New PSU no go, but ....
Thread Starter
New PSU no go, but ....
In a moment of weakness I offered to check a friend's computer which "wouldn't start". The description suggested a dead PSU, and component substitution seemed a good place to start. I fitted a 'spare' PSU I had on hand and all system were immediately 'go'. I left it there to keep him airborne, and eventually sourced a nice new Corsair VX450 of slightly higher power capacity than the original and, by all accounts a reliable and highly regarded unit. I had previously installed one in my own primary computer here. Yesterday the time came.
His M/B (now a few years old) uses the older 20 pin power socket with a second 4 pin socket elsewhere on the M/B fed with its own supply lead and plug. The Corsair has a 24 pin plug, so the 'surplus' 4 pin part of the connector was unclipped and ignored. A double check on everything and then the moment of switch-on. Main power leds came on, the CPU fan started up, but things went no further. Wait a short eternity - no progress. Check, recheck. Nothing. The only obvious sign of distress - the HDD led also on. Extreme puzzlement and embarrassment! Had I inadvertently stuffed something? (sense of profound horror at the implications)
The old loan PSU was reinstalled and status returned to 'ops normal' (profound sense of relief!). I returned home and tried the Corsair PSU on a machine I have here with 24 pin connector on board power socket. Worked a charm. Returned to friend's residence with PSU from that machine (860 watt Shaw dual fan model also with 24 pin connector). Installed same, detaching and ignoring the extra 4 pin plug as before. Machine worked immediately. We agreed that it could stay there. Problem solved - but doubts remain.
Begs the question as to why the Corsair supply appears not to 'work' unless all 24 pins are loaded. But why then did the M/B power up top the degree that it did - key board leds as normal, 5 volts on USB external, but nothing further, when a seemingly identical Shaw PSU with similar connections and installation worked OK. The only obvious major difference is that the Shaw was rated at 860 watts - almost 200% of what the machine had in the past (it's a very basic domestic device - nothing suggestive of a larger than normal power requirement - onboard video, one HDD, One DVDrom, 2.9 Ghz Pentium, single memory module etc.).
I need to add of course that due attention was paid to all aspects of the installation - correct plugs in the correct sockets etc. (I've been doing this sort of thing for a while now) One suspects a faulty new PSU, but it works OK on the test bed machine so a return is out of the question.
Has anyone else ever encountered anything like this ??
Thanks in advance.
A very puzzled FOR
His M/B (now a few years old) uses the older 20 pin power socket with a second 4 pin socket elsewhere on the M/B fed with its own supply lead and plug. The Corsair has a 24 pin plug, so the 'surplus' 4 pin part of the connector was unclipped and ignored. A double check on everything and then the moment of switch-on. Main power leds came on, the CPU fan started up, but things went no further. Wait a short eternity - no progress. Check, recheck. Nothing. The only obvious sign of distress - the HDD led also on. Extreme puzzlement and embarrassment! Had I inadvertently stuffed something? (sense of profound horror at the implications)
The old loan PSU was reinstalled and status returned to 'ops normal' (profound sense of relief!). I returned home and tried the Corsair PSU on a machine I have here with 24 pin connector on board power socket. Worked a charm. Returned to friend's residence with PSU from that machine (860 watt Shaw dual fan model also with 24 pin connector). Installed same, detaching and ignoring the extra 4 pin plug as before. Machine worked immediately. We agreed that it could stay there. Problem solved - but doubts remain.
Begs the question as to why the Corsair supply appears not to 'work' unless all 24 pins are loaded. But why then did the M/B power up top the degree that it did - key board leds as normal, 5 volts on USB external, but nothing further, when a seemingly identical Shaw PSU with similar connections and installation worked OK. The only obvious major difference is that the Shaw was rated at 860 watts - almost 200% of what the machine had in the past (it's a very basic domestic device - nothing suggestive of a larger than normal power requirement - onboard video, one HDD, One DVDrom, 2.9 Ghz Pentium, single memory module etc.).
I need to add of course that due attention was paid to all aspects of the installation - correct plugs in the correct sockets etc. (I've been doing this sort of thing for a while now) One suspects a faulty new PSU, but it works OK on the test bed machine so a return is out of the question.
Has anyone else ever encountered anything like this ??
Thanks in advance.
A very puzzled FOR
Hippopotomonstrosesquipidelian title
Join Date: Oct 2006
Location: is everything
Posts: 1,826
Likes: 0
Received 0 Likes
on
0 Posts
The power distribution across the 20-pin and 24-pin connectors is very different between ATX 1.x and ATX 2.x (though the pinouts are the same, of course, for the 20 pins they have in common). Older systems tend to draw power from 5V and 3.3V rails, whilst newer systems tend to draw power from 12V and do on-board conversion. So if you are swapping an old ATX supply with a newer, similarly-rated supply, I suppose you might bump into supply problems. Banging in an apparently over-specced ATX 2.x supply might fix it because although you're not pulling power off the 12V, 5V and 3.3V lines in the expected manner, the 5V and 3.3V lines do have sufficient capacity.
Did you need to connect the CPU power plug for CPUs of a certain age? Couldn't tell from your posting. If I were to guess at the CPU your friend has, I'd think an LGA 775 of some kind, around 3.4 GHz??
Well, a quick post so it's probably wrong :-)
Did you need to connect the CPU power plug for CPUs of a certain age? Couldn't tell from your posting. If I were to guess at the CPU your friend has, I'd think an LGA 775 of some kind, around 3.4 GHz??
Well, a quick post so it's probably wrong :-)
Join Date: Jul 2007
Location: Stockport
Age: 72
Posts: 43
Likes: 0
Received 0 Likes
on
0 Posts
Check that the new PSU is giving a power OK signal (5V) on pin 8. This tells the MB that the voltages are stable and the CPU can start. The test MB may not require the power OK signal. The voltages are present anyway.
The 24 pin thing is a bit confusing. One mboard I have specifically says just use a 20 pin, don't need the other 4.
But it didn't work. Made up a 4 pin to squeeze in at the end of the socket. Works fine like that.
Back to your pal's mb. Did you connect the 4 pin (two black two yellow) to the second power connector on the board? (Known as the P4 plug btw)
But it didn't work. Made up a 4 pin to squeeze in at the end of the socket. Works fine like that.
Back to your pal's mb. Did you connect the 4 pin (two black two yellow) to the second power connector on the board? (Known as the P4 plug btw)
Thread Starter
Thanks to each of you for your responses. The collective wisdom and helpfulness within this forum continues to make it one of the most informative I read - and the most friendly.
All contributions are helpful, and whether hasty or no, it seems to me that the explanation by Bushfiva is the most probable one. I haven't had to study the details about the various power connectors and distribution systems in depth before and this has been educative as well as helpful.
Thanks for the tip about the pin 8 check BTW - that will be followed up when I next have the lid off my 'test' computer which hosts the new PSU,. and also for the reminder about the P4 plug - yes that one was dutifully attended to on each attempt.
best regards,
FOR
All contributions are helpful, and whether hasty or no, it seems to me that the explanation by Bushfiva is the most probable one. I haven't had to study the details about the various power connectors and distribution systems in depth before and this has been educative as well as helpful.
Thanks for the tip about the pin 8 check BTW - that will be followed up when I next have the lid off my 'test' computer which hosts the new PSU,. and also for the reminder about the P4 plug - yes that one was dutifully attended to on each attempt.
best regards,
FOR
I think that it's possible that you have answered your own question.
Given that we hope that PSUs get more clever in power management, is it possible that your fancy new Corsair (a brand name that you'd avoid if you were English and of an age). Is it possible (yes) that it's looking at it's outputs and saying "that's a bit odd, there are four pins not doing anything, and I really really don't know what to do?"
P'raps leave your swap - in and rest on your laurels? After IIABDFI. (If it...).
Given that we hope that PSUs get more clever in power management, is it possible that your fancy new Corsair (a brand name that you'd avoid if you were English and of an age). Is it possible (yes) that it's looking at it's outputs and saying "that's a bit odd, there are four pins not doing anything, and I really really don't know what to do?"
P'raps leave your swap - in and rest on your laurels? After IIABDFI. (If it...).
