If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#21
|
|||
|
|||
On Fri, 07 Oct 2005 23:16:11 GMT, "Phil Weldon"
wrote: 2. A display adapter using the nVidia 6800 uses far less power (perhaps half) compared to one using the nVidia 6800 Ultra. Yes, the PSU calculator shows a vanilla 6800 as drawing 49w and an Ultra drawing 77w. 3. New GPUs aren't 'requiring their own power'; some display adapter have several connecting pins for +12V DC; this connects to the same +12V DC rail that drives use. This replaces power that was previously drawn from the +5V DC and/or +3.3V DC rails. The display adapters have on-board DC-to-DC down-convertor/regulators that work more efficiently with +12V DC input than with +5V DC or +3.3V DC input. The connecting wires and traces can be smaller. The higher input voltage provides more stable operation. There is power wasted and less heat produced using +12V DC than with the lower voltages. Just because a display adapter requires a +12V DC connector doesn't mean it draws more power than one which does not - it may even draw less for the reasons just given. Using multiple +12V DC connectors SAVES power because less in wasted in contact and conductor resistance. I've not delved too much into how they work, but I was under the impression that they needed their own power connectors because they draw more watts than can be supplied directly through the (lower voltage) AGP slot. I know they have step down converters to turn the 12v back into vcore usable voltages, and that step-down is more efficient than step-up. 6. | Hmm, yes, questionable. The calc suggested 310w used, so I put in a 350w | cheapo and when it got hot it would power down. Using the mobo power | check, when it was hot the PSU delivered less power, to the point it | would trip. I bought a more expensive 350w dual rail PSU and it was much | more stable. In all probability your system was inadequately ventilated, though it may also be that your power supply was falsly rated. What was the room ambient temperature and the temperature of the air leaving the power supply? Did you check, or did you just rely on what you were told? Power supplies are rated at a specific temperatures. At higher temperatures they must be derated. A power supply that is more than adequate at 60 C can be inadequate at 75 C. At the time I was measuring typically ambient temperatures around 26-28c, CPU 28c, GPU 30c, case 34c, PSU output 32c, drives 35c. I put an extra fan blowing out in bottom rear and the case temp dropped to 30c and the drives 32c. Seems the GPU being a long card and placed roughly in the middle of the case cut the airflow bottom to top, so the extra fan helped the airflow in the lower half of the case. The cheapo PSU would often be outputting air at 42c+ and the case temp was a few degrees higher, it also did not deliver stable 3.3v/12v when it ran hot (for some reason the 1.5v and 5v always stayed stable, the 3.3v would rise to around 3.6v and the 12v would drop as low as 10.5v). The new PSU runs a lot cooler and is more stable in it's power delivery. -- Alfie http://www.delphia.co.uk/ Philosophy is a battle against the bewitchment of our intelligence by means of language. |
#22
|
|||
|
|||
On Sat, 08 Oct 2005 04:28:54 GMT, Al Kaufmann
wrote: On Fri, 07 Oct 2005 23:16:11 GMT, "Phil Weldon" wrote: In all probability your system was inadequately ventilated, though it may also be that your power supply was falsly rated. What was the room ambient temperature and the temperature of the air leaving the power supply? Did you check, or did you just rely on what you were told? Power supplies are rated at a specific temperatures. At higher temperatures they must be derated. A power supply that is more than adequate at 60 C can be inadequate at 75 C. I bought the Antec TruePower supply and I knew that the power supply fans run at the lowest speed appropriate to load and conditions. I guess they do not speed up until the power supply gets hot. Not a good idea if you are into overclocking. I have enough case fans and a very good cpu heatsink and fan but still my system gets about 10C hotter than a similar system with an Enermax power supply and I have been scratching my head about that for a long time now. I even thought I had faulty case fans because they were not pushing much air. I finally read the fine print in the power supply manual and the "special fan only connectors" the power supply provides also regulate the case fans to the lowest speeds! I just wonder how the power supply can tell what temperature I want inside my case. Guess I will be disconnecing those fan only power connectors and using the regular connectors. Ak TBH the one problem I have with Antec PSUs is that their fan speeds (both for their internal fans and fans on the fan only leads) seem to be set far too conservatively: even when things are getting pretty hot, the fans are still not running very fast. Using the PSU temperature to set case fan speeds seems a bit daft to me anyway. I don't use their fan only leads for those reasons, and on my Antec PSU I've connected the fans directly to 12V - I've changed them for better fans, as well. It's a little bit on the noisy side ATM, so at some point I'll probably put a resistor in to reduce the voltage to them, and the noise, a little. Patrick - take five to email me... |
#23
|
|||
|
|||
"AAvK" wrote in message news:iML1f.712$UF4.402@fed1read02... What power supply do you have in this rig? sounds too low end. McG. It's an "Avus Premier Performance Pro Gold 450w"... really a damn good one so far, 'cept fer the damn problem I've got. -- Giant_Alex cravdraa_at-yahoo_dot-com That doesn't mean it will supply the needed current to either the 12v or 5v rails. Another thing is that the motherboards own voltage regulators may not be enough to supply the needed current to the card. Remember, this is a 1.5v card. The mobo has regulator IC's for PCI/PCI-e bus or PCI and AGP buses. Your problem may not show up until the IC is loaded and hot. I haven't heard of Avus. I have heard of and use Thermaltake and Antec. I have a 420W and 430W in use on 2 different rigs, no problems with the cards I'm using. It isn't so much the wattage, but the actual current each rail is designed to supply. Both my PS's have all this info on a label on the PS itself. McG. |
#24
|
|||
|
|||
Switching PSU have to have a minimal load to function, but as long as that
minimal load is met, it shouldn't be a problem. Your load is more than enough to meet the minimal load it would need. -- there is no .sig "Alfie [UK]" wrote in message ... On Fri, 07 Oct 2005 18:51:03 GMT, "Phil Weldon" wrote: I was advised by others not to put in a 500w PSU as I first wanted to do. Yes PSUs have under-draw protection, but prolonged use at low power levels supposedly reduces the lifetime of the PSU. As the 2nd link you provided states; PSUs MAY not have over-temperature protection, where over-temperature is the result of current overload or fan failure. Try a read of http://www.formfactors.org/developer...0Ratified.pdff and http://www.formfactors.org/developer...public_br2.pdf . Pay particular attention to: "System power supply needs vary widely depending on factors such as the application (that is, for desktop, workstation, or server), intended ambient environment (temperature, line voltage), or motherboard power requirements." -- Alfie http://www.delphia.co.uk/ It's called Irony. You know. Like goldy or silvery, only it's made out of iron. |
#25
|
|||
|
|||
"dave" once tried to test me with:
It'll definately be the power supply - we had a problem with an nVidia card and a 450W PSU, the games ran OK but when playing Far Cry for two hours the PSU blew up (obviously the PSU was being overdriven and wasn't regulated poroperly) ! We now have a 750W PSU and everything is fine - If you wish to maintain cutting edge performance without continually replacing PSU you're probably better getting a 1KW PSU :-) holy **** -- Knight37 - http://knightgames.blogspot.com Once a Gamer, Always a Gamer. |
#26
|
|||
|
|||
That doesn't mean it will supply the needed current to either the 12v or 5v rails. Another thing is that the motherboards own voltage regulators may not be enough to supply the needed current to the card. Remember, this is a 1.5v card. The mobo has regulator IC's for PCI/PCI-e bus or PCI and AGP buses. Your problem may not show up until the IC is loaded and hot. I haven't heard of Avus. I have heard of and use Thermaltake and Antec. I have a 420W and 430W in use on 2 different rigs, no problems with the cards I'm using. It isn't so much the wattage, but the actual current each rail is designed to supply. Both my PS's have all this info on a label on the PS itself. McG. http://www.avus.com/default.asp?PageNo=DEFAULT ... looks like they don't make this PSu anymore. It does have the big sticker visible on the side, but I can't understand most of the tech garble like you can. This mobo is an Abit IS7 ("regular" apart from other model types), know anything about it, or where I can read about those IC details? http://www.abit-usa.com/products/mb/...ies=1&model=79 Thanks for the help tho... -- Giant_Alex cravdraa_at-yahoo_dot-com |
#27
|
|||
|
|||
"AAvK" wrote in message news:PO_1f.758$UF4.171@fed1read02... That doesn't mean it will supply the needed current to either the 12v or 5v rails. Another thing is that the motherboards own voltage regulators may not be enough to supply the needed current to the card. Remember, this is a 1.5v card. The mobo has regulator IC's for PCI/PCI-e bus or PCI and AGP buses. Your problem may not show up until the IC is loaded and hot. I haven't heard of Avus. I have heard of and use Thermaltake and Antec. I have a 420W and 430W in use on 2 different rigs, no problems with the cards I'm using. It isn't so much the wattage, but the actual current each rail is designed to supply. Both my PS's have all this info on a label on the PS itself. McG. http://www.avus.com/default.asp?PageNo=DEFAULT ... looks like they don't make this PSu anymore. It does have the big sticker visible on the side, but I can't understand most of the tech garble like you can. This mobo is an Abit IS7 ("regular" apart from other model types), know anything about it, or where I can read about those IC details? http://www.abit-usa.com/products/mb/...ies=1&model=79 Thanks for the help tho... Usually I find this stuff in the back pages of the user manuals, when it is there to be found. I've dl'd the manual for the ABit IS7 series and they don't show this information. 'White Paper' specs are harder to get for consumer items like add in cards and motherboards. Actual schematics are nearly impossible to obtain. I know what the regulator IC's look like and it's usually easy to tell if one is blown. It will have a burnt sooty looking spot near the case on one or two of the legs. This is getting way deeper than any normal consumer will ever go though. Troubleshooting at this level is very difficult for anyone outside the manufacturers production facility. I probably shouldn't have gone this way with you. There are other "likely suspects" to look at and account for first. And, at best, the most thorough process of elimination you can perform will not identify the actual problem. It may get you to the basic section, such as vid card or something in the PSU. Your problem can also be the video card itself. It can still be the PSU. Have you tried to use a different power cable connector than the one you have plugged into the card now? If it were simply resetting the system (arbitrary restart) would have me thinking video card. That it is shutting the system down has me thinking power. The only place you can really look is the PSU itself and any of its connectors. My thoughts on this include that you have already turned off some features that are known to cause stability problems with some ATI cards. I really hope this turns out to a simple problem with a simple solution, like a poor connection of the power cable to the vid card. Let us know what you find, ok? McG. -- Giant_Alex cravdraa_at-yahoo_dot-com |
#28
|
|||
|
|||
Actually, the active components that may fail in DC-DC
downconvertor/regulators are not ICs, but power transistors. The associated capacitors may fail if subjected to prolonged overheating. Obviously a failed power transistor in a DC-DC downconvertor/regulator is not the problem for the original poster as after shut-down the system can be restarted successfully. Phil Weldon "McGrandpa" wrote in message ... | | "AAvK" wrote in message | news:PO_1f.758$UF4.171@fed1read02... | | That doesn't mean it will supply the needed current to either the 12v or | 5v | rails. Another thing is that the motherboards own voltage regulators may | not be enough to supply the needed current to the card. Remember, this is | a 1.5v card. The mobo has regulator IC's for PCI/PCI-e bus or PCI and | AGP buses. Your problem may not show up until the IC is loaded and hot. | I haven't heard of Avus. I have heard of and use Thermaltake and Antec. | I have a 420W and 430W in use on 2 different rigs, no problems with the | cards I'm using. It isn't so much the wattage, but the actual current | each rail is designed to supply. Both my PS's have all this info on a | label on the PS itself. | McG. | | http://www.avus.com/default.asp?PageNo=DEFAULT ... looks like they don't | make | this PSu anymore. It does have the big sticker visible on the side, but I | can't understand | most of the tech garble like you can. This mobo is an Abit IS7 ("regular" | apart from | other model types), know anything about it, or where I can read about | those IC details? | http://www.abit-usa.com/products/mb/...ies=1&model=79 | | Thanks for the help tho... | | Usually I find this stuff in the back pages of the user manuals, when it is | there to be found. I've dl'd the manual for the ABit IS7 series and they | don't show this information. 'White Paper' specs are harder to get for | consumer items like add in cards and motherboards. Actual schematics are | nearly impossible to obtain. I know what the regulator IC's look like and | it's usually easy to tell if one is blown. It will have a burnt sooty | looking spot near the case on one or two of the legs. This is getting way | deeper than any normal consumer will ever go though. Troubleshooting at | this level is very difficult for anyone outside the manufacturers production | facility. | I probably shouldn't have gone this way with you. There are other "likely | suspects" to look at and account for first. And, at best, the most | thorough process of elimination you can perform will not identify the actual | problem. It may get you to the basic section, such as vid card or something | in the PSU. | | Your problem can also be the video card itself. It can still be the PSU. | Have you tried to use a different power cable connector than the one you | have plugged into the card now? If it were simply resetting the system | (arbitrary restart) would have me thinking video card. That it is shutting | the system down has me thinking power. The only place you can really look | is the PSU itself and any of its connectors. | My thoughts on this include that you have already turned off some features | that are known to cause stability problems with some ATI cards. | I really hope this turns out to a simple problem with a simple solution, | like a poor connection of the power cable to the vid card. Let us know what | you find, ok? | McG. | | | -- | Giant_Alex | cravdraa_at-yahoo_dot-com | | | |
#29
|
|||
|
|||
the 3 leg 7905 regulator 'IC' contains power transistor(s). I know what
you're saying, but i haven't seen any of the power transistors actually on the board in the traces to the slots. I have seen several of the regulator ic's. McG. "Phil Weldon" wrote in message link.net... Actually, the active components that may fail in DC-DC downconvertor/regulators are not ICs, but power transistors. The associated capacitors may fail if subjected to prolonged overheating. Obviously a failed power transistor in a DC-DC downconvertor/regulator is not the problem for the original poster as after shut-down the system can be restarted successfully. Phil Weldon "McGrandpa" wrote in message ... | | "AAvK" wrote in message | news:PO_1f.758$UF4.171@fed1read02... | | That doesn't mean it will supply the needed current to either the 12v or | 5v | rails. Another thing is that the motherboards own voltage regulators may | not be enough to supply the needed current to the card. Remember, this is | a 1.5v card. The mobo has regulator IC's for PCI/PCI-e bus or PCI and | AGP buses. Your problem may not show up until the IC is loaded and hot. | I haven't heard of Avus. I have heard of and use Thermaltake and Antec. | I have a 420W and 430W in use on 2 different rigs, no problems with the | cards I'm using. It isn't so much the wattage, but the actual current | each rail is designed to supply. Both my PS's have all this info on a | label on the PS itself. | McG. | | http://www.avus.com/default.asp?PageNo=DEFAULT ... looks like they don't | make | this PSu anymore. It does have the big sticker visible on the side, but I | can't understand | most of the tech garble like you can. This mobo is an Abit IS7 ("regular" | apart from | other model types), know anything about it, or where I can read about | those IC details? | http://www.abit-usa.com/products/mb/...ies=1&model=79 | | Thanks for the help tho... | | Usually I find this stuff in the back pages of the user manuals, when it is | there to be found. I've dl'd the manual for the ABit IS7 series and they | don't show this information. 'White Paper' specs are harder to get for | consumer items like add in cards and motherboards. Actual schematics are | nearly impossible to obtain. I know what the regulator IC's look like and | it's usually easy to tell if one is blown. It will have a burnt sooty | looking spot near the case on one or two of the legs. This is getting way | deeper than any normal consumer will ever go though. Troubleshooting at | this level is very difficult for anyone outside the manufacturers production | facility. | I probably shouldn't have gone this way with you. There are other "likely | suspects" to look at and account for first. And, at best, the most | thorough process of elimination you can perform will not identify the actual | problem. It may get you to the basic section, such as vid card or something | in the PSU. | | Your problem can also be the video card itself. It can still be the PSU. | Have you tried to use a different power cable connector than the one you | have plugged into the card now? If it were simply resetting the system | (arbitrary restart) would have me thinking video card. That it is shutting | the system down has me thinking power. The only place you can really look | is the PSU itself and any of its connectors. | My thoughts on this include that you have already turned off some features | that are known to cause stability problems with some ATI cards. | I really hope this turns out to a simple problem with a simple solution, | like a poor connection of the power cable to the vid card. Let us know what | you find, ok? | McG. | | | -- | Giant_Alex | cravdraa_at-yahoo_dot-com | | | |
#30
|
|||
|
|||
'McGrandpa' wrote:
| the 3 leg 7905 regulator 'IC' contains power transistor(s). I know what | you're saying, but i haven't seen any of the power transistors actually on | the board in the traces to the slots. I have seen several of the regulator | ic's. _____ A 7905 is a +5V three pin LINEAR regulator, NOT a switching regulator or a DC-DC down convertor/regulator. Phil Weldon "McGrandpa" wrote in message ... | the 3 leg 7905 regulator 'IC' contains power transistor(s). I know what | you're saying, but i haven't seen any of the power transistors actually on | the board in the traces to the slots. I have seen several of the regulator | ic's. | McG. | | "Phil Weldon" wrote in message | link.net... | Actually, the active components that may fail in DC-DC | downconvertor/regulators are not ICs, but power transistors. The | associated | capacitors may fail if subjected to prolonged overheating. Obviously a | failed power transistor in a DC-DC downconvertor/regulator is not the | problem for the original poster as after shut-down the system can be | restarted successfully. | | Phil Weldon |
Thread Tools | |
Display Modes | |
|
|