Good troubleshooting, I've just begun to work on boards and I've killed my fair share of projects by trying to reflow. Every check the board came out healthy was a relief lmao.
Cover the back contact in between the pin on the socket, i think as you close the socket the ground pin(cpu) contact the back of the data pin because there is no plastic to prevent from contacting as the cpu will be moved to the left
@@MainboardMedic The little plastic wall between the ground pin in the socket (which you plucked out) and the data pin was broken off from the beginning. That's why the ground pin on the CPU could short to the data pin in the socket when being pushed to the left by closing the lever. You'd have to bend up the protruding right-side part of the data pin in the socket so the ground pin on the CPU won't be pushed into it (instead of removing the pin from the CPU). But since it didn't fiy the RAM no-detect, there's obviously a second issue.
The cause of the short was obvious from the video: The back side of the data pin hole was bent over to the right. It's quite clear when you compare it to the pins around it - You can see that it's shiny on the right side of the data pin hole where the metal has been bent down flat onto the plastic of the socket, whereas none of the others are. When the lever is lowered, the entire CPU is shifted slightly *to the left*, which was causing that ground pin on the CPU to move to the left, and therefore contacting the part of the data pin hole that was bent to the right. Remove that small bent piece of metal on the right side of the data pin hole and you'll clear the short. You didn't need to remove the ground pin hole from the socket, nor the pin from the CPU - just clear the bent metal between the holes and the problem would be fixed. Your camera doesn't have sufficient magnification to show the problem to you clearly, but to anyone who's done this kind of detail motherboard work, as I have, the problem was pretty obvious. This kind of work is much easier with a stereoscopic magnifier. Cameras are fine, but they don't provide depth perception Magnifiers are expensive, however. Still, a workable one can be cobbled together from parts on eBay.
I would've flipped the MB and tapped the back and see if anything falls out. Next I would've cut the 2 ribs in the socket so the 3 unused pins are one big slot. Next I would've slotted the hole in the socket cover so it doesn't press on the pin any more, ie, it's the cover pushing the CPU to the side that makes the connect.
23:20 what is the shine bend wall or soldering under pins ? your magnifier is horrible and unstable, also will help side light to remove reflections and you can move it from diferent angle . side look on those pins wil help and more zoom on shining metal wall between those pins. short is internal in cpu groung pins are together. nice work and conclusions, i didnt know dumy ram short tester exist, thanks for tip what buy next
yes, no. the ground pin next to the target pin, is connected internally to other ground pins. as the socket lid moves the cpu aary left the pin makes contact with the tiny pit of contact on the right side of the target .. ?? maybe.
Couple things... Is yourr signall path correct. 2nd thing was the cpu latch closed when you tested it without the cpu. 3rd thing have you tested the transisor that feeds that line 4th thing did you look for heat on the back of the board when injecting current.
A worthy attempt :) I understand that AMD has the RAM controller in the CPU itself but Isn't there an IC next to the RAM banks like a driver chip? or perhaps a mosfet or a resistor maybe? It's possible that it may not be shorted but malfunctioning, It may be worth a shot. I have a feeling the CPU socket is OK, rather it seems the fault may be elsewhere. Take care Buddy. Cheers!
First, I would to clean the socket with a cover removed with a compressed air (maybe a conductive particle stuck some where). If that did not help, I would try (without memory and cpu installed) to put a needle to the hole of the pin, which causes the short and then to close the lock of the socket. Then, when the short appears, i would try to burn out the short with injecting high current to it. I do not believe the 3&4 ram slots would start to work, but the short could be removed without the need to remove the pin from the cpu. Anyway, great video! Thanks!
See my other post, the issue was with the bent piece of metal on the right side of the data pin hole contacting the ground pin of the CPU when the socket was closed.
I notice the copper coming out on the right side of the hole seems to come out further than the rest of the holes. Is it possible that when you close the socket that's causing the short.
brother plss help me i have same isue with mpg x570 with ryzen 2600 all slots works butt with 5000 series only first 2 slots works what you think what is issue
I personally reckon that CPU socket is good, I`d make a guess that the short is on the other side of the CPU (so to speak) then hitting the short and the only place your regestering is on the DDR side, of course the issue is then hiding ing the fog of the CPU, so where that line comes out back out (of the cpu) could be a real mystery, also I will add this is all guess work on my part as heck, I`m a lorry driver and don`t fix motherboards!
When you look at the image of the affected pin before the reflow you can see a metal tab on the right hand. I wonder if it was the tab from the right hand side of the pin you removed that has been pushed over to the far left of that pin socket. When you insert the CPU perhaps the leg pushes that tab into the affected pin. You removed the sides but perhaps not the back. There is another one 3 up and from the pin you removed also that looks wrong. I wonder if the previous owner had a CPU with bent legs that caused this damage. I think you should have another look. Look at the image at timestamp 31.37. Just like for LGA sockets look for anything different. I have an AM4 board with a dismantled socket and all the spots for the pins are the same. It does not have the tabs which to me look to be out of place. The other pin is MB_DATA42 and it might be stopping the two affected ram slots from working.
Good troubleshooting, I've just begun to work on boards and I've killed my fair share of projects by trying to reflow. Every check the board came out healthy was a relief lmao.
Cover the back contact in between the pin on the socket, i think as you close the socket the ground pin(cpu) contact the back of the data pin because there is no plastic to prevent from contacting as the cpu will be moved to the left
yes, that's what I think too.
Thats also what I was thinking...
Well the ground pin was removed by me at the end if you finished to watch the Video and the short was gone!
@@MainboardMedic yes but on cpu
@@MainboardMedic The little plastic wall between the ground pin in the socket (which you plucked out) and the data pin was broken off from the beginning. That's why the ground pin on the CPU could short to the data pin in the socket when being pushed to the left by closing the lever. You'd have to bend up the protruding right-side part of the data pin in the socket so the ground pin on the CPU won't be pushed into it (instead of removing the pin from the CPU). But since it didn't fiy the RAM no-detect, there's obviously a second issue.
The cause of the short was obvious from the video: The back side of the data pin hole was bent over to the right. It's quite clear when you compare it to the pins around it - You can see that it's shiny on the right side of the data pin hole where the metal has been bent down flat onto the plastic of the socket, whereas none of the others are. When the lever is lowered, the entire CPU is shifted slightly *to the left*, which was causing that ground pin on the CPU to move to the left, and therefore contacting the part of the data pin hole that was bent to the right. Remove that small bent piece of metal on the right side of the data pin hole and you'll clear the short.
You didn't need to remove the ground pin hole from the socket, nor the pin from the CPU - just clear the bent metal between the holes and the problem would be fixed.
Your camera doesn't have sufficient magnification to show the problem to you clearly, but to anyone who's done this kind of detail motherboard work, as I have, the problem was pretty obvious.
This kind of work is much easier with a stereoscopic magnifier. Cameras are fine, but they don't provide depth perception Magnifiers are expensive, however. Still, a workable one can be cobbled together from parts on eBay.
Your persistence is an inspiration! Keep up the good work.
I would've flipped the MB and tapped the back and see if anything falls out.
Next I would've cut the 2 ribs in the socket so the 3 unused pins are one big slot.
Next I would've slotted the hole in the socket cover so it doesn't press on the pin any more, ie, it's the cover pushing the CPU to the side that makes the connect.
23:20 what is the shine bend wall or soldering under pins ? your magnifier is horrible and unstable, also will help side light to remove reflections and you can move it from diferent angle . side look on those pins wil help and more zoom on shining metal wall between those pins. short is internal in cpu groung pins are together. nice work and conclusions, i didnt know dumy ram short tester exist, thanks for tip what buy next
You could try temporarily cutting the trace near the socket to see if it makes the fault disappear
I was able to ger rid off the short at the end!
yes, no. the ground pin next to the target pin, is connected internally to other ground pins. as the socket lid moves the cpu aary left the pin makes contact with the tiny pit of contact on the right side of the target .. ?? maybe.
9:34 see dirt, it's black in colour in the holes in slots here
Couple things...
Is yourr signall path correct.
2nd thing was the cpu latch closed when you tested it without the cpu.
3rd thing have you tested the transisor that feeds that line
4th thing did you look for heat on the back of the board when injecting current.
A worthy attempt :) I understand that AMD has the RAM controller in the CPU itself but Isn't there an IC next to the RAM banks like a driver chip? or perhaps a mosfet or a resistor maybe? It's possible that it may not be shorted but malfunctioning, It may be worth a shot. I have a feeling the CPU socket is OK, rather it seems the fault may be elsewhere. Take care Buddy. Cheers!
First, I would to clean the socket with a cover removed with a compressed air (maybe a conductive particle stuck some where).
If that did not help, I would try (without memory and cpu installed) to put a needle to the hole of the pin, which causes the short and then to close the lock of the socket. Then, when the short appears, i would try to burn out the short with injecting high current to it. I do not believe the 3&4 ram slots would start to work, but the short could be removed without the need to remove the pin from the cpu.
Anyway, great video! Thanks!
The short was resolved at the end! And the power was only 2.5A so i wouldnt be able to burn it out anyway!
i suspect its chipset trace low ohm when loop closed.
take chipset heatsink off and thermal check w the pin reconnected.
chipset has 15ohm, this is 0,44ohm it is onli probes resistant 😞
See my other post, the issue was with the bent piece of metal on the right side of the data pin hole contacting the ground pin of the CPU when the socket was closed.
Good Job! 🤩
I notice the copper coming out on the right side of the hole seems to come out further than the rest of the holes. Is it possible that when you close the socket that's causing the short.
Also those bent pins that you didnt want to mess with to much, have you checked what those pins do they might be the cause.
They are also responsible for data lines with the DDR and they did show up as fine on my tester!
Good video man 🗿
brother plss help me i have same isue with mpg x570 with ryzen 2600 all slots works butt with 5000 series only first 2 slots works what you think what is issue
have you checked with different 5000 series procecors?
I personally reckon that CPU socket is good, I`d make a guess that the short is on the other side of the CPU (so to speak) then hitting the short and the only place your regestering is on the DDR side, of course the issue is then hiding ing the fog of the CPU, so where that line comes out back out (of the cpu) could be a real mystery, also I will add this is all guess work on my part as heck, I`m a lorry driver and don`t fix motherboards!
When you look at the image of the affected pin before the reflow you can see a metal tab on the right hand. I wonder if it was the tab from the right hand side of the pin you removed that has been pushed over to the far left of that pin socket. When you insert the CPU perhaps the leg pushes that tab into the affected pin. You removed the sides but perhaps not the back. There is another one 3 up and from the pin you removed also that looks wrong. I wonder if the previous owner had a CPU with bent legs that caused this damage. I think you should have another look. Look at the image at timestamp 31.37. Just like for LGA sockets look for anything different. I have an AM4 board with a dismantled socket and all the spots for the pins are the same. It does not have the tabs which to me look to be out of place. The other pin is MB_DATA42 and it might be stopping the two affected ram slots from working.
That might very well be the case, that the privious owner had a damaged CPU! That would explain a lot!
oh, lol, I was supposed to wait for you to remove the pin in the vid. before my last comment.. sorry
Might some kind of fluff
My Asus Maximus Vi Extreme has this problem ☠️
or the chip is compromised.