I originally thought this was just a PSU problem, so I posted part of the following on the Power Supply forum. Then during my "involuntary work stoppage", I read several pages of threads on this forum, and think this might be the more appropriate place for troubleshooting the problem. I don't mean to intentionally cross-post.
Original post on PSU forum:
---------------------
Here's the situation:
I just finished assembling my build last night. I used a PC Power & Cooling Silencer 750 Quad. Motherboard is Intel DX38BT, CPU is Intel QX9650. Other components connected at the time were one SATA II Seagate 750 GB HD, and one Samsung SATA DVD burner.
I connected up all the connectors: the three to the motherboard (big 24-pin, 8-pin CPU, and 4-pin molex for the PCI bus), one SATA power each to the HD and the burner. All case headers were connected (power switch, HD LED, etc.).
My current system, w/a 510watt PSU, is plugged into an APC RS1200 (1200 watt) Back-UPS. I unplugged it, plugged in the new system w/the PC Power 750watt PSU, and flipped the switch.
Immediately, my APC Back-UPS started screeching, and the "Overload" light lit up. I shut off the new system, quadruple-checked all my connections, and tried again. This time, nothing. No "overload" warning, no nothing.
The first time I started, as best I can recall (having been distracted by the Back-UPS alarm) no fans or HD activity was present even for that short period. The second time, definitely not.
I really dug in and started checking conections, and making sure the graphics card and RAM were seated. Everything checked out. Every connection was where it was supposed to be.
I went to PC Power's site, and they have this troubleshooting advice. I followed each step to the letter, including step #6: shorting between the green and black wires on the 24-pin connector. The PSU fan did not come on.
This seems to indicate a bad PSU. But there are two odd things I want to note: one, when the connectors were still plugged into the motherboard, the tiny green LED next to the 24-pin mobo connector was lit up when the PSU was turned on. Two, when I pressed the power switch on the computer case, the power-on LED blinked on and off, rather than coming on steadily.
So can a PSU be bad, and still supply some power? Enough to light up the 24-pin mobo LED, and make the power switch blink? And what does the power switch LED blinking indicate? (I'm sure it means something; I just don't know what.)
I had originally planned to order the PSU directly from PC Power, like I did last time. But they now have resellers listed on their web site, and one was Amazon, which is where I ordered from. It cost less, and no tax plus free two-day shipping. But it came in a huge box, w/hardly any packing, and now I wonder if it got bumped around so much that it ruined the PSU? Hard to believe; PC Power builds like a tank, but possible, I suppose.
I know most of you experienced guys have gone thru this--putting it all together and flipping the switch, only to discover some part was DOA. This is my first time, and I am crushed. But if it has to be one part, I'd rather it be the PSU, than the motherboard. Can you guys give me some guidance as to if it is indeed the PSU that is dead? (Make sure to read the troubleshooting steps I followed on PC Power's site, so you know what I did to test.) And if it is the PSU, I am sending it back to Amazon and ordering direct from PC Power this time--they actually know how to pack an electronic component.
---------------
Okay, that was the original post on the PSU forum. I think I've fallen prey to "sophomore syndrome"--"sophomore", in the sense that this is my second build, and the first build went off w/out a hitch. IOW, I thought I knew what I was doing. First mistake.
I'm going to tell you guys every remotely stupid thing I can think of that I did, so you can help me troubleshoot. I've read several pages of threads in this forum, paying special attention to those that cry out "No POST!", and some of the situations sound similar, tho not the same.
I'm now thinking I have a short. But I don't know where. If you followed the link in the above part of the post, to PC Power's troubleshooting page, you'll see that I did the "shorting across the green and black wire" step. I gather from the wiring diagram that this is supposed to force the PSU to turn on, but the step also says to hook up two high-capacity hard drives, presumably to load the minimum one amp the PSU needs to turn on. I still have to check the specs on the two Seagate HDs I used for this--mistake #2--to make sure they add up to one amp load.
I'm thinking I should pull everything out of the case and bench-test--which is what I should have done in the first place. Mistake #3. Questions on that:
1) I've seen posts that say, when bench-testing, to place the motherboard on the anti-static bag it came in. But then I've also seen at least one post that said, no, the bag is only to dissipate a charge, and if you put a powered mobo on it, it will actually conduct electricity across the solder points and traces, and to put the mobo on a piece of cardboard. So which is it: bench-test the mobo on top of the anti-static bag, or on a piece of cardboard?
2) I thought the first best way to uncover the short (other than removing the mobo from the case), would be to take out the graphics card and the RAM sticks. From the postings I've read here, the mobo will still start (assuming it's not dead), it will just start beeping that there's no RAM. But at least I'll know if it even CAN start. Is this the correct way to go? But I should leave in the CPU and its HSF, right?
3) Trying #1 and #2 may be moot--is my PSU dead? (BTW, I did plug my old system back into the APC Back-UPS, and it works fine, so the Back-UPS breaker did not trip from the original overload.) I know it could be that the "shorting" trick to get it to come on may not have worked, because the two SATA HDs I used may not have tried to pull enough power. Will it hurt anything at this point to pull out the mobo and the possibly-dead PSU, and try bench-testing them?
Okay, time to 'fess up mistake #4: when I was first installing the mobo in the Lian-li PC-V200Bplus case I bought, and was trying to figure out where to position the stand-offs, I held the board by the Northbridge HS as a handle (mistake #4a), and tried to hover the board over the motherboard compartment in the Lian-li case, as I used a Sharpie to mark thru the mobo mounting holes, down thru to the screw holes in the case. It's possible that at some point during this process, the mobo tipped and its underside came into contact w/the case metal (mistake #4b).
Now I know I sound like a complete idiot, and probably I am. But I WAS wearing an electrostatic wrist strap the entire time, and also constantly touched both the new case, as well as my old system's case, which was plugged in and running at the time. I was especially carefull to do this right before any crucial action--like marking the mobo holes. The case itself was laying on a 100% cotton towel--no acrylic content--on top of a small table (actually my dog's grooming table) which is rubber-topped, and has thick rubber caps on the bottom of the legs.
Still... from all the posts that I read here, and the fact that my APC should easily have been able to handle the wattage requirements of the new system--does this sound like a short? If so, where should I start looking?
Is the PSU dead, or is it in some overload protection state that it will come out of as soon as it gets the right load, without a short? If it is dead, will it hurt the mobo trying to bench-test w/it?
And if the motherboard is dead, how do I actually tell? Is the only way to do it, is start replacing parts to eliminate suspects? Like replace PSU, then bench-test stripped mobo? If something shorted, how likely is it that it made it all the way to the CPU? I'm praying you say "not likely", because that QX9650 cost almost $1100. This would be an expensive lesson.
Before I do any of that, while the motherboard is still in the case, where is the short likely to be? I ask this so I can visually inspect those spots before I pull the mobo, to see if I can find anything? In thinking about it, the I/O shield struck me as seeming kind of odd--nothing I can put my finger on, just that it seems to have more of those little "fingers" than I recall the old one having. BTW, I did make sure that the big "finger" did NOT dive down into the LAN port, but is riding on top of it, inside the case.
I know this has been overly long, but I'm trying to cover all the points I can think of to help you guys help me sort this out. I also know that you experienced guys have probably at one time or another, experienced the sinking sensation of throwing a switch on a new build for the first time, and having nothing happen. Please help me figure out what I did wrong. Thank you.
Original post on PSU forum:
---------------------
Here's the situation:
I just finished assembling my build last night. I used a PC Power & Cooling Silencer 750 Quad. Motherboard is Intel DX38BT, CPU is Intel QX9650. Other components connected at the time were one SATA II Seagate 750 GB HD, and one Samsung SATA DVD burner.
I connected up all the connectors: the three to the motherboard (big 24-pin, 8-pin CPU, and 4-pin molex for the PCI bus), one SATA power each to the HD and the burner. All case headers were connected (power switch, HD LED, etc.).
My current system, w/a 510watt PSU, is plugged into an APC RS1200 (1200 watt) Back-UPS. I unplugged it, plugged in the new system w/the PC Power 750watt PSU, and flipped the switch.
Immediately, my APC Back-UPS started screeching, and the "Overload" light lit up. I shut off the new system, quadruple-checked all my connections, and tried again. This time, nothing. No "overload" warning, no nothing.
The first time I started, as best I can recall (having been distracted by the Back-UPS alarm) no fans or HD activity was present even for that short period. The second time, definitely not.
I really dug in and started checking conections, and making sure the graphics card and RAM were seated. Everything checked out. Every connection was where it was supposed to be.
I went to PC Power's site, and they have this troubleshooting advice. I followed each step to the letter, including step #6: shorting between the green and black wires on the 24-pin connector. The PSU fan did not come on.
This seems to indicate a bad PSU. But there are two odd things I want to note: one, when the connectors were still plugged into the motherboard, the tiny green LED next to the 24-pin mobo connector was lit up when the PSU was turned on. Two, when I pressed the power switch on the computer case, the power-on LED blinked on and off, rather than coming on steadily.
So can a PSU be bad, and still supply some power? Enough to light up the 24-pin mobo LED, and make the power switch blink? And what does the power switch LED blinking indicate? (I'm sure it means something; I just don't know what.)
I had originally planned to order the PSU directly from PC Power, like I did last time. But they now have resellers listed on their web site, and one was Amazon, which is where I ordered from. It cost less, and no tax plus free two-day shipping. But it came in a huge box, w/hardly any packing, and now I wonder if it got bumped around so much that it ruined the PSU? Hard to believe; PC Power builds like a tank, but possible, I suppose.
I know most of you experienced guys have gone thru this--putting it all together and flipping the switch, only to discover some part was DOA. This is my first time, and I am crushed. But if it has to be one part, I'd rather it be the PSU, than the motherboard. Can you guys give me some guidance as to if it is indeed the PSU that is dead? (Make sure to read the troubleshooting steps I followed on PC Power's site, so you know what I did to test.) And if it is the PSU, I am sending it back to Amazon and ordering direct from PC Power this time--they actually know how to pack an electronic component.
---------------
Okay, that was the original post on the PSU forum. I think I've fallen prey to "sophomore syndrome"--"sophomore", in the sense that this is my second build, and the first build went off w/out a hitch. IOW, I thought I knew what I was doing. First mistake.
I'm going to tell you guys every remotely stupid thing I can think of that I did, so you can help me troubleshoot. I've read several pages of threads in this forum, paying special attention to those that cry out "No POST!", and some of the situations sound similar, tho not the same.
I'm now thinking I have a short. But I don't know where. If you followed the link in the above part of the post, to PC Power's troubleshooting page, you'll see that I did the "shorting across the green and black wire" step. I gather from the wiring diagram that this is supposed to force the PSU to turn on, but the step also says to hook up two high-capacity hard drives, presumably to load the minimum one amp the PSU needs to turn on. I still have to check the specs on the two Seagate HDs I used for this--mistake #2--to make sure they add up to one amp load.
I'm thinking I should pull everything out of the case and bench-test--which is what I should have done in the first place. Mistake #3. Questions on that:
1) I've seen posts that say, when bench-testing, to place the motherboard on the anti-static bag it came in. But then I've also seen at least one post that said, no, the bag is only to dissipate a charge, and if you put a powered mobo on it, it will actually conduct electricity across the solder points and traces, and to put the mobo on a piece of cardboard. So which is it: bench-test the mobo on top of the anti-static bag, or on a piece of cardboard?
2) I thought the first best way to uncover the short (other than removing the mobo from the case), would be to take out the graphics card and the RAM sticks. From the postings I've read here, the mobo will still start (assuming it's not dead), it will just start beeping that there's no RAM. But at least I'll know if it even CAN start. Is this the correct way to go? But I should leave in the CPU and its HSF, right?
3) Trying #1 and #2 may be moot--is my PSU dead? (BTW, I did plug my old system back into the APC Back-UPS, and it works fine, so the Back-UPS breaker did not trip from the original overload.) I know it could be that the "shorting" trick to get it to come on may not have worked, because the two SATA HDs I used may not have tried to pull enough power. Will it hurt anything at this point to pull out the mobo and the possibly-dead PSU, and try bench-testing them?
Okay, time to 'fess up mistake #4: when I was first installing the mobo in the Lian-li PC-V200Bplus case I bought, and was trying to figure out where to position the stand-offs, I held the board by the Northbridge HS as a handle (mistake #4a), and tried to hover the board over the motherboard compartment in the Lian-li case, as I used a Sharpie to mark thru the mobo mounting holes, down thru to the screw holes in the case. It's possible that at some point during this process, the mobo tipped and its underside came into contact w/the case metal (mistake #4b).
Now I know I sound like a complete idiot, and probably I am. But I WAS wearing an electrostatic wrist strap the entire time, and also constantly touched both the new case, as well as my old system's case, which was plugged in and running at the time. I was especially carefull to do this right before any crucial action--like marking the mobo holes. The case itself was laying on a 100% cotton towel--no acrylic content--on top of a small table (actually my dog's grooming table) which is rubber-topped, and has thick rubber caps on the bottom of the legs.
Still... from all the posts that I read here, and the fact that my APC should easily have been able to handle the wattage requirements of the new system--does this sound like a short? If so, where should I start looking?
Is the PSU dead, or is it in some overload protection state that it will come out of as soon as it gets the right load, without a short? If it is dead, will it hurt the mobo trying to bench-test w/it?
And if the motherboard is dead, how do I actually tell? Is the only way to do it, is start replacing parts to eliminate suspects? Like replace PSU, then bench-test stripped mobo? If something shorted, how likely is it that it made it all the way to the CPU? I'm praying you say "not likely", because that QX9650 cost almost $1100. This would be an expensive lesson.
Before I do any of that, while the motherboard is still in the case, where is the short likely to be? I ask this so I can visually inspect those spots before I pull the mobo, to see if I can find anything? In thinking about it, the I/O shield struck me as seeming kind of odd--nothing I can put my finger on, just that it seems to have more of those little "fingers" than I recall the old one having. BTW, I did make sure that the big "finger" did NOT dive down into the LAN port, but is riding on top of it, inside the case.
I know this has been overly long, but I'm trying to cover all the points I can think of to help you guys help me sort this out. I also know that you experienced guys have probably at one time or another, experienced the sinking sensation of throwing a switch on a new build for the first time, and having nothing happen. Please help me figure out what I did wrong. Thank you.
, I don't want to fry the replacement PSU either.
Facebook
Twitter