-on switch, another that slowly began to emit a stench like burning diesel fuel. I HATE ATX. -- Jim Strickland jim(a)DIESPAMMERSCUMcalico.litterbox.com ----------------------------------------------------------------------- Vote Meadocrat! Bill and Opus in 2000 - Who ELSE is there? -----------------------------------------------------------------------

Max Eskin <max82(a)surfree.com> wrote: I've read the ATX spec, and I can't find anything in it that requires this sort of flaky behavior, or even encourages it the slightest bit. What you are complaining about isn't the ATX standard, but the crappy implementations that typical motherboards and PSUs have. After 18 years of history of IBM PCs and clones, surely you don't expect the things to start being well-engineered? For my own part, I only use Tyan and Asus ATX motherboards, and ATX PSUs from PC Power & Cooling, and of the problems you describe, the only one I've experience is failure to turn off. I haven't attempted to determine the cause. My server machine has been running 24x7 for about 20 months with no unplanned outages except for two power failures that lasted more than 90 minutes. (My software is set to trigger a shutdown at 25% remaining UPS battery level, although I probably should increase the threshold to 75% to reduce the cycling of the gel-cells.) Even after the power outages it started up automatically with no problems. I suspect that all of these problems are due to poorly engineered PSUs, although it is quite possible that some of the fault lies with the motherboard. Poorly enginered PSUs certainly aren't a new phenomenon; the ATX standard has merely given them a few new ways to demonstrate their lack of quality. The PC P&C PSUs seem much better engineered than the dime-a-dozen PSUs that come in the cases. Of course they cost a lot more, but that's no surprise. It seems that in just about every type of product, the majority of consumers are willing to put up with inferior crap as long as it is cheap, so those of us that want quality are forced to pay a much higher price, even though if everyone demanded a quality product it would only increase the cost slightly. Sigh.

A non-text attachment was scrubbed... Name: not available Type: text Size: 871 bytes Desc: not available Url : http://www.classiccmp.org/pipermail/cctalk/attachments/19990204/843de10a/at…

Date sent: Thu, 4 Feb 1999 21:14:05 +0000 (GMT) Send reply to: classiccmp(a)u.washington.edu From: ard(a)p850ug1.demon.co.uk (Tony Duell) To: "Discussion re-collecting of classic computers" <classiccmp(a)u.washington.edu> Subject: Re: Need some O.T. help Tony is right on everything. What makes a reliable machine is 90% preparations and selections of quality stuff, don't have to rob the bank, sheesh! And 10% is correct choices on what software to load onto that HD. I think someone is selling yours too many junk stuff. Oh, Biostar is junk, Sorry. I have already dealt with Biostar and PcMate, Yecch! If someone claims DIMM is PC100, it may not be and just saying the speed of the chip like. 10ns is not it. Must be just PC100 with all the strings attached (reqirements as documented by intel/micron). 1st choices on boards: Asus 1st choice, Microstar (MSI) middling, Aceropoen distrant last. I have no problems shutting down or aborting, (holding the power button til it turns off), the power if I forgot to do something inside or adjust something. BTW mine is P2B, PII 350 PC100 64mb, G200. I now don't remember netscape 3.01 gold crashing or blue screens or any software that cause problems. Oh, aceropen ATX PSU in HX95 case. That PSU in it have very decent big heatsinks. :-) ATX PSU's blowing up is also are junk. I have shorted out many good PSUs and they survived it. (LM339/TI494 design with dual choppers via base trigger isotransformer) I have blown just once a cheapie PSU by hitting that power swtich one too many in close off/on times. on / off / on / off / on pop! While fixing LED's status in a machine. Those I just said about good PSUs they survive that abuse perfectly which this is required when I'm repairing boards on bench. I haven't cracked open both junk and decent ATX PSU's yet. Have anyone got one cheapie atx PSU that are broken to send to me for analysis? Jason D.

Also the SIEMENS PC-D, a early '80s 80186 MS-DOS PC had some hard/software controll over the PS - basicly the on/off switch was in controll, but you where able to set (via a controll reg) a 'power hold device' to override the users power of command, until the Software ended operations. Nice features, and I used it some times - also there has been a FEIN-N signal on the serial ports for remote power on. Basicly the machine was a downgreded UNIX work station ... Gruss H. -- Ich denke, also bin ich, also gut HRK

At 01:16 AM 2/4/99 +0000, Tony wrote: I found one that I wasn't at all impressed with: the rectifier and 2 filter caps were not switched but connected to the AC mains at all times. There was a fuse. The "power" switch was a logic level to the regulator on the main's side. There were no IC's on the low voltage side besides an optocoupler feedback and the linear types for the 2 negative voltages. Haven't found the time yet to repair it. -Dave

A non-text attachment was scrubbed... Name: not available Type: text Size: 1009 bytes Desc: not available Url : http://www.classiccmp.org/pipermail/cctalk/attachments/19990204/59faa710/at…