Two Mini PCs that Won't POST - Let's Fix Them!

this video is sponsored by PCB way hello there and welcome back to another video in today's video I've got two broken mini PCS that were very kindly sent to me by fellow YouTuber Carrie Holzman so first off thank you Carrie for making this video possible Carrie reached out to me and offered to send these computers over for repair as he tried to fix them before and was unable to get them working so let's see if we can get these systems working I'm going to start out with this one which is a belink eq12 Mini PC that has an Intel n100 GPU 16 GB of ddr5 RAM and a 500 GB nvme SSD after getting the computer out of the box and plugging in both HDMI and power I pressed the power button which illuminated and the fan spun up however there was no display on the monitor even after a while Carrie told me that this computer had its Ram switched out and after that it never posted again even after reinstalling the original Ram so with the problem confirmed let's start digging around inside this computer to see what's causing it to not post after after removing several screws and two plastic panels I was granted access to the motherboard where the RAM and SSD are both conveniently positioned on the top side it looks like the system is loaded out with a 16 GB stick of crucial brand ddr5 4800 Ram which isn't too shabby if all else fails hopefully that memory stick is still good and I can salvage it Carri went through the standard troubleshooting steps with a system that's exhibiting this problem as he's a very experienced computer technician but I'm going to go through the basic troubleshooting steps that one should do when met with a power but no post or display issue for those of you who may want to learn these steps I took the ram out and put it back into its slot which in turn receips it I then tried to power on the system again but there was still no display now if the system had more than one ram slot I would have tried the dim in all of the different slots to see if the issue revolved a dead memory slot or Channel however the intel n100 only has one ram Channel and this belink computer only has one slot on that channel so I can't do that here I also don't have any other systems that use sodm ddr5 on hand and somewhat for that reason I don't have any spare sodum ddr5 modules that I know work so for the time being we're going to have to assume this Ram stick is fine I also know that Carrie tried a couple different Ram modules when he was diagnosing this computer and none of them worked so presuming this Ram stick to at least not be the primary source of the issue doesn't seem too crazy to me I'll provide links to the videos that carried it on these computers in the description so that you can see what he tried out on them in full detail I then went ahead and removed all of the other parts of the computer such as the m.2 drive Wi-Fi card and the ram just for now I then removed the motherboard from the case and started to look at it to see if there was any noticeable damage to any part of the PCB but it didn't seem like there was I also went ahead and located the seos battery so that I could test whether or not it was dead it is possible that the seos battery was faulty and when the ram was switched out it coincidentally finally failed which is causing this no poost situation these battery terminals are really really tiny and quite hard to work with so sadly I can't use my custom seos B battery tester instead I grabbed my multimeter and with some super fine probes I was able to get a reading on the battery's voltage which is completely good at about 3.05 volts with the seos battery known to be good I left it unplugged from the board for a little while longer to hopefully give the BIOS more of a chance to clear in that time I chose to go ahead and pull off the CPU Cooler for the sake of inspecting the board underneath it and double-checking the thermal paste application thermal past shouldn't be causing this issue with the known circumstance but why not give it a check while investigating the components hidden under the heat sink Additionally the Intel n100 in this system is soldered to the motherboard so a CPU receip isn't something we can try if it was a regular desktop system with a socketed CPU that's another thing that I would have given a shot I pulled off the heat sink by unscrewing its mounting screws in a cross pattern which is extra important in this situation because this cooler is direct dieye it would be much easier to crack the diey by unscrewing it unevenly because it's a direct die cooler so always be careful when dealing with these kinds of heat sinks under the cooler nothing seemed out of the ordinary which was mostly to be expected I applied new paste remounted the cooler assembly plugged our seos battery back in and reinserted the dim I then decided to try powering on the system once more this time with only the bare minimum necessary components on the board I also powered it up outside of the case just being careful not to let anything short out while doing so once more we had a power light and I could hear the fan spinning on the bottom but there was no display I was still pretty unsure of what could be going on here at this time so I went ahead and decided to give the power adapter a check though it might not make the most sense in this case because we know that changing the ram provoked the issue since it's quite easy for me to confirm that the power adapter is good I might as well rule it out to do this I could have used my multimeter to measure the voltage that the adapter outputs under no load and I also could have used my osilloscope to measure whether or not the power supply had an unusual amount of Ripple in it but instead I chose to do something different because I have adjustable bench power supplies I found an old Barrel jack cable that matched the plug of the adapter that the uses I then took that cable and hooked it up to to my bench power supply whose voltage I matched to the ac adapter's voltage and current limit I set to the adapter's rated maximum current I then double checked that the voltage coming out of the barrel jack cable I had connected to my power supply was correct and that it was the right polarity with my multimeter and powered the beelink up using it the benefit of doing this is that I not only now know that the voltage being fed to the is clean but I also get to watch the amount of current that it's drawing when it's booting up there's a very low chance that this computer has a direct short in it based on the fact that it mostly spins to life minus the display of course but the power Supply's current readings should confirm this with the beelink powered up and doing whatever it was doing I watched the current measurement on the power supply I didn't put my camera on it for whatever reason but the was drawing about 1 amp from the power supply which at 12 volt seems like a normal amount of power to draw in startup also the belink AC adapter is rated to do 3 amps so it was only drawing 1/3 of what the spec power adapter can provide further making this current draw unsuspicious so with this now known I'm confident that the issue resides on the motherboard board I'm not a trained electronics repair technician and I've not done much regarding motherboards and their inner workings at all I designed and build my own circuits from scratch but understanding someone else's circuits with no reference materials and figuring out how to fix them is a different Beast altogether in any case repairing this board is absolutely worth a shot so I went ahead and tried something that made sense considering the circumstances replacing the ram module obviously requires putting some force on the sodm slot and I wondered whether or not that Force might have been the final straw for an already weakened solder joint on the slot so I proceeded to do my best to Reflow the ram slot with my soldering iron and way too much flux it was an ugly job at best but I was confident that none of the pins were shorted and that the board shouldn't be in bad shape though it was quite messy sadly though this reflowing of the sodum socket yield did no results and the system continued to power on but refused to post at that time I wasn't sure what to do and so I moved on to the next computer which I'm going to get to here soon but there is more on the belink that still needs to be shown in the following few weeks I I researched a little bit on motherboards and motherboard repair mostly looking at laptop motherboards because there isn't much repair information on these mini PCS yet and these Mini PC boards are very similar to laptop motherboards I also managed to acquire some new soldering gear and a microscope for some other projects but even though the gear doesn't make the difference and instead the person operating it does I felt a lot more confident about messing with these Mini PC boards so let's take a closer look at the I started by figuring out how the main power supplies in the worked the 12volt comes in from the adapter at the barrel Jack and then appears to Branch out into these two different DC todc converters with the belink motherboard out on my bench I plugged it into power and turned it on with the beelink turned on I used my multimeter to check the voltages on the board the 12 volts made it into the motherboard and then it seemed to go into a 5volt regulator which was closer to the barrel Jack and after that there was a second converter that was doing 3.3 volts I then went off on a bit of a tangent with my thought process and tried to reinforce the soldering on the two HDMI ports for some reason I'll chalk this one up to me just not knowing exactly what to do next but needless to say messing with the bad looking solder joints on these HDMI ports wasn't helpful and the system still didn't post while trying to post the board after messing with the HDMI port I left the CPU fan off and felt the heat sink to see whether or not the computer's CPU was getting warm at all this would indicate whether the CPU was getting power or not and I mainly decided to gauge whether vcor was present or not through this method rather than measuring the power phases for a decent reason the power phases are hard to measure with the heat sink right next to them and this is a direct die cooler I can't power the system up even for a short time with no heat sink on without immediately overheating the CPU with a CPU that has an IHS you can power the system on without a heat sink attached for a couple seconds to get some measurements but you can't easily do that on a direct die cooled CPU the CPU was getting really quite warm so it was definitely getting voltage at this point I didn't really know what to do with the system as I just didn't understand where to go next so I decided to pull out the ddr5 sodum module and the 500 GB nvme SS and at least for now call the belink system a loss I do have an update on the belink system at the end of this video though because I learned another detail that I originally missed and did some more probing on the board off camera but before we get to that let's move on to the other system because we've been working on this belink for quite some time now the other Mini PC that Carrie gave me is this Zulu xr1 Max Mini PC and this is a pretty darn nice computer it's got a ryzen 7 5800 U with eight cores and 16 threads 16 GB of ddr4 RAM and and a 1 TB nvme SSD I'm really hoping to get this system fixed as I think it's a pretty awesome little Powerhouse of a computer I'll start out with the Zulu in the same way that I did the belink by going through the standard troubleshooting steps for the issue that it's exhibiting I connected the Zulu to power and HDMI and hit the power button the system spun right up and the small OLED screen in the front of it illuminated I let the system sit for a bit but it did the exact same thing as the it didn't provide any display to the monitor I know for a fact that the monitor was on the HDMI cable was good and that the input was set correctly too so it's definitely another case of a computer with a power but no post problem however this Zulu has that little OLED display which is showing me just how much life this system still has in it the fan speed percentage and the temperature are both updating and the values they're outputting are believable so it's clear that the system is alive and doing something just not providing a display I think that the system might be fixable which is something I'm very glad to see I went ahead and pulled off the top and back panels of the little computer so that I could pop the main assembly out of the casing with with the computer part of the computer out of its case we can see that this thing is constructed in a bit of an unconventional way there are two boards the main motherboard and an iio board which also has the nvme drive on it these boards are stacked on top of each other then there are four ribbon cables coming up from the io board which plug into the motherboard in the video Carri did on this computer he mentions how annoying this construction can be and I totally agree working on this system definitely wasn't my favorite repair job ever because man those ribbon cables are really annoying in any case with the motherboard part removed from the rest of the system I began to inspect it we've got a single 16 GB stick of ddr4 3200 MHz Ram in the system which is actually really convenient I have some known working ddr4 sodum modules in my Lenovo think Center tiny so I can test some known good Ram modules in here to see if it's an issue with either the ram stick or one of the two sodm slots I also quickly took the sodm from the Zulu and put it into the Lenovo which posted using it with no issue so I can confirm that that Ram stick is good I'm also going to pretty much skip over the testing of the other Ram and the Zulu and the testing of both the Zulu's slots because it yielded no results and the Machine still didn't post plus I'm sure you all know how to change out a sodum module and then power on a computer I also went ahead and pulled out the seos battery both in the hopes of clearing the seos settings and also so that I could give it a test to ensure the voltage is correct my polarizer made it pretty hard to see and the reading only popped up for a split second because it was really hard to get good contact with the pins on this cosos battery but it measured at 3.04 volts which is a good voltage I put the computer back together and tried to post it with the the newly reset seos but there was still no display coming out of this PC I then let a little bit of time pass and did my bit of research on motherboards and repairing them and now is when things are going to start to get really interesting because this motherboard has a lot of pretty standard looking chips and circuitry chips and circuitry I was able to find some general troubleshooting information on but quickly before I hop over to the microscope and look at this board I'd like to thank the sponsor of today's video PCB way PCB way is a company who offers several different highquality Custom Manufacturing Services chances are if you're someone who enjoys poking around on pcbs and repairing them you might be interested in creating some pcbs of your own if that's the case PCB way is Affordable and highquality PCB manufacturing service is something that you should absolutely check out I've been using their PCB manufacturing service for a bunch of my projects and I've been nothing but pleased with the results check out PCB way and their highquality Manufacturing Services at the link in the description all right back to the Mini PC with my new soldering gear and a small bit of newly acquired knowledge I pulled the Zulu back out and started to do some deeper troubleshooting I powered it on quickly to ensure that the issue was still persisting as sometimes computers will do that weird thing where they're broken you change nothing but putting them on a shelf for a few weeks pull them out again and they work just fine however this time that wasn't the case and the system still powered on but wouldn't post the OLED display was displaying the fan speed and CPU temperatures still and was also actively updating which is a good sign for this system's repairability I went ahead and pulled out the motherboard once more and began inspecting it in detail one other thing that I allowed myself to do as well was removing the CPU Cooler as the Big Slab of copper that Zulu used while a very high quality feeling cooler covers a lot of the motherboard's probable components removing the CPU heat sink reveals a pretty high number of chokes for Mini PC such as this one all of these chokes are usually associated with different power supply circuits that convert the 19 volts main input into voltages that the internal components can actually use just as with the though I cannot power on this system to probe all of these chokes with the multimeter this is because the CPU heat sink has to be off in order to reach these components and powering on a system like this with the cpu's die entirely bare is a very bad idea thankfully after digging around in my heat sinks for projects box I pulled out this little one that was previously from a Raspberry Pi and it fits this diey beautifully with a little added thermal paste it sat very nicely on top of the die and allowed me to probe these voltages now this heat sink is not capable of cooling this ryzen 7 at all and I'm only using it in short segments of having the system powered on to buy me enough time to get some voltage readings 19.5 volts was present at the input of the system system and a lot of the voltage rails were off while it wasn't turned on which isn't surprising however one of the green inductors next to the two ribbon cable connectors was outputting about 5 volts presumably for USB that's powered through the ribbon cables the other green inductor was seemingly a 3.3 volt rail which works with a lot of the system but also likely the nvme SSD that runs primarily on 3.3 volts this also explains its positioning next to two more of the board interconnect cables with the system still off I was also able to locate the 1.8 volt rail and a rail that was outputting about 0.76 volts as well I then hit the system's power button and with the system powered on as is evident by the CPU fan spinning I measured the chokes that should be on the V core rail it started out on 1.1 volts and then climbed to 1.3 Vols which seems like relatively normal behavior for a ryzen v core rail during a post after letting the system cool down for a minute I powered it on again and checked another choke which was another 0.76 volt rail after that I checked one that was located very close to the ram slot assuming it would be 1.2 volts for the ddr4 memory which it was at this point I was relatively confident that the voltages across the board were all good and that we had no power issues at least not at the power supply parts now where should I go from here well I found this video from the YouTube channel lapix which goes over some basic laptop motherboard troubleshooting and this video has been immensely helpful so check it out at the link in the description in short the video said to check voltages and if those are good we can check What's called the startup chip which is also known as the super iio chip the important thing to check for here is that the super. chip is getting power which should be 3.3 volts and this can generally be checked by just probing a few of the capacitors around it which is what I did I replaced the cpu's heat sink and fan assembly and powered on the system while checking some voltages on the super. Chip's capacitors we were getting 3.3 volts on most of them and not all of the capacitors are between 3.3 volts and ground so it's very likely that the power to this chip was fine while I was probing around this area I discovered something that looked unusual and would prove to be a very crucial part of this pair an area that looked like it had either been blown partially or been hand soldered and had some old flux left behind here you can see the spot that I saw and needed to troubleshoot there was a capacitor and a resistor right next to each other and it looked as if the resistor had either gotten really hot or been hand soldered I wasn't sure yet but whatever was going on here didn't seem quite right I did my best to measure the resistance of the resistor while it was in the circuit which never yields highly accurate results because the resistance measurement will catch other things on the board than this resistor but I wanted to see if I could find any place to start from finding schematics for this computer was never going to happen so I had to do my best to reverse engineer from what I've got here which looked almost like a failed resistor I didn't get a shot of it but the resistance measurement on my multimeter did some weird things while measuring this resistor it would constantly switch between o something random in the mega ohms range and then about 9.3 kiloohms so I decided to desolder this resistor clean up the area around it with some alcohol and lastly replace it with a 10 kohm resistor as a starting point some isopropyl cleaned up the spot l so it didn't seem like anything had burned and it was just old flux that was giving the area that burned color with the old resistor removed I had to find a way to get the temporary replacement into the circuit I don't have SMD resistors on hand yet I've got some coming now but at this time all I had were quarter watt through hole resistors which are not quite the right size as you can see but nevertheless I managed to find a way to make the 10K resistor fit onto the motherboard in a way that it could still be attached to the case this resistor was never going to be a permanent solution and was just a quick value test to see if 10K was right or not if this resistor worked I'd order the right part and install that at a later time now that the 10K resistor was installed and the system was partially reassembled I plugged in power and hit the power button the fan twitched and then nothing else happened holding the power button down kept the fan spinning though it almost seemed like the circuit that I had been messing with was some kind of latch with the power state from the power button and 10K was too high resistance to hit the switch and have the on state latch into place and I use air quotes here because this is almost certainly not quite what's happening but you can probably see what I mean and where I'm coming from I then decided to try and measure the resistance of the resistor that I had removed from the board and it measured at 0.2 ohms quite odd I was also certain that this wasn't the metal bed of the microscope by the way as it's powder coated and the multimeter read it as o when directly measuring it so it's not a short from the bed of the microscope all right so clearly 10K is too high and 0.2 from that resistor seemed to have some issues so how about a value of a lot lower than 10K something more like 1K I went through the same install process as I did with the 10K resistor but this time with a 1K resistor and it exhibited similar Behavior to what the 10K resistor did I'll also quickly mention here that I measured the resistance between the pads of this resistor we're dealing with and the power buttons terminals they were connected though through two very high resistance Pathways one about 500 kohms and the other about 36 khms as a small experiment I decided to try out shorting the pads of the resistor Al together with some solder it could have been possible that the 2 ohm tiny resistor was good but the solder joints or the old flux was causing it the issues I shorted the resistor pads or what was left of them because I broke one of them off and tried powering on the system I was not ready for what was about to happen as is evident by the camera being solely framed on the PC and not the monitor at all but after pressing the power button and waiting a few seconds with the solder Bridge the Mini PC posted for the first time at least in my hands it also went ahead and booted into the windows 11 install that had been on it when Carrie last used it so I was quite confident that this little computer was fixed I won't spoil it too much here but as I was messing with this PC later I noticed that it has another issue that needs to be addressed that will be a follow-up video because this one is already absurdly long I'm on page 10 of this script this is insane in any case at least one issue with the Zulu has been fixed and soon I'll have another video working on the other issue which is storage related now if you're still here I've got to say that that is awesome listening to me ramble for this song straight is is impressive but earlier I said that I had an update on the belink so I'll just let live audio me take it from here on that all right so it's been a little bit since I first started working on this belink and after doing the stuff on the Zulu I decided to come back to this after talking a little bit with Carrie about what exactly broke this so I originally thought it had broken just after a regular Ram swap just taking out the old RAM and putting the new Ram in however it turns out that there's actually more to it than that the thing that most likely caused it to be broken by that procedure was the fact that Carrie left it plugged in while he did the ram swap so that explains the fact that the ram swap kind of messed it up it makes it a lot more reasonable uh originally I was thinking that just the force of inserting and removing the ram might have messed up the socket which is why my first instinct was hey let's Reflow some of these pins see if it broke a faulty solder joint but that was not the case and I don't know I don't feel like I did the most thorough job at diagnosing this past that because I just wasn't sure where to go but I ended up picking it up again off camera and messing with it for a couple hours I'm not going to go all the way through everything that I did and found out but I want to update and talk a little more about this because I now have a theory on what went wrong with this system and what's broken on it right now sadly the thing that's broken is something that I don't think I can replace it's definitely something that is not worthwhile to replace but I want to go through my thought process because I want to know what some of you think I'm certain some of you are much more advanced electrical technicians than I am and I'd like to hear your thoughts on on what I have theorized based on my findings with this uh with this motherboard I still don't have any Ram in here uh mainly because I I think that this computer is going to end up being a lost cause at least my theory definitely puts it in that category but then again my theory could always be wrong so we're going to find out why we're going to hang on to this guy for a while so this board is kind of an interesting board it uh ites doesn't have a massive big old giant it startup chip like the Zulu one did but it does have one and it's just kind of Hidden Away right here underneath the seamos battery so I located that and I messed around a bit on that I found a component that looked a little weird I actually found a schematic for kind of the application here uh of this exact startup chip and I located that component and it was basically just another zero ohm resistor at least I believe so I tried jumping the contact with solder but that didn't change anything the main thing that I did was I actually went and I measured the voltage on all four of these power stages so I originally thought that you know these are probably all four for the CPU which arguably doesn't make the most sense because the n100 in here is a very low powerered CPU it probably doesn't need four power stages in any case I didn't really check all of them I just assumed that number one they were all for the CPU and number two that they were all functional because of how warm the CPU got this CPU was putting out a reasonable amount of heat so it was very much alive and active doing something again it was clearly getting power so I wasn't too concerned about the uh V core stages however I went ahead and I measured the output voltage of all of these this one right here uh doesn't seem to be related to CPU power much at all it might power something on the CPU but this just uh seems to be the systems kind of board wide 1.8 volt rail very common rail the Zulu had it I know a lot of graphics cards have 1.8 volts on them so this this phase right here was kind of just the the 1.8 volt Rail and then these two phases right here these two guys right here are both working together to produce about 0.85 volts which I believe to be VC cor they are connected together the resistance from the output to ground is about 15 HMS and they're producing about 0.85 volts when the system is on and just not displaying so I'm pretty confident these are vcor the n100 runs at a very low clock speed and it's an alder Lake chip which is a relatively uh efficient architecture so it doesn't need tons of voltage it's not quite efficient as some others so 085 volts to get the clock speeds the n100 is doing on the four cores that the n100 has I think it's four cores anyway 85 volts it's got to be vcor I'm pretty confident of that plus it's the only voltage that had two phases dedicated to making it working together on that so clearly I think vcor and then there's this one down here which is a little bit different this one's actually got his driver circuitry right next to the mosfet these two have the driver circuitry on the back of this board this one is a little bit of a different phase here and this one was not outputting any voltage at all straight zero volts I thought that was very very weird I thought that this might be the reason that we weren't working that it just wasn't functioning reason that it wouldn't post however I think this power stage is actually working fine and I'll get into that a little more in just a second by the way I hope this audio is going to be okay I've tried something different rather than using my cheap lav mic I've got a uh overhead boomed small diaphragm condenser but yeah this phase wasn't outputting any power at all and so I traced the circuitry for these three phases cuz this one's entirely separate this guy's doing its own thing this is just the 1.8 volt rail these three I traced them back through the drivers all the way over here to that's as much Zoom as I've got this chip right here this chip is our main power control chip this controls the vrms for the CPU and such and over here you can see the two driver circuits for our for the the two phases that power VOR and then we've got this guy and some bigger capacitors on the undersides of the phases but this right here this chip I looked up the model number let me see what it is again is an RT 3624 be and I looked up that number and I found on the manufacturer's website I believe I believe it was the manufacturer's website I found some data on the Chip And I discovered something very interesting this supports two different voltage rails uh on its output so it's got it can support up to four phases on one rail and two phases on another and so I did some digging with these pins and a multimeter and these driver circuits and I found out that the two phases that I suspect to be VC cor were connected to two of the phase outputs on the main rail that this supports and then the other one was connected to the other rail the auxiliary Rail and after doing a little more research just looking up some stuff trying to find out what was going on here why there was a separate phase that's doing its own thing I discovered that separate phase is likely to be vxg I think it's axg vxg is the integrated Graphics voltage and vxg again I'm just presuming it's vaxg based on the stuff I was able to find is gone so we're not going to get any video because we don't have any integrated Graphics hence the power but no video makes a lot of sense now all right I thought so something's gone wrong with one of our power rails so we're not getting power to the integrated graphics and therefore we're not getting any video out and here's the thing with that theory I don't know if that would make any sense because switching out the ram while the system's powered on can cause a myriad of issues or sorry not powered on plugged in but killing our integrated Graphics power phase is not one that I would immediately expect it just doesn't make sense how does changing your RAM out while the system's plugged into Power kill the phase that controls the graphics power that just doesn't make sense so I actually think that the phases are working fine plus another thing I probed the output uh I probed the gate signal and the outputs from the from the controller and these two for the V cor we're getting about a 650 khz signal with about a 90% positive duty cycle on their Gates I don't know if these are n Channel or P Channel mosfets I'd assume they're n Channel because that's generally more efficient so I'd assume these are running at about 90% duty cycle and then on the output of the controller was a 10 is% positive duty cycle so these drivers are clearly inverting drivers they invert the signal from the from the controller and they send it to the mosfets as an inverted signal when I probed the output of the controller for this rail I discovered that it was pinned to 100% positive hence this mosfet was getting a 100% or sorry 0% positive duty cycle meaning that it should be entirely off and actually that means that these are nend Channel because if the voltage is low on the gate it should be off which it was because we weren't getting any voltage so it's clear that the control chip isn't even trying to tell that power phase to do anything in fact it's telling the power phase to not do anything and I think that seems to be a conscious command from this controller I think that this controller is not being told to turn on the igpu voltage Rail and this is where my theory starts to get more uncertain for me personally I'm not very experienced with the exact way that post Works in these computers I I not sure exactly about the order but I have some ideas and I'm just going to share them here I could be wrong so let me know what you think if you think you know what it actually is I think the post in these computers when we turn it on and it starts T testing things I think that it tests the CPU first so we have VC cor of course so that we can test the CPU and once we test the CPU I would assume we move on to system memory after the CPU because I mean it's one of the key components it should be one of the next logical steps and then after memory I think then we move on to the G to the igpu I'm pretty sure that there's other things thrown in there along the way but the important thing is I think it goes CPU memory igpu and I think the memory is failing I think the memory is failing the post test and therefore the BIOS is not telling our startup chip or super IO chip if you want to call it that to turn on through this controller the rail that's providing voltage to the igpu and so we never get any video the best explanation I have for why the memory would be failing is I think that when the ram was switched out with power engaged I think that the memory controller on the CPU got killed so I think the memory controller on the CPU is gone and because of that I think we're not making it past um memory checks in fact actually now that I'm talking about it we might not even be making it past CPU checks so now that I've kind of explained that that's my main Theory I think that when the ram was switched out with power on I think the cpu's memory controller got killed and we when we boot the computer it runs its post tests and either the CPU or the memory I think it's more likely that it would catch it during the memory tests is failing and because of the fact that the igpu relies so heavily on system memory to work I would expect that we test the igpu after CPU and memory pass if one of those fails I don't think we ever get to the point where we test the igpu I think the post hangs there again I'm not 100% sure on that but I think that's what's happening and because the post is hanging there I think that the super iio chip here is never telling our power chip here to turn on the phase for the igpu because the super. chip on this board does seem to handle some things regarding power and power phasing and it I believe I believe that it talks to the vrm controller and therefore I think that the is dead because of a dead CPU and number one I am not skilled enough at SMD soldering uh and really this would be hot air rework soldering to solder a new CPU onto this board and number two the only replacement n100 chip that I could find was $225 and a new one of these systems with 16 gigs of ddr5 and a 500 gig SSD costs $250 on Amazon so I think that this system is as sad as it is a goner I don't think that there's anything left that can be done about this system I don't think we can fix it again uh if you think that my theory is wrong and have another thing for me to try I will happily give it a shot because I mean why not we don't really have anything to lose it's already broken might as well try but I think I'm relatively confident in that theory being at least mostly correct and because of that I don't think that this board will be coming back because replacing the CPU is just not economical at all but yeah that is the update that I have on the belink I played around with it for a while just discovered some things I didn't bring the camera out cuz I didn't feel like filming uh at that time of night and I didn't think I'd get it fixed anyway and I didn't but I just wanted to put that in here because I feel like originally I didn't do the most thorough job at working on the and I changed that I've now really messed with this board a lot and I've learned a lot more about how the power stuff works and the interface of the super IO chip so I think the CPU is dead well that's all that I have for you today one more thanks to PCB way for sponsoring this video and one more thanks to Carrie for sending out these PCS I had a lot of fun with them in any case I hope to see you next time goodbye

Channel: Computer Hardware Tips and Tutorials

Views: 104,852

Keywords:

Id: PYSsdnb5AwQ

Length: 35min 41sec (2141 seconds)

Published: Mon Mar 25 2024