| Forum Home | ||||
| Press F1 | ||||
| Thread ID: 25250 | 2002-09-29 20:06:00 | Need new Power Supply if upgrade Mboard & Chip? | Steve_L (763) | Press F1 |
| Post ID | Timestamp | Content | User | ||
| 84500 | 2002-09-30 05:33:00 | You might have to upgrade, I know that there are some motherboards that support bothh the old poswer supply's and the New ATX supplys (I forgot the techy name for the older ones... :s) If that's the case, I bought a Tower Case and a 250Watt Power Supply for $60 and haven't looked back! Cheers Chilling_Silence |
Chilling_Silence (9) | ||
| 84501 | 2002-09-30 05:35:00 | im sure that sounds a bit out of contorl to most people! :) | MoNk (2077) | ||
| 84502 | 2002-09-30 05:39:00 | When you consider that an accurate multimeter can cost more than the motherboard (which does not have voltage measuring as its primary function) tweak'e is quite right. A high-count meter will also respond to variations that wont be seen on the on-board monitor function, and its these excursions that can be critical. Its easy enough to measure it approximately with a sensor but I wouldn't rely on it as being exact. Same with temperature. The motherboard will only be using basic a-d conversion for reporting compared to a good multimeter. |
godfather (25) | ||
| 84503 | 2002-09-30 06:28:00 | > So you sit next to your computer while running it > normal > games > apps etc holding yuo multimeter > measure it???? = stupid > > Or do you get an app and use the motherboards voltage > sensors? < do that. Not so stupid actually. that is exactly what I have done in the past when playig with PSUs. Obviously for general use, or t determine if there is a mjor prob in a preiously working unit, the mobo sensors are OK, but if you have a PC that is playing up, and the problem does not seem to be anywhere else, the only way to do it is with an hour or so monitoring the PSU under various conditions with a decent multimeter. > I suggest you try some real world experiences before > you link to people you "think" are right. Do you > beleive everything you read on the net. I didn't link to anyone, but to an article I thought might help steve. You should take your own advice before you make a bigger fool of yourself than you already have. G P |
Graham Petrie (449) | ||
| 84504 | 2002-09-30 06:31:00 | arrogant. | MoNk (2077) | ||
| 84505 | 2002-09-30 07:01:00 | Cool it fellahs. As regards voltage readings by the BIOS etc or voltage settings by jumper, just think a moment, voltages to 3 significant figures? No way are they going to be accurate. Eg my cpu core voltage reads 3.49 instead of 3.3; 1 in 300 roughly, 0.3%, better than a Fluke multimeter? No way, so dont believe all that you read :) I would "believe" an accuracy of maybe 5%, without the benefit of a proper calibration. PSU voltages; it is the spikes that can cause problems, you cant even see them on an ordinary analogue scope, you would need a bit of special transient recording instrumentation to see those. All a multimeter will do is to measure 'steady state' droop of the voltage regulation with current load. |
Terry Porritt (14) | ||
| 84506 | 2002-09-30 07:13:00 | 3.3 is whats set 3.49 is actual. mine is 3.3 set, reading 3.33 |
MoNk (2077) | ||
| 84507 | 2002-09-30 07:35:00 | My watch shows "6:34 39" . I know that the time is about 6:34 (because the watch has a drift . . . it's nearly up to the minute fast at which I correct it) . I say "It's about half past six" . Digital displays show numbers . The numbers are only as correct as the circuit which produces them . A digital voltmeter's accuracy depends on its reference source, and the uncertainty of the comparator, and the number of bits it uses . The temperature dependence of band-gap references is such that they need careful compensation . They don't put references that good on a motherboard . That costs money . That reading of 3 . 49 on a 3 . 3 voltage is about 6% high . Pretty good . But not as good as a proper meter . My Fluke cost $535 . I'd trust it for 3 digits (though I'd still use a scope to look at the line If I saw funny readings) . The sort of A-D converter used on a MB would be about as good as the "A-D" converter used for resistance measurements on a joystick interface . But I would not trust that to better than "1-1/2" digits . A better supply will be better . But one that's good enough will be good enough . An overloaded supply will shut down . The test will be at startup . All the disk drive motors will be pulling their start current, which will be several times the normal running current . (Just thought I'd put in something on topic :D) . |
Graham L (2) | ||
| 84508 | 2002-09-30 10:58:00 | Well well guys, you sure have given me lots to think about. LOL! I posted my Q early this morning and now checking F1 for the first time there are 17 replies, with controversy and even a few flames. Egad, I thought my Q was fairly simple. Ha! So what if I go the cheap way and stick with my old 240w ps? What is the worst that can happen - would the ps just fry and sizzle rather than other components? If there is a chance that an underpowered ps would cause damage further along, e.g. the new mboard & chip & RAM, then of course I would get a new 300 w ps. Oh, BTW, I know that if I were to shut my eyes and just VISA card a $1,000 (or more) upgrade, then all would be nice and safe. I just think that many others here on F1 are in, or will be in, the same boat: wanting to upgrade to a faster mboard-chip - without shelling out big bucks. So what do you say now? Thanks for your assistance...and....the entertainment! ;-) |
Steve_L (763) | ||
| 84509 | 2002-09-30 11:10:00 | the odds are you will need a new case anyway so it would best to get a psu with the new case. running an underpowered psu risks causing brownouts which can cause severe damage to parts. |
tweak'e (174) | ||
| 1 2 3 | |||||