|
Post by Macsbeach98 on Jan 19, 2019 20:48:48 GMT -5
Nice
|
|
|
Post by Aleslammer on Feb 24, 2019 20:23:42 GMT -5
Wellll! 8 pin CPU connector on the Asus P5k64 WS got hot enough to fuse the plastic of the PS cable to the plastic of the MB connector, scratch one Asus MB and possible my my backup PS. Sucks have a few scores on the HD.
|
|
|
Post by Macsbeach98 on Feb 24, 2019 22:55:58 GMT -5
Thats not much good Brad that worries me with my P5E3 Pro its only got a 4 pin connector. Board was still going at the end?
|
|
|
Post by cbjaust on Feb 25, 2019 7:26:44 GMT -5
I did that to the 5v rail of an Antec 380w PSU that powered an EPOX 8K5A3+ (Socket A Athlon XP) because no EPS 4 pin 12v. Both still work though.
|
|
|
Post by obscureparadox on Feb 25, 2019 8:22:09 GMT -5
Wellll! 8 pin CPU connector on the Asus P5k64 WS got hot enough to fuse the plastic of the PS cable to the plastic of the MB connector, scratch one Asus MB and possible my my backup PS. Sucks have a few scores on the HD. Sounds like some sort of issue somewhere as you'd never draw enough power through it generally for that to happen. Worst case scenario is a Q6600 or similar 65nm quad core at around 5GHz, that won't draw more than 400w under load in cinebench so no idea how that's happened
|
|
|
Post by ground on Feb 25, 2019 8:32:16 GMT -5
Should be possible to remove the connector from the board and put a new 8 pin on the PSU if you really wanted.
|
|
|
Post by alpi on Feb 25, 2019 11:54:08 GMT -5
Wellll! 8 pin CPU connector on the Asus P5k64 WS got hot enough to fuse the plastic of the PS cable to the plastic of the MB connector, scratch one Asus MB and possible my my backup PS. Sucks have a few scores on the HD. You had some contact issue I think. That's why connector became so hot. Anyway, it's a big loss. Hard ocing with a Q6600 is not a joke as we can see.
|
|
|
Post by Aleslammer on Feb 25, 2019 13:19:53 GMT -5
Image of what stayed behind after it came apart. Smelled something like hot plastic but had the window open to cut down on the heat build up in the room and thinking someone may of got rid of some junk mail in a wood stove.
|
|
|
Post by Macsbeach98 on Feb 25, 2019 16:15:41 GMT -5
Yep connector replacement should fix it its usually caused by a bad contact between the plug and socket. Resistance = Heat
Tip. When you sweat another socket off a dead board plug a 8 pin plug into it off a dead supply that will keep the pins straight in the connector.
|
|
|
Post by alpi on Feb 25, 2019 16:45:17 GMT -5
Image of what stayed behind after it came apart. Smelled something like hot plastic but had the window open to cut down on the heat build up in the room and thinking someone may of got rid of some junk mail in a wood stove. Hmm.. Only one interesting thing is in the story. All the 4 pins were burnt just as the same level. :s So maybe it wasn't a contact issue ? :s I guess if one pin is heating up till the level to burn / melt the whole connector area, it should be seen in cables or in the pcb too. Strange story.
|
|
|
Post by Macsbeach98 on Feb 25, 2019 17:38:51 GMT -5
They all aint making great contact as the one that is making worse contact heats up resistance rises then the others have to compensate they all end up getting hot and all get burnt in the end. Also the place of most resistance is in the plug contact itself thats why it isnt seen in the cables or board just at the connections. If you are a patient man the socket doesnt look damaged make up some fine chisels and hooks and you could probably pick all the black melted plastic out scrap the posts clean and it would be fine.
Bad contact in those sort of sockets isnt that uncommon and can also hold your OC down a bit.
|
|
|
Post by ground on Jun 17, 2019 16:01:57 GMT -5
Most likely had my R9 290X Lightning die on the weekend. It still pulls a little bit of power epowered with a lab PSU, but I'm almost sure the core got hit with most likely 12V when the VRM decided to blow up. Off course it had to happen while dialing in the CPU, GPU at stock sitting in bios , not even running 3D load. At least one of the mosfets was cracked on the inside and it smelled really bad (the whole VRM), so I suspect this thing is toast. Gonna try epowering it tomorrow but hopes aren't high. With this random death as well as watching my friends have an RX 480 die as well as one of my 5970s I sorta feel like 3D really isn't for me. Gonna stay with 2D only from now on, less stuff to go wrong.
|
|
|
Post by Mr.Scott on Jun 17, 2019 17:44:05 GMT -5
You'll get bored. You'll be back to 3D again. I bounce all the time too.
|
|
|
Post by Macsbeach98 on Jun 17, 2019 18:09:52 GMT -5
Interesting you have removed all the mosfets I take it the 2 wires you have soldered on one on the ground rail and one on the output of the mosfets are where you are applying power from your Lab supply what voltage are you giving it nothing over 1.3v I hope? It certainly wont run like that. Also what is wrong with that memory choke the left of the three in the picture or is that just one of those picture anomalies. Also nice job removing the mosfets.
|
|
|
Post by ground on Jun 19, 2019 8:40:56 GMT -5
Never been into 3d, wanted to give it a go and then this randomly happens, not even putting any load on it. Prefer 2D, one less thing that can break And yeah those wires were mainly to check if the core still pulls any current or if it's guaranteed toast. We tried 1.0-1.3V with the lab PSU, it pulled a couple amps still but not enough to output display for sure. Memory chokes are just a mirror + phone cam artifact I guess, they look fine to me. Thats what a decent quality Hot air station excels at really easy to remove mosfets, memory chips, chokes etc from basically anything
|
|
|
Post by Macsbeach98 on Jun 19, 2019 17:32:46 GMT -5
I would say it would Epower Ok a lot of the time when a mosfet blows the GPU is still Ok. I would expect @ 1v with your lab supply the GPU should pull around 10amp check the resistance between the output and ground it should be around 1 ohm or less the GTX 460 I put one on the other day was .6 Ohm. At least its not a Chil controller that controls both GPU and Mem circuits otherwise you would need a Gen5 epower and do both circuits If you try epowering it you have to cut the powergood track on the original controller.
|
|
|
Post by ground on Jul 11, 2019 7:22:13 GMT -5
Unkilled my GD65 last Thursday, though it now looks like a dumpsterfire with the hole in the VRM Now I just have a R3E with a hole in the VRM left to fix... As well as that, with a big surprise my 290X Lightning lives on Epower Epower only quickly mounted to confirm the core isn't dead since the VRM is repairable, just need to source some mosfets for it now. It should've died, the highsides blew and should've sent 12V directly to the core, but I guess OCP on the PSU turned on just in time or something like that... Really glad it lives!
|
|
|
Post by dr4g00n on Jul 11, 2019 8:11:09 GMT -5
Make sure you try installing the drivers before calling the 290x a success. I bought an RX 570 that had a 12v short and epowered it a bit ago. It outputs a display just fine until I try to install the drivers, then it dies.
|
|
|
Post by ground on Jul 11, 2019 8:46:52 GMT -5
Make sure you try installing the drivers before calling the 290x a success. I bought an RX 570 that had a 12v short and epowered it a bit ago. It outputs a display just fine until I try to install the drivers, then it dies. Stopped testing there since this made me confident enough to try putting new mosfets on, especially since there weren't particularly many or thick cables connecting the epower to the card.
|
|
|
Post by Bones on Jul 29, 2019 22:27:28 GMT -5
It's official. I killed the PGX board, more or less. Enough clear seeped in it messed up the socket area and when trying to get it out that messed up a few of the little copper contacts down in the socket, making it stop at post code C5. It's possible but not likely at all it could be fixed, however it's still a project I'll have to go after one day. The other boards came out fine and are working so..... Just had to be the PGX didn't it? Figures.
|
|
|
Post by Vinster on Jul 29, 2019 23:05:51 GMT -5
It's official. I killed the PGX board, more or less. Enough clear seeped in it messed up the socket area and when trying to get it out that messed up a few of the little copper contacts down in the socket, making it stop at post code C5. It's possible but not likely at all it could be fixed, however it's still a project I'll have to go after one day. The other boards came out fine and are working so..... Just had to be the PGX didn't it? Figures. Get a socket on ebay and test out your station with heating bed
Vin
|
|
|
Post by Bones on Jul 29, 2019 23:11:30 GMT -5
Now that's a thought. Not in a hurry to get anything done with it and I'd have to deal with all the clear on the board too - It's actually ready for subzero, fully coated and everything. Still a thought though.
|
|
|
Post by Vinster on Jul 29, 2019 23:16:17 GMT -5
Now that's a thought. Not in a hurry to get anything done with it and I'd have to deal with all the clear on the board too - It's actually ready for subzero, fully coated and everything. Still a thought though. if it's already gone, you have nothing to lose. find a socket on ebay from China with free shipping.. it'll give 30-45 days to do other stuff before it comes in.
you can't change a socket without a heating bed and air station... just be sure to used kapton tape an any parts on the bottom of the socket if there are any.
Vin
|
|
|
Post by Bones on Jul 29, 2019 23:17:58 GMT -5
There certainly is a few to say the least. I'll have to try it at least - As you said, nothing to lose at this point.
|
|
|
Post by Macsbeach98 on Jul 31, 2019 17:16:50 GMT -5
Sorry to say but I dont like your chances of replacing the socket if the clearcoat is under it the heat will turn it to charcoal and burn it to the board. I tried to remove the NB off my last P4C that I had put LET on it and it had wicked under the chip it came off easily but was useless when I got it off. I am not saying dont try to its good for practice and to try out your profile. Johnny Guns gives a good talk on profiles here.
|
|
|
Post by Bones on Jul 31, 2019 19:46:43 GMT -5
I'm going to look into some work on the socket before trying to heat things up. I suspected the clear would be a problem like it is now.
All I can say is I know how it happened and how to avoid it in the future but it took this happening to learn it. Don't worry guys, I've done some crazy stuff in my time, this could be another.
|
|
|
Post by Macsbeach98 on Jul 31, 2019 21:29:06 GMT -5
When I paint them Rodney I use a strip of shoptowel and wind it up into a string and wrap that around the socket then tape it up so paint dont get anywhere near it. I have got a P5Q3 wifi that I got LET under the socket and everytime I take it cold and sit it upside down on the pot to thaw when I pull it apart there is always let up on the bottom of cpu and on the pins I just use a soft brush and acetone to clean the pins from that day on I keep LET and paint well away from the sockets.
|
|
|
Post by Vinster on Jul 31, 2019 23:43:40 GMT -5
it's just a clear coat, I'd use either paint thinner or acetone to remove what you need. should clear it up. Then you could use a brake cleaner to make sure the surface is really clean.
Vin
|
|
|
Post by Macsbeach98 on Aug 1, 2019 8:48:27 GMT -5
Good luck cleaning it all off if its 2 BGA rows under the socket. And if its in the socket I would guess that it wicked up from under it I am sure Rodney would of had the top of it covered well.
|
|
|
Post by Bones on Aug 1, 2019 9:08:50 GMT -5
Actually it seeped in from the top even though I had it covered. What happened is for some reason a corner came up slightly, just enough a little managed to get in and down into the socket on one side.
As said, for the next time I now know how to stop that from happening and will prep it how I'd need to.
|
|