The true value is the content and the learning you get out of these. If you fail to repair a board, you get a supply of spare parts from it which is very useful for future jobs. I suspect that repairs have been attempted on the majority of faulty mainboards you find in places like ebay, particularly if they come in lots. The everyday user with a board that has failed does not know which part is faulty and generally won't take the system apart to sell individual components. These are most often prebuilts. Such a user will throw it out, or send the system in for repairs. Some repair centers will do component-level work but more often they will swap the board for a working one. Bad boards may get sent or sold to a facility that does component-level work. Boards they can't fix may be resold individually or in lots. Unless you work in such a repair center, expect tons of boards with failed repairs.
Usually the Oszillator is used in a "free-running" mode, this means the Oszillator shall start oszillating by itself as soon as the voltage is applied.
@@MainboardMedic I flashed BIOSes for 3xx and 4xx series (best was a MSI x470), Asus MSI Gigabyte. So far, from this tiny experience (20 boards ca.), if a flash fails it's always because of an installed Cpu. If it still fails I remove the Ram. Now if it fails with cpu on I remove cpu + ram immediately. My guess is that 1.8V goes charging some caps and or is sunk by the cpu making under 1.8 or diminishing the total Amps. These are DC +V, maybe a tiny bit makes a difference. I'm no technician, just observations.
Non-powered semiconductors can put some heavy load onto all connected data signals. With the BIOS EEPROM, it shares a 4-bit SPI bus. Also it's important not to use more pull-up resistors as the specification allows. If the board contains the pull-up resistors and the programmer also contains the pull-up resistors, but it's used as a in-circuit programmer, then the programming might be unsuccessful.
Can you put an artificial signal to the PWM IC to get the VRM working. I was thinking a while back about all the checking AND gate transistors on a MB, and don't all 3 pins have to be high for an OK signal, ie, if pin 1 and 2 is high turn on pin 3. Like one high and the next low to turn on doesn't make sense. So you could check them quickly looking for 3 highs.
Not to sure if you are mixing 3 pin mosfets and AND gates here. There are many 3 pin sot-23 mosfets that act as pulling low to ground resistors. Meaning only 1 pin the gate is high and the other 2 are low, which is the normal operating procedure. The AND gates are most off the time DUAL AND gates which means they have 2 Inputs that need to be high in order for the output to be pulled to VCC. So There are 2 more pins that are mendatory which are ground and VCC. Simulating VCORE is not easy sadly. You cant just inject voltage like on other voltage lines. There are a lot of signals comng into the PWM controller and also some going out!
It's enough to overdrive the Enable signal. Then you should be able to measure the output voltage of the power stage in question. Also you can check for activity of the PWM signals and for the voltage feedback signal. Usually 0.8 Volt.
Please be aware of, that there are single-, dual-, triple and quad logic gates ICs possibly in use. But also transistor- and diode-logic. So I suggest to think more in a "semiconductor" manner, not only for each function. If the semiconductor behaviour is as expected, the function will be also as expected.
hello i watched your video about the broken asrock board from ebay and wanted to ask: I have ASRock b450 pro4 r2.0 does the fix you did apply to my motherboard
If a crystal "Oscillator" isn't outputting a wave form and it's getting it's required Voltages, The lil sh't needs to be replaced. Mind you 1st check if all capacitors supporting the sub circuit have not shorted.
Sehr gut gesehen! War tatsächlich in dem Moment ein Fehler von mir, den ich aber off screen bereits behoben hatte! Hatte tasächlich einen unterschied gemacht, aber den sieht man bereits im Video! Aber vielen dank!
Hi, where are you located, my Wife's B550 Pro4 died. It has only 70 Ohm between power and 3,3V which seems quite low, but idk how to find the defective component, no damage visible and nothing getting irregularly warm
i have b450f and similar am4 board have PCH resistance 17-25 ohm, working properly. please check that PCH line, if under 10 ohm it's very likely dead PCH.
Thats why you should deny previously repaired boards. The rabbit hole you get into (time) and the price of the board never make economical sense. Only thing you get out of it is more knowledge and share what you figured out.
First, it's asus, second, it's amd, third the solution is to check the enables and signals from the schematic, it could be a component open or shorted, don't assume or take anything for granted, verify it and follow the scientific method.
Don't worry, I think each of us has failures in repairs but that's the only way we can learn.
Thank you for the videos and good luck in the future.
The true value is the content and the learning you get out of these. If you fail to repair a board, you get a supply of spare parts from it which is very useful for future jobs.
I suspect that repairs have been attempted on the majority of faulty mainboards you find in places like ebay, particularly if they come in lots. The everyday user with a board that has failed does not know which part is faulty and generally won't take the system apart to sell individual components. These are most often prebuilts. Such a user will throw it out, or send the system in for repairs.
Some repair centers will do component-level work but more often they will swap the board for a working one. Bad boards may get sent or sold to a facility that does component-level work. Boards they can't fix may be resold individually or in lots. Unless you work in such a repair center, expect tons of boards with failed repairs.
Yes the experience that i get on the way is very valueable, but its just frustrating not getting a result!
14:46 WinRAR - interesting. I've not installed WinRAR for about 20 years - 7zip having replaced it entirely.
27:00 Would it be possible for you to share the picture/the source of the AM4 power sequence in the video description?
Its from a russian telegram group! There is not more to it then the picture itself!
Usually the Oszillator is used in a "free-running" mode, this means the Oszillator shall start oszillating by itself as soon as the voltage is applied.
Hey i really Like your videos and i tryed to repair a board alswall but unfortunatly without succes
When reflashing, I use a clamp, I remove the cpu and ram.
If it fails usually it's because a cpu is installed. Ram rarely so.
You get different results, if the CPU is still in the system when doing in curcuit flashing? I need to check that for myself then!
@@MainboardMedic I flashed BIOSes for 3xx and 4xx series (best was a MSI x470), Asus MSI Gigabyte.
So far, from this tiny experience (20 boards ca.), if a flash fails it's always because of an installed Cpu. If it still fails I remove the Ram. Now if it fails with cpu on I remove cpu + ram immediately. My guess is that 1.8V goes charging some caps and or is sunk by the cpu making under 1.8 or diminishing the total Amps. These are DC +V, maybe a tiny bit makes a difference.
I'm no technician, just observations.
@@Neksus-M06 I do need to test that! Thank you very much for sharing your experience!
Non-powered semiconductors can put some heavy load onto all connected data signals. With the BIOS EEPROM, it shares a 4-bit SPI bus. Also it's important not to use more pull-up resistors as the specification allows. If the board contains the pull-up resistors and the programmer also contains the pull-up resistors, but it's used as a in-circuit programmer, then the programming might be unsuccessful.
@@jurgenkruger3932 This makes sense, thanks for the technical explanation.
When u mention the flux stinks bad, so what flux do u use regularly? ✌️👍
Its a german brand called Martin Flux, dont thinl you can get it outside off germany!
@MainboardMedic thanks. I will take a look into that.
You can take-off the upper half of the CPU-socket to make a thorougly inspection of all solder-joints under the microscope.
It's also important to remove the remaining flux.
Can you put an artificial signal to the PWM IC to get the VRM working.
I was thinking a while back about all the checking AND gate transistors on a MB, and don't all 3 pins have to be high for an OK signal, ie, if pin 1 and 2 is high turn on pin 3. Like one high and the next low to turn on doesn't make sense. So you could check them quickly looking for 3 highs.
Not to sure if you are mixing 3 pin mosfets and AND gates here. There are many 3 pin sot-23 mosfets that act as pulling low to ground resistors. Meaning only 1 pin the gate is high and the other 2 are low, which is the normal operating procedure.
The AND gates are most off the time DUAL AND gates which means they have 2 Inputs that need to be high in order for the output to be pulled to VCC. So There are 2 more pins that are mendatory which are ground and VCC.
Simulating VCORE is not easy sadly. You cant just inject voltage like on other voltage lines. There are a lot of signals comng into the PWM controller and also some going out!
It's enough to overdrive the Enable signal. Then you should be able to measure the output voltage of the power stage in question. Also you can check for activity of the PWM signals and for the voltage feedback signal. Usually 0.8 Volt.
Please be aware of, that there are single-, dual-, triple and quad logic gates ICs possibly in use. But also transistor- and diode-logic. So I suggest to think more in a "semiconductor" manner, not only for each function. If the semiconductor behaviour is as expected, the function will be also as expected.
hello i watched your video about the broken asrock board from ebay and wanted to ask: I have ASRock b450 pro4 r2.0 does the fix you did apply to my motherboard
on some other channel, cant remember which tho, or if it was same board, the SIO was dead and after replacing the board shined
Usually the off-board pre-programmed controllers are marked with a colored dot.
CPU debug sockets doesn't been a good thing for u?
If a crystal "Oscillator" isn't outputting a wave form and it's getting it's required Voltages,
The lil sh't needs to be replaced. Mind you 1st check if all capacitors supporting the sub
circuit have not shorted.
Bin mir zu 100% sicher das du das Falsche Bios ausgewählt hast.
bei 13:33 wählst du nen Z390 Bios aus.
No bios is perfect
Sehr gut gesehen! War tatsächlich in dem Moment ein Fehler von mir, den ich aber off screen bereits behoben hatte! Hatte tasächlich einen unterschied gemacht, aber den sieht man bereits im Video! Aber vielen dank!
@@MainboardMedickein Ding :)
Hi,
where are you located, my Wife's B550 Pro4 died.
It has only 70 Ohm between power and 3,3V which seems quite low, but idk how to find the defective component, no damage visible and nothing getting irregularly warm
I am located in germany!
Sent me an Email if you are interested in sending it over!
@MainboardMedic
Wo finde ich deine Mail Adresse?
i have b450f and similar am4 board have PCH resistance 17-25 ohm, working properly. please check that PCH line, if under 10 ohm it's very likely dead PCH.
I have same problems with Asus prime a320m-k r2.0
CPU socket replacements are not easy.
Thats why you should deny previously repaired boards. The rabbit hole you get into (time) and the price of the board never make economical sense. Only thing you get out of it is more knowledge and share what you figured out.
Jap i really should do that!
Agreed, but makes for a good video 🎉
First, it's asus, second, it's amd, third the solution is to check the enables and signals from the schematic, it could be a component open or shorted, don't assume or take anything for granted, verify it and follow the scientific method.