|
pc
newsgroups
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Upgrade memory on ECS K7S5A mobo?I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz
Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A motherboard. It is an older board (about 3 years). The specs say it can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the memory mentioned above work on this board? Thanks for any feedback. Marcus In article <1783abaf.0502112210.7786d***@posting.google.com>, Marcus
says... Show quote > I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz > Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A > motherboard. It is an older board (about 3 years). The specs say it > can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the > memory mentioned above work on this board? > Yes. -- Conor An imperfect plan executed violently is far superior to a perfect plan. -- George Patton On 11 Feb 2005 22:10:24 -0800, holysmoke***@hotmail.com
(Marcus) wrote: >I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz In theory it should work but in practice a lot of those>Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A >motherboard. It is an older board (about 3 years). The specs say it >can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the >memory mentioned above work on this board? > >Thanks for any feedback. > >Marcus boards were quite buggy, particularly with regards to memory. I'd suggest CAS 2 or 2.5 as a better attempt with that board, possibly still needing to manually lower the timings to CAS3, OR if that didn't work you could then try another motherboard. It could work though, should work, but may not. Be sure to test with memtest86 for several hours.
Show quote
"kony" <spam@spam.com> píse v diskusním príspevku I was using memtest86 with no errors, but later tried goldmemorynews:lfsr01h742h2v1ktgpm6flu38ktb7pem61@4ax.com... > On 11 Feb 2005 22:10:24 -0800, holysmoke***@hotmail.com > (Marcus) wrote: > > >I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz > >Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A > >motherboard. It is an older board (about 3 years). The specs say it > >can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the > >memory mentioned above work on this board? > > > >Thanks for any feedback. > > > >Marcus > > In theory it should work but in practice a lot of those > boards were quite buggy, particularly with regards to > memory. I'd suggest CAS 2 or 2.5 as a better attempt with > that board, possibly still needing to manually lower the > timings to CAS3, OR if that didn't work you could then try > another motherboard. It could work though, should work, but > may not. Be sure to test with memtest86 for several hours. > (shareware), which found some problems. The way to get all working was increasing CPU voltage (1.65->1.70) (AthlonXP 2600+, MSI KT6 Delta) Another CPU - Duron 1300, Abit K7T133, GF4MX440 - had big problems with running 3D apps. The way was increasing voltage again - 1.75->1.825. But I am so unhappy this happens. There are good PSUs (350W first case, 400W fortron second case). I really dont understand where is the problem... Zdenek Sojka On Sat, 12 Feb 2005 14:56:26 +0100, "Zdenek Sojka"
<Xzsoj***@seXznam.cz> wrote: Show quote > It's good that you found the problems but based on what was>"kony" <spam@spam.com> píse v diskusním príspevku >news:lfsr01h742h2v1ktgpm6flu38ktb7pem61@4ax.com... >> On 11 Feb 2005 22:10:24 -0800, holysmoke***@hotmail.com >> (Marcus) wrote: >> >> >I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz >> >Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A >> >motherboard. It is an older board (about 3 years). The specs say it >> >can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the >> >memory mentioned above work on this board? >> > >> >Thanks for any feedback. >> > >> >Marcus >> >> In theory it should work but in practice a lot of those >> boards were quite buggy, particularly with regards to >> memory. I'd suggest CAS 2 or 2.5 as a better attempt with >> that board, possibly still needing to manually lower the >> timings to CAS3, OR if that didn't work you could then try >> another motherboard. It could work though, should work, but >> may not. Be sure to test with memtest86 for several hours. >> > >I was using memtest86 with no errors, but later tried goldmemory >(shareware), which found some problems. needed to solve them, the memory wasn't the problem. I will typically suggest someone run Prime 95's Torture Test for several hours to check the CPU/related. > Are they running hot?>The way to get all working was increasing CPU voltage (1.65->1.70) (AthlonXP >2600+, MSI KT6 Delta) >Another CPU - Duron 1300, Abit K7T133, GF4MX440 - had big problems with >running 3D apps. >The way was increasing voltage again - 1.75->1.825. But I am so unhappy this >happens. There are good PSUs >(350W first case, 400W fortron second case). > >I really dont understand where is the problem... > >Zdenek Sojka The cooler the CPU the lower the voltage needs be. It's not a very large difference but if it were borderline... could just be the boards though. Your problems are unusual compared to those I've seen, usually the CPU stays quite stable at stock voltage and frequency. Hello,
thanks for Prime'95, it is a great program, it helped me so much! The test I used is 'In-place large FFT', which seem to be causing most errors. Tested the Duron 1300 - voltage range 1.70-1.80 locks up PC in a while, 1.825 stops test with error in few minutes, 1.85V causes error in 10s of minutes. Temperatures 1.75V - 52C normal, 62C stressed 1.85V - 57C normal, 70C stressed Barton 2600+ - at 100 and 133 FSB, all is ok for hours on 1.35V, but with 166, I have to change to 1.675V. (at 200, voltage 1.725, but still not able to run without errors... I dont want to overclock anyway) Temperatures 1.35V (133) - 43C normal, 48C stressed 1.675V (166) - 46C normal, 56C stressed I dont think the temperatures are too high... do you? Thanks Zdenek Sojka Show quote "kony" <spam@spam.com> píse v diskusním príspevku news:vi1t01507a8jokq4avvmslmnas245rs91p@4ax.com... > On Sat, 12 Feb 2005 14:56:26 +0100, "Zdenek Sojka" > <Xzsoj***@seXznam.cz> wrote: > > > > >"kony" <spam@spam.com> píse v diskusním príspevku > >news:lfsr01h742h2v1ktgpm6flu38ktb7pem61@4ax.com... > >> On 11 Feb 2005 22:10:24 -0800, holysmoke***@hotmail.com > >> (Marcus) wrote: > >> > >> >I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz > >> >Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A > >> >motherboard. It is an older board (about 3 years). The specs say it > >> >can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the > >> >memory mentioned above work on this board? > >> > > >> >Thanks for any feedback. > >> > > >> >Marcus > >> > >> In theory it should work but in practice a lot of those > >> boards were quite buggy, particularly with regards to > >> memory. I'd suggest CAS 2 or 2.5 as a better attempt with > >> that board, possibly still needing to manually lower the > >> timings to CAS3, OR if that didn't work you could then try > >> another motherboard. It could work though, should work, but > >> may not. Be sure to test with memtest86 for several hours. > >> > > > >I was using memtest86 with no errors, but later tried goldmemory > >(shareware), which found some problems. > > It's good that you found the problems but based on what was > needed to solve them, the memory wasn't the problem. I > will typically suggest someone run Prime 95's Torture Test > for several hours to check the CPU/related. > > > > >The way to get all working was increasing CPU voltage (1.65->1.70) (AthlonXP > >2600+, MSI KT6 Delta) > >Another CPU - Duron 1300, Abit K7T133, GF4MX440 - had big problems with > >running 3D apps. > >The way was increasing voltage again - 1.75->1.825. But I am so unhappy this > >happens. There are good PSUs > >(350W first case, 400W fortron second case). > > > >I really dont understand where is the problem... > > > >Zdenek Sojka > > Are they running hot? > The cooler the CPU the lower the voltage needs be. It's not > a very large difference but if it were borderline... could > just be the boards though. Your problems are unusual > compared to those I've seen, usually the CPU stays quite > stable at stock voltage and frequency. > And...
the cooler as some standard ~ 10USD at Duron, maybe there is a lot of dust on it. The Barton has its boxed AMD cooler, installed from shop, few months old. Show quote "Zdenek Sojka" <Xzsoj***@seXznam.cz> pí¹e v diskusním pøíspìvku news:cuq7ov$r8$1@ns.felk.cvut.cz... > Hello, > > thanks for Prime'95, it is a great program, it helped me so much! > > The test I used is 'In-place large FFT', which seem to be causing most > errors. > > Tested the Duron 1300 - voltage range 1.70-1.80 locks up PC in a while, > 1.825 stops test with error in few minutes, 1.85V causes error in 10s of > minutes. > > Temperatures > 1.75V - 52C normal, 62C stressed > 1.85V - 57C normal, 70C stressed > > Barton 2600+ - at 100 and 133 FSB, all is ok for hours on 1.35V, but with > 166, I have to change to 1.675V. > (at 200, voltage 1.725, but still not able to run without errors... I dont > want to overclock anyway) > > Temperatures > 1.35V (133) - 43C normal, 48C stressed > 1.675V (166) - 46C normal, 56C stressed > > I dont think the temperatures are too high... do you? > > Thanks > Zdenek Sojka > > "kony" <spam@spam.com> píse v diskusním príspevku > news:vi1t01507a8jokq4avvmslmnas245rs91p@4ax.com... > > On Sat, 12 Feb 2005 14:56:26 +0100, "Zdenek Sojka" > > <Xzsoj***@seXznam.cz> wrote: > > > > > > > >"kony" <spam@spam.com> píse v diskusním príspevku > > >news:lfsr01h742h2v1ktgpm6flu38ktb7pem61@4ax.com... > > >> On 11 Feb 2005 22:10:24 -0800, holysmoke***@hotmail.com > > >> (Marcus) wrote: > > >> > > >> >I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz > > >> >Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A > > >> >motherboard. It is an older board (about 3 years). The specs say it > > >> >can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the > > >> >memory mentioned above work on this board? > > >> > > > >> >Thanks for any feedback. > > >> > > > >> >Marcus > > >> > > >> In theory it should work but in practice a lot of those > > >> boards were quite buggy, particularly with regards to > > >> memory. I'd suggest CAS 2 or 2.5 as a better attempt with > > >> that board, possibly still needing to manually lower the > > >> timings to CAS3, OR if that didn't work you could then try > > >> another motherboard. It could work though, should work, but > > >> may not. Be sure to test with memtest86 for several hours. > > >> > > > > > >I was using memtest86 with no errors, but later tried goldmemory > > >(shareware), which found some problems. > > > > It's good that you found the problems but based on what was > > needed to solve them, the memory wasn't the problem. I > > will typically suggest someone run Prime 95's Torture Test > > for several hours to check the CPU/related. > > > > > > > >The way to get all working was increasing CPU voltage (1.65->1.70) > (AthlonXP > > >2600+, MSI KT6 Delta) > > >Another CPU - Duron 1300, Abit K7T133, GF4MX440 - had big problems with > > >running 3D apps. > > >The way was increasing voltage again - 1.75->1.825. But I am so unhappy > this > > >happens. There are good PSUs > > >(350W first case, 400W fortron second case). > > > > > >I really dont understand where is the problem... > > > > > >Zdenek Sojka > > > > Are they running hot? > > The cooler the CPU the lower the voltage needs be. It's not > > a very large difference but if it were borderline... could > > just be the boards though. Your problems are unusual > > compared to those I've seen, usually the CPU stays quite > > stable at stock voltage and frequency. > > > > And...
the cooler as some standard ~ 10USD at Duron, maybe there is a lot of dust on it. The Barton has its boxed AMD cooler, installed from shop, few months old, but there have been problems since I got it. Missing textures in UT at most (that's the only application I get problems... but very often :-( The way I got there is something bad with my computer was very hurting: I have some same data on both computers. After a defragmentation run, the data became different. Archives with CRC errors etc... Show quote "Zdenek Sojka" <Xzsoj***@seXznam.cz> pí¹e v diskusním pøíspìvku news:cuq7ov$r8$1@ns.felk.cvut.cz... > Hello, > > thanks for Prime'95, it is a great program, it helped me so much! > > The test I used is 'In-place large FFT', which seem to be causing most > errors. > > Tested the Duron 1300 - voltage range 1.70-1.80 locks up PC in a while, > 1.825 stops test with error in few minutes, 1.85V causes error in 10s of > minutes. > > Temperatures > 1.75V - 52C normal, 62C stressed > 1.85V - 57C normal, 70C stressed > > Barton 2600+ - at 100 and 133 FSB, all is ok for hours on 1.35V, but with > 166, I have to change to 1.675V. > (at 200, voltage 1.725, but still not able to run without errors... I dont > want to overclock anyway) > > Temperatures > 1.35V (133) - 43C normal, 48C stressed > 1.675V (166) - 46C normal, 56C stressed > > I dont think the temperatures are too high... do you? > > Thanks > Zdenek Sojka > > "kony" <spam@spam.com> píse v diskusním príspevku > news:vi1t01507a8jokq4avvmslmnas245rs91p@4ax.com... > > On Sat, 12 Feb 2005 14:56:26 +0100, "Zdenek Sojka" > > <Xzsoj***@seXznam.cz> wrote: > > > > > > > >"kony" <spam@spam.com> píse v diskusním príspevku > > >news:lfsr01h742h2v1ktgpm6flu38ktb7pem61@4ax.com... > > >> On 11 Feb 2005 22:10:24 -0800, holysmoke***@hotmail.com > > >> (Marcus) wrote: > > >> > > >> >I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz > > >> >Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A > > >> >motherboard. It is an older board (about 3 years). The specs say it > > >> >can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the > > >> >memory mentioned above work on this board? > > >> > > > >> >Thanks for any feedback. > > >> > > > >> >Marcus > > >> > > >> In theory it should work but in practice a lot of those > > >> boards were quite buggy, particularly with regards to > > >> memory. I'd suggest CAS 2 or 2.5 as a better attempt with > > >> that board, possibly still needing to manually lower the > > >> timings to CAS3, OR if that didn't work you could then try > > >> another motherboard. It could work though, should work, but > > >> may not. Be sure to test with memtest86 for several hours. > > >> > > > > > >I was using memtest86 with no errors, but later tried goldmemory > > >(shareware), which found some problems. > > > > It's good that you found the problems but based on what was > > needed to solve them, the memory wasn't the problem. I > > will typically suggest someone run Prime 95's Torture Test > > for several hours to check the CPU/related. > > > > > > > >The way to get all working was increasing CPU voltage (1.65->1.70) > (AthlonXP > > >2600+, MSI KT6 Delta) > > >Another CPU - Duron 1300, Abit K7T133, GF4MX440 - had big problems with > > >running 3D apps. > > >The way was increasing voltage again - 1.75->1.825. But I am so unhappy > this > > >happens. There are good PSUs > > >(350W first case, 400W fortron second case). > > > > > >I really dont understand where is the problem... > > > > > >Zdenek Sojka > > > > Are they running hot? > > The cooler the CPU the lower the voltage needs be. It's not > > a very large difference but if it were borderline... could > > just be the boards though. Your problems are unusual > > compared to those I've seen, usually the CPU stays quite > > stable at stock voltage and frequency. > > > > On Mon, 14 Feb 2005 14:13:28 +0100, "Zdenek Sojka"
<Xzsoj***@seXznam.cz> wrote: >And... Sure, Prime 95 is very useful in that it not only checks>the cooler as some standard ~ 10USD at Duron, maybe there is a lot of dust >on it. >The Barton has its boxed AMD cooler, installed from shop, few months old, >but there have been problems since I got it. Missing textures in UT at most >(that's the only application I get problems... but very often :-( > >The way I got there is something bad with my computer was very hurting: >I have some same data on both computers. After a defragmentation run, the >data became different. Archives with CRC errors etc... whether the system crashes when getting hot, but that it checks it's work so calculation errors are revealed. Your CPUs should run completely stable at stock voltage, _NEVER_ should you need to increase the voltage in a properly working system. Your temperatures didn't seem all that high, not for stock speed stability. Overclockers might need lower temps but that's because they're pushing the CPU, the slower it's running the more tolerant of temps it will usually be. You might check those fans, the dust buildup, and determine whether it's actually heat or power that's causing your problems. I'm not familiar with the MSI KT6 motherboard so I can't help you there. Ah, finally have time to clean that Duron from dust.
btw, I let run Prime 95 for whole day on Athlon on standard voltage and it run without any error... Thanks for replys anyway :-) Zdenek Sojka Show quote "kony" <spam@spam.com> píse v diskusním príspevku news:5l3211t3kn7g41l1hdpgkt4vqq4irdge91@4ax.com... > On Mon, 14 Feb 2005 14:13:28 +0100, "Zdenek Sojka" > <Xzsoj***@seXznam.cz> wrote: > > >And... > >the cooler as some standard ~ 10USD at Duron, maybe there is a lot of dust > >on it. > >The Barton has its boxed AMD cooler, installed from shop, few months old, > >but there have been problems since I got it. Missing textures in UT at most > >(that's the only application I get problems... but very often :-( > > > >The way I got there is something bad with my computer was very hurting: > >I have some same data on both computers. After a defragmentation run, the > >data became different. Archives with CRC errors etc... > > Sure, Prime 95 is very useful in that it not only checks > whether the system crashes when getting hot, but that it > checks it's work so calculation errors are revealed. > > Your CPUs should run completely stable at stock voltage, > _NEVER_ should you need to increase the voltage in a > properly working system. > > Your temperatures didn't seem all that high, not for stock > speed stability. Overclockers might need lower temps but > that's because they're pushing the CPU, the slower it's > running the more tolerant of temps it will usually be. > > You might check those fans, the dust buildup, and determine > whether it's actually heat or power that's causing your > problems. I'm not familiar with the MSI KT6 motherboard so > I can't help you there. > > Zdenek Sojka wrote:
> When errors are occurring, you should avoid doing any writing to.... snip ... > > The way I got there is something bad with my computer was very > hurting: I have some same data on both computers. After a > defragmentation run, the data became different. Archives with > CRC errors etc... the disk. Defragmenting will be fatal to your data if you have a memory fault, for example. That is one more reason you should have ECC memory. Please don't top-post. Your answer belongs after, or intermixed with, the material to which you reply, with non-germane portions snipped. -- "If you want to post a followup via groups.google.com, don't use the broken "Reply" link at the bottom of the article. Click on "show options" at the top of the article, then click on the "Reply" at the bottom of the article headers." - Keith Thompson
Show quote
"CBFalconer" <cbfalco***@yahoo.com> píse v diskusním príspevku Well, I use top-post because its simplier for me. Dont have to scroll wholenews:4211667A.7896C1C7@yahoo.com... > Zdenek Sojka wrote: > > > ... snip ... > > > > The way I got there is something bad with my computer was very > > hurting: I have some same data on both computers. After a > > defragmentation run, the data became different. Archives with > > CRC errors etc... > > When errors are occurring, you should avoid doing any writing to > the disk. Defragmenting will be fatal to your data if you have a > memory fault, for example. That is one more reason you should have > ECC memory. > > Please don't top-post. Your answer belongs after, or intermixed > with, the material to which you reply, with non-germane portions > snipped. > text to see where are replies... And usually I do remember what that article was about. Ok, to do you a favor, I replied you in this style :-) The interesting thing is that errors are made in files, but DMA is used and CPU should not touch the data in any way. But, maybe, it failed sometimes earlier, when I had overclocked memory on Athlon. And data on Duron are compressend (NTFS), so here comes the problem. ECC memory is not suited for me. Normal memory has a parity, doesnt it? I dont eve know whether my MoBos support them (I guess not) Zdenek Sojka Marcus wrote:
> I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz Marcus,> Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A > motherboard. It is an older board (about 3 years). The specs say it > can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the > memory mentioned above work on this board? > > Thanks for any feedback. > > Marcus I too purchased some memory for the same model of MB. Below are my comments from a Posting dated 12-20-04 I made in a different NG: ___________________________ "I have the above MB and currently have two sticks of 256MB PC2700 DDR installed in it. One stick is Kingston Value Ram and the other is a Nanya stick. The computer is solid as a rock running with an Athlon 2200+. I just purchased a single stick of 512MB Kingston Value Ram (also PC2700), and installed it alone in the MB. It recognized the full 512 MB upon boot. After encountering many errors I finally tested the new DDR with DocMem 2.2 and it detected errors at about 256MB. Thinking I had purchased a bad stick, I exchanged it for another like it and found I encountered the same type of error with the DocMem test. After having searched the news groups for comments about Kingston DDR and this MB, I found no comments critical of it, but I did find a poster ask someone who was having problem with DDR if he had double sided DDR? Is there a problem with DDR with chips on both sides with this MB? I tried loosening the settings in CMOS to see if that would allow the DDR to work, but it did not work. I even slowed the settings down to 100 + 100 rather than 133 + 133 and it still failed test. Since PC2700 is faster than a 133 FSB setting requires, I would have thought it would have run easily. Anyone have some insight on using such RAM on this MB? Thanks." ___________________________ Taking the story further, it seems that there is no guarantee that you will or will not have problems. Kingston is generally recognized as decent memory and this stick was rated CL2.5. After the second stick also failed, I took it to a neighbor who tested it on his computer and it passed DocMem several times. The conclusion is that like others have said, the MB is temperamental and might not accept the memory. On the other hand, it might? Ken "Marcus" <holysmoke***@hotmail.com> wrote in message (Don't quote me on this!)news:1783abaf.0502112210.7786d3f1@posting.google.com... >I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz > Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A > motherboard. It is an older board (about 3 years). The specs say it > can handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the > memory mentioned above work on this board? > > Thanks for any feedback. > > Marcus The K7S5A Pro won't boot with a 166FSB - a limit of an on-board oscillator or timer i read. So other than the 133FSB, you'd then need to (incrementally) overclock the FSB using the BIOS (You'd need to flash an overclocking BIOS - easy enough). I remember a ~150FSB being a realistic overclock - the K7S5A Pro would still boot. Once booted into Windows (if that's your OS), you can (or should be able to) software overclock the FSB to 166 (maybe more?) successfully using Speedfan utitlity. (Untested by me). Or you can just let the memory run at 133FSB - but not benefit from it's full potential. I'm itching to ditch my K7S5A Pro and 512MBs of PC133 and get an Athlon64.... I don't think there's point in wasting money upgrading such an antique motherboard! Martin. On Sat, 12 Feb 2005 14:12:01 -0000, "Martin"
<zed***@o2.co.uk> wrote: <snip> >I don't think there's point in wasting money upgrading such an antique Plus quite a few of them had turquise and silver sleeved>motherboard! G-Luxon caps that seemed to prefer a vented disposition (failure). kony wrote:
> On Sat, 12 Feb 2005 14:12:01 -0000, "Martin" Your comment regarding "Turquoise and silver sleeved" caps prompted me > <zed***@o2.co.uk> wrote: > > <snip> > >>I don't think there's point in wasting money upgrading such an antique >>motherboard! > > > > Plus quite a few of them had turquise and silver sleeved > G-Luxon caps that seemed to prefer a vented disposition > (failure). to look at my MB. I have a few of those type caps, but they are not in the regulator circuit. On the MB you are citing, are they everywhere or in some area other than the regulator??? I was just wondering if I have a MB I must watch. Thanks. On Sat, 12 Feb 2005 18:05:35 GMT, Ken <user@domain.invalid>
wrote: Show quote >kony wrote: There are other makes that have also been turquoise and>> On Sat, 12 Feb 2005 14:12:01 -0000, "Martin" >> <zed***@o2.co.uk> wrote: >> >> <snip> >> >>>I don't think there's point in wasting money upgrading such an antique >>>motherboard! >> >> >> >> Plus quite a few of them had turquise and silver sleeved >> G-Luxon caps that seemed to prefer a vented disposition >> (failure). > > Your comment regarding "Turquoise and silver sleeved" caps prompted me >to look at my MB. I have a few of those type caps, but they are not in >the regulator circuit. On the MB you are citing, are they everywhere or >in some area other than the regulator??? I was just wondering if I have >a MB I must watch. Thanks. silver, like Sanyos. Well they might've actually been a little closer to hunter green, but anyway... Among the caps I've seen failed, G-Luxons, particularly the turquoise w/silver, do so quite often. I can't very well claim they're ALL defective, could just be unsuitable for their applied use in some circuits, on some boards. K7S5A has had them all over the boards, of larger (>= 1000 mfd) caps there are roughly 16 or more of those G-Luxons... though at some point ECS replaced 3 of them with Purple & Gold striped Ost caps. Anywhere there was high ripple they'd be subject to failure, and heat shouldn't help either since they're electrolytics. There were several most-likely to fail about the CPU socket, at the top and bottom of the memory slots, nearer the northbridge and AGP slot. Others not (as) likely to fail (or at least not rupturing by the time the board had failed from the OTHER ones failing sooner) might be around the USB pin headers, between PCI slots. I would be weary of relying on any board that uses significant numbers of those caps. I believe G-Luxon also has other cap models with higher impedance and that sometimes a board manufacturer will use the turquoise simply because they bought in high enough volume to use them. Otherwise I've seen boards using better Rubycon, Nichicon, United Chemcon caps in the subcircuits with highe ripple but the Black and Gold (those I presume to be their higher impedance model caps) G-Luxons seem to have fared well elsewhere on the board. I suspect what it boils down to is that the G-Luxon Turquoise and Silver either have ESR that is too high for per applications they're used in, misspec/cost-cutting by the board manufacturers, OR G-luxon themselves has mis-spec'd them... or maybe QC is just so bad that a high percentage of the product is substandard- I dont' know and prefer to just avoid them. I believe G-luxon has also switched to using Pinkish-purple sleeves too, perhaps to mimic Sanyo Oscons, I would guess. Sleeve colors are only a first step in identifying problem areas, but overall it can tend to come back to which manufacturer one chooses, as ECS/PCChips seems to be one of the worst when it comes to barely engineering their boards to work, such that a board may work fine initially but too little margin means the caps wearing out will reduce stability incrementally, and randomly, the worst kind of problem for someone to troubleshoot if the caps haven't visably vented. Back to your original question, you're probably ok if the Turquoise G-Luxons aren't in the regulation subcircuits but their presence could be a sign of other cost-cutting measures that could effect the regulation subcircuit cap spec'd, too... or if it ain't broke don't fix it? No easy answer there until a problem appears and can be traced back to the caps. On Sat, 12 Feb 2005 14:12:01 +0000, Martin wrote:
Show quote > I couldn't agree more with all of your comments.> "Marcus" <holysmoke***@hotmail.com> wrote in message > news:1783abaf.0502112210.7786d3f1@posting.google.com... >>I see what appears to be a very good deal for 512MB PC3200 DDR 400MHz >> Memory CL3. I was thinking of getting 2 sticks for my ECS K7S5A >> motherboard. It is an older board (about 3 years). The specs say it can >> handle max of 1GB RAM DDR. The max FSB is 166mhz I think. Will the >> memory mentioned above work on this board? >> >> Thanks for any feedback. >> >> Marcus > > (Don't quote me on this!) > > The K7S5A Pro won't boot with a 166FSB - a limit of an on-board oscillator > or timer i read. > So other than the 133FSB, you'd then need to (incrementally) overclock the > FSB using the BIOS (You'd need to flash an overclocking BIOS - easy > enough). I remember a ~150FSB being a realistic overclock - the K7S5A Pro > would still boot. > > Once booted into Windows (if that's your OS), you can (or should be able > to) software overclock the FSB to 166 (maybe more?) successfully using > Speedfan utitlity. > (Untested by me). > > Or you can just let the memory run at 133FSB - but not benefit from it's > full potential. > > I'm itching to ditch my K7S5A Pro and 512MBs of PC133 and get an > Athlon64.... > I don't think there's point in wasting money upgrading such an antique > motherboard! > > Martin.  |
|||||||||||||||||||||||
Other interesting topics