In the future, before recapping, check all the voltage rails as close to the chips as possible and look for noise on the power rails. If theres no noise, the caps arent the problem.
Aluminium electrolytics are "formed" - the forming current grows an oxide layer on one of the plates which is actually the dielectric. Forming thicker oxide gives you a higher breakdown voltage rating, but decreases capacitance. If the capacitor is left a long time with no voltage applied, the caustic electrolyte etches the oxide layer - capacitance increases, but you lose voltage rating.
I've seen exactly that on a socket A board that wasn't turned on for 15+ years before I took out the caps and tested them. They were Nichicon 5V 3300uF ones and looked fine, but they measured ~10000-12000 uF each (good ESR). I thought "thats strange" and put them back into the board. When I turned the board on for the first time, all the caps swell and vented before the POST was even done, live and right in front of my eyes in a matter of seconds. I thought of degradation and increased suface area myself, but nobody wanted to believe me ("no, its excessive voltage loss that simple testers detect as increased capacitance"). But as you see in the vid, that cap had only 2% voltage loss.
I have the same motherboard in my file server ( and I did a partial recap last year ), the peculiar thing about this motherboard is that it only supports older low density ddr3 memory , with more "modern" high density ddr3 ram it doesn't post, so Adam might have actually fixed the board but he used the "wrong" memory sticks to test it.
yup, same with dualcore 1156 cpu's, doesn't like 4 Gigabyte singlesided RAM. The 4GB must be doublesided. I have 2 out of 4 775 mainboard that works till this day. 3 of them are G41 chipset, one has both ddr2 and ddr3 (combo m/b)
I wanted to point out that it may be worth checking to see if the bios reset jumper is not on as this will sometimes result in a blank screen at boot (no bios screen output).
I was recently given an old socket 478 PC that I made into a retro Windows 98 machine. It had blown caps so I swapped them out. Yours certainly came out easier than mine. My board had a huge ground plane that was sucking the heat away. Also managed to get a short somewhere by connecting both legs to said ground plane, but reflowing and removing excess sorted that. Got it working, but you are right it is not worth the repair. I just did it for pleasure. I did an old graphics card not so long back and I would suggest that you can't always rely on the silkscreen to be accurate with the polarity. I hadn't taken a picture and just put them in the way the board suggested. You quickly know when you've done it wrong. All three caps violently turned themselves inside out. What fun.
I have a socket a board with bloated caps, it worked ok at first, it then stopped working. I am going to replace the bad ones. And while it may not be financially worth it to recap boards like these, I think it’s still worth doing.
i work at a repair shop, we dont do motherboard repairs but i want to so i do it as a hobby (or so im learning about it). I really hope what im doing is fine by just finding a confusing question and going down a rabbit hole until one thing can make sense, and then i guess meander my way back up... but I find it much better to watch more mirror-like, or unsuccessful attempts at something that would more closely relate to my own thoughts. Bc usually what they do works and you do the same and it doesnt. Im running on a tangent here, i gotta watch all your vids to hopefully understand somethin in the future. if anyone has good insight of how to get started and gradually get into MO repairs id be glad
Years ago, I had to fix a LGA775 motherboard ASRock 775i65G R2.0 with some bloated caps. It was not booting up. The top caps near cpu seem fine but not the ones around agp slot. It was painful to replace the 3 caps because of the limited space. I unsoldered the legs and put some used ones from dead boards and the board started to work fine again.
Glad you've released the video anyway. You can't always win. It was still entertaining and maybe it'll teach someone who wants to get into board repair to not be discouraged, if something doesn't work out.
Couple of points: 1. great video nonetheless. It's nice to see more "failed" attempts, just for the entertainment factor and to help other techs realize failure is perfectly acceptable, other than giving the same predictable success each time. 2. the symbol for a polarized cap is with a curved bar (like an umbrella) for the negative leg. 3. my 2 cents on the MB: possibly failed clock generator (need scope to check it) or BGA gone bad on NorthBridge (given the memory error code, since the NB is the memory controller, if that's anything to go by). Could try taking the heatsink off the NB and blasting it with hot air just for funzies - nothing to lose at this point. I once got one going by doing this and it did much the same thing: gave me a code on the PCI tester, but nothing else. Cheers
Vloss is related to ability of that capacitor to retain voltage level according to it's rated voltage, sometimes you measure capacitance and is right on point but cap won't be able to hold its voltage and that's shown by vloss BUT the way those Chinese testers find that I usually have doubts because caps have wide range of voltage ratings ie. vloss of 10% on 250V cap and on 6.3V cap will have very different implication considering the way those testers work (not sure but at least I believe they treat all caps the same and not test accordingly to theirs range of voltage ratings), although you can use to compare 2 or more caps of the same ratings to predicts which one has bad vloss and so pick out which is bad. Good video share I always learn a lot from you brother
I was needing to test a few capacitors and totally forgot I had a tester similar to that, thanks for reminding me. VLOSS = Initial capacitor voltage loss immediately after a load pulse due to the energy required to complete the switchover. It's an arbitrary figure. As far as the cap with the exceedingly high capacitance (1200 uf), In a capacitive circuit, when capacitance increases, the capacitive reactance XC decreases which leads to an increase in circuit current. So if you knew where you pulled that one from you may find you have a short or faulty component. Always test all components purchased from ebay prior to use, Thank You for the video...
26:07 Why isn't your tester calibrated? Short the 3 test points together and follow the on-screen instructions. Takes about 30 seconds to do it then you will get more accurate readings. Bear in mind it will ask for a small value non-polarised ceramic capacitor at one point and you need to clamp that in then press the button to continue. Use a common 104 cap (0.1uF). This tester is based on a German open-source design using an Atmel ATMEGA328 microcontroller and it actually a pretty nice piece of test gear for the price.
I think the most important thing with repairing things (regardless of its age and what it is that youre repairing) is that you have absolutely no idea if it will work or not even after you've done the best repairs that you can, especially when it comes to diagnosing electrical goods. There could be a hundred other things wrong with it and its usually the smallest most obscure thing that is never spoke about that plays the tinest of roles that kills the entire system when it goes down or prevents it ever from being powered on again. Nothing ventured, Nothing gained. Chalk it down to a learning experience. Do or DO NOT, there is no TRY
Yup. Diagnostics are often more a matter of "fix it, then try to sell that solution to the client" rather than "figure out what's wrong and offer to fix it for a price".
This was a timely video. I have exactly the same problem with power electrolytics on an old Intel MOBO. The system was behaving really oddly with intermittent boot errors and repeat corruptions of the boot partition on the SSD! Had a look at the board and sure enough I have bulging caps. Think I will replace them and see if that fixes the problem.
These sort of board repair vids are very interesting, well to me anyway! We can all 'fix' stuff by exchanging a dead part, but to actually repair the dead part is the skill!
Take the memory out... please continue part 2.. I would, do BIOS and then remove the Memory cards, clean them a bit, or replace.. if nothing then give up... Oh! was hoping for a part 2, ah well I still enjoy watching, and learning. Thank you Graham.
Halfway through the video I suddenly realised this board looks awefully familiar, turns out I have the exact same board lying around 😂 Graham, do you need a favour? 😉
well done Graham I really enjoyed this video, I am trying to learn about this stuff, I am 70 years of age in bed with me bad back and find your videos very relaxing, keep up the good work, cheers from Ireland
Thanks for this video. I'm very new to doing my own repairs and have not had much luck with most board repairs. It makes me feel better knowing it's not just me that wastes time and money on attempting a repair.
I repaired my old gaming board by replacing a few Capacitors. It worked. Until a few days later, I decided to hook up a harddrive and set up windows 7. The PC refuses to post. I believe something bad had happened to the bios. I cleared the cmos through the jumper, removed the battery, re-seated everything. There are no beep codes, and the button when you press once doesn't turn off the machine. It turns on but doesn't display anything. One thing I noticed which was very strange before it got confused... Was that during booting from the DVD-ROM, the display would go blank, the power button would do nothing even if you held it down, LEDs will still be lit and the CPU fan will still be spinning. Now i need to find a programmer tool that supports that specific bios chip.
Another thing that I think is the cause is that it was saving settings to the bios and the same thing happened whilst it was saving those settings to the bios
It is possible, for example, that particular capacitor that was leaking (3.1 Ohm), as displayed, may have damaged some other chip, by allowing voltage higher than the gate, could handle, and basically cooked, the logical aspect of the chip, and thus the reason it will not Post. And, agree with comments below, sometimes, seeing a repair fail, is interesting, as one that works.
Great as always Graham, while you may not have been successful I still learnt something. I always thought you were supposed to mount everything flush to the board
Depends on the components. Flush to the board is nice and neat, but certain components like electrolytics, crystal oscillators, anything that's not 'solid state' likes to have a tiny gap so it can move or expand ever so slightly.
There are lots of electronics "out there" that have a small space between the component and the printed circuit board. There are also a multitude of electronics with no space between the components and the board.
Thank you very much for this explanation, I find it better than that given in universities. My understanding of the capacitor behavior was a little blurred, now I understand better.
I would highly recommend using a solder slurper (sucker). They will make your caps and other part removal very easy. I started in electrics in 1961, from the old tube TV's to commercial transceivers to CB's. I really like your repair show. You know more about computer repair than most . Note: replacing and re-soldering is easier as well. Most of the old solder is removed.
The cap testing really showed the failure. What you also could have tried is leaving out the graphics card ( although it gave you an E 1 ) and put the monitor on the internal graphics.
The problem with all the 5V rail caps being faulty means that there also was no surge buffer in case the 5V rail or the 5VSB of the PSU went wonky. With the amount of components on those era boards running directly off of 5V that's a concern.
I have seen some dead PSUs beeing started up with a heat from a hairdryer, can you read any differencee between a failed cap and a hot failed cap? Also about the motherboard, the ripple produced from a regulator with a failed capacitor can cause damage to the thing they are powering, guess how I found out.
One advice, not from me but from a french UA-camr (well Peertubeer now), when he have a failing device (he does mainly old hi-fi/vhs and playstation stuff) that doesn't work properly with bad cap, he heat the cap with a hair dryer to make them "work" temporary and see if there is some life, usually, it work really well, if this advice can help you :) You're video are great to watch, i'm an IT student, last year to do, and i really want to do the same as you because i love this, and your video help me to learn a lot for board repair
There *can* be a downside to low ESR capacitors. Some negative feedback loops (some voltage regulators for example, not all, depends on the device in question) depend on the ESR zero in their frequency compensation scheme for some phase boost. These systems might become unstable or suffer poorer step response with low ESR capacitors (due to decreased phase margin).
Interesting. This is advanced stuff. Had never thought about whether having low ESR caps could ever be a 'bad thing'. Operational amplifiers came to mind when you mentioned 'negative feedback loops', (which I recollect only require resistors and ceramic caps)...so this looks like a subject I need to learn more about...fancy stuff
@@minilab9030 Any system that regulates something typically employs negative feedback (When the output is too small it has to be increased and vice versa). Delays (phase shift) in the feedback loop will cause instability if they add up to -360° (because the feedback becomes positive at that point). Therefore such systems have to be designed so that the loop gain is lower then unity at the frequency where the phase shift reaches -360° (in reality even sooner then that, you want some margin, called "phase/gain margin"). These techniques are called "frequency compensation". Because the LC filter in a switching voltage regulator causes a -180° phase shift all by itself, it's properties are very important to the compensation scheme. The capacitor's ESR introduces a zero in the transfer function that can boost the phase back to -90° at higher frequencies. Changing ESR moves this zero. This applies to op-amps aswell, but most well-known opamps are already internally compensated (using a dominant pole, basically they are slowed down to ensure stability). So as long as you only use local feedback and don't do anything "stupid" (such as adding gain in the loop) they should be stable without much effort.
Been there, done that, sorry to see you having the same experience. It sucks when you invest time and usually quality caps that you'll surely need later somewhere... all for nothing. At least this time people learned plenty of stuff. Thanks!
you should check 1.5v ramvcc, ram slot pin no. 51,54,57,60,62 (or on ram vrm coil) 0.75v ram vtt ram slot pin no. 120, 240 3.3v vddspd ram slot pin no. 236 bios cpu cpu socket gmch e1 mostly a bios problem in this motherboard
My guess would be that most likely there is a broken trace on the circuit board. Components usually don't fail unless there has been a power surge or lightning hit.
Welcome to the complicated world computer motherboards. Here anything and everything can fail without any apparent reason. Hell, even a game that was designed to run on a specific graphics card can kill it. These circuits use a combination of hardware and software to work. There are very tight tolerances even on the arrangement of glass fibre bundles on the board. Then there is a very specific timing and sequence in which the board is powered up to running state (S0 state). Then there is BIOS which will go corrupt just by sitting idle for few years. Any deviation in any of the above from its designed parameter will result in board not able to be brought up. And I didn't even mention about dry solder joint or a broken trace yet.
This video reminded me that: I have also changed 6 caps (the others were fine and I was bored) of my ASUS A8N SLI Premium board and it was resurrected.
Should be available from most places, ebay, aliexpress, banggood, etc. Here's an example eBay link: www.ebay.co.uk/itm/234284775298 Only difference is that I've modified mine with a USB recharged lipo, where the base unit is 9v powered.
It’s the chipset, it needs to be reflowed or replaced , I recall well back in 2006 I got three motherboards fixed through gigabytes official repair centere and all of them had the same symptoms!
The board to me is getting enough stable power as it stays on. Normally, you'd get beeps aswell if it was ram or at least something with no ram in at all. I think the board is just bricked. Still, intresting video :)
Answering your questions: Can capacitance go up on a failed capacitor? Answer: Yes if your tester works out the value based on charge time. If the capacitor is electrically leaky it will take longer to charge giving a false reading on the tester. What is V Loss? Answer: Loss of charge over a given time in percentage. Another indication of it being electronically leaky.
I just,as in 15 minutes ago, replaced a 1155 H61 Foxconn motherboard with a 1155 B75 board for sata 3 and USB 3. NVMe slot included, I only run one lower end game, World of Warplanes so it is running quite nice for very little cash.
There are many reasons people want their old computers to be revived. I for one am trying to revive my old P4 864 PERL motherboard after almost 14 years of its demise. I am no motherboard expert, but with lots of effort and time I could bring up a totally dead board back to a stage where it beeps when RAM is removed. If I could revive it, I will be able to run my old DOS and windows games and software in it.
I don't know where all the techs are. But I can't tell you how many of these Asus 1155 boards i saw fail like this. It's an Asus Plague with this. I went through your attempts as well with some of them before realizing this is an Asus plague with many of their 1155 boards, matx especially.
Had a very similar ASUS LGA 1155 board fail this same exact way. I just replaced it but was hoping to see what the issue might have been. Good try anyway.
Thank you for posting it! I'm hooked, you have taught me a lot by your videos! One question though, why don't you ever measure the power over the mosfets/buck converters and various components surrounding them, like you do on laptops. Is it because the error codes lets you narrow down to the right ballpark Again, thank you for making electronics available for us, curious people:-) Måns
Mainly because if you have "fan spin" where the system has turned on, it's probably not a power issue. It could still be a rail that's working well enough for the system to think it's OK, but is low or full of ripple, but yea, we're getting to rabbit holes now.
@@Adamant_IT It's not that unreasonable (but will get you into a rabbit hole for sure), I had old motherboard where the fan turned on no problem but the CPU was stone cold. I never bothered to test it further though considering it was an old Athlon x64.
are you sure that the capacitors where putted in the right polarity ? becouse the minus is white strip also paint in board and needs to corespond with the white line of the capacitor
Voltage rating is the _max_ the cap can handle, so it needs to be equal or higher than the original. Capacitance (Farads) should be the same. Caps have a big tolerance on them though (20% is common) and more often than not, just being in the correct order of magnitude is good enough for a power rail.
1) The labor and shipping cost is way more expensive than the capacitors themselves; therefore, using solid electrolyte capacitors to replace the faulty ones. 2) The main reason for the capacitors to fail - the work load and environment temperature are reaching the design margin; therefore, buying low ESR solid electrolyte capacitors and increase the spec (e.g. 6.3V 820 uF --> 9~12V, 1000~2000 uF) should be able to prevent the similar failure in the long run. 3) Using proper soldier removing tool. 4) Dead CPU won't turn on cooling fan. 5) Buy a PCI/PCIe diagnostic card for PC.
If you have a failure like this a CPU, PSU, SSD and MEM are all OK and recapping doesn't fix then what other components could be at fault. Could it be support chips like the thing under the blue heat sink?
It was probably never the capacitors, i've owned this exact board, and it had heavy use for most of those years, sold it recently and still worked like new, i also own some slightly earlier boards that have run big overclocks for many years, those boards are still fine to this day, i think the capacitor problems go back a few years before this boards time
My uncle has one, LGA 775 but not the exact same one. It's still working fine even though some capacitors are bulging and even leaking / the insides are coming out of the cross top. I've warned him though, that the motherboard is in the verge of dying.
There's a lot of small cues that give it away, but the main one is the shape. Most BIOS chips are either WSON8 (flat rectangle) or SOP8 (square, 4 legs each side) The only other SOP8 packages you're likely to see are mosfets, and you can spot those because they'll be part of a power circuit (and have capacitors, inductors, etc nearby). The last cue is just to check what's written on the chip. There's only a couple of common brands for BIOS ROM chips, such as MXIC, Microchip, Winbond, and they tend to have that brand name written on the chip (unlike mosfets, which only have a part number) so if you see a chip out on its own, and it has winbond written on it, that's your guy. Keep in mind that a lot of laptops have more than one BIOS chip as well though.
I have that problem where I don't know when to give up but I really want to know what is wrong with the board! How about testing for chipset / rail voltages voltages? If all looks good there then it's cool to stop... pretty please!?
No plans to come back to this for now, but yea, my next guess is Northbridge issues. Likely a dead chip or back solder ball though. It was producing some heat, so it's getting power.
for the first clean the both dim slot with alcohol or thinner and then put in memory again turn on power motherboard if still error accour then Please check with selector multimeter countinuity on dim voltage, usualy the both dim voltage are broken connection, so just put wire on pin 3.3 or 3.7 volt on dim1 slot in the back motherboard to dim2 , its work with me, or , may be can help
Surface mount 'solid state' electrolytics fail more often then you might think. Whereas through-hole electrolytics sometimes offer visual clues to their failure (bulged domes and leaking electrolyte), the surface mount solid state electrolytics are trickier to diagnose (because no visual clues present) and replace (because they muist be removed by careful clockwise or anti-clockwise rotation so as not to rip away the tissue-paper delicate copper pads on which they are surface mounted). If the board refuses to work after replacing the through-hole electrolytics, I would absolutely remove (with great care - see Mr Carlson's lab video on surface mount/solid state cap removal) the surface mount/solid state electrolytics, before testing them with an ESR meter. Then replace with an equivalent solid state...or indeed a through-hole....and test the board again. I much prefer repairing electronics which do NOT have surface mount electrolytics, because two many times, I have used pliers to twist off solid state caps, only to discover I have ripped away the delicate copper pads x 2 connecting them to the board. Repairing this, in a two-layer boards, is a fucking nightmare....hours of avoidable labour...hence enormous care is required when removing them, and of course, they cannot be tested for ESR in circuit, which means you HAVE to remove them to test for ESR, and cannot simply remove 'one leg' before testing, because there is not enough space to remove only one leg before testing.
Done a few full board re-caps in my day, all completely successful, except that one time where it's never the CPU, except where it's the CPU 🤣 No joke, after a re-cap, still had a toasty CPU, but the entire thing started with a PSU that came 1 step from a firecracker so it had it coming. New PSU+new CPU+ re-cap and it was back on the road. Only thing I was curious with your Asus board here is if it acted the same way with or without any mem modules. I have seen a post issue like that previously, and being an extremely early ddr3 board, it's QVL would be quite limited and without post, it would be impossible to tell what version of BIOS was on there, but it looks like you went to the trouble of pulling the chip completely and reflashing it, so I'm just assuming you used the very latest firmware available for that Mobo.... Too bad this repair was a bust, but I used to love doing re-cap jobs back in the day.. I still enjoy picking up a 75in 4 year old 4k smart Vizio TV for free off Craigslist and re-capping the PSU for under $2 just to cure the itch, and of course to hoard 75in TVs and put them in every room in and outside my house that I spent less than $10 on all of them total🤣 Totally love your channel, learned so much from you Graham! Thanks for all you do and keep up the fun good work!
Have you tried cleaning the PCI-E slot that could be a reason why you don't get a display on the GPU because of corroded pins in the slot fixed it with contact cleaner of WD-40 and a wire brush that could fit in between the gap
Hi, I'm struggling to debug this problem on my laptop. As soon as I unplug the charger, the CPU speed drops to 0.48GHz (BD PROCHOT triggered) To make it work again up to its speed (2.2G) I need to unplug charger AND battery, and power up again with either battery or charger. This happens with Linux too, so it's definitely an hardware problem, the BIOS battery is ok. Do you have any suggestions? Thanks!
😂🤣😆I know, Graham "It's never the CPU" Memory error, is it possible you have two different Memory Ram Sticks inserted? Just double check everything.. I could've sworn, you took one out, and inserted a new one. I would have replaced both.. why check one old, and add a new one.. wouldn't that affect the memory value in some way? Just asking, trying to figure this out myself, just an amateur.🤗
I think you should have at least checked some canned caps, expecially the one near the parallel port that is suspiciously tilted. It just drives me nuts how motherboard manufacturers indicate the positive side of caps on the silk screen while every other manufacturer of anything indicate the negative as the cap itself. Whyyyy???
The issue here is not caused by electrolytic capacitors, but rather by low quality capacitors. I have an Hi-Fi system with about 25 years, and all capacitors are electrolytic. There are no bulged capacitors, because they have used high quality capacitors (the unit was manufactured in Japan). These computer motherboards and many other devices (like TVs) have failing capacitors due to the manufacturer cutting costs, by using cheap chinese capacitors.
excellent shows even if the repair did not work you can still learn something. if the board has sentimental value put it in a frame and hang it on the wall like some of your wall art!!!
I've recently got a faulty motherboard (lga1150) I'm getting ready to post-mortem or possibly repair. Used air compressor to clean out daughter's computer like done several times. Next time she booted it started sporadically turning on and off and once in a while makes it into bios before shutting off again(led fans also got brighter sometimes). With the knowledge I've learned from your videos I hope to come to a definitive cause of death.
From my experience with flashing BIOS chips outside the board with a programmer; it sometimes ends up with a different result than the original BIOS of a working board.
Correct, the unique identifiers of the board won't be preserved. Those would need to be taken from the backup and written in manually with a hex editor.
In the future, before recapping, check all the voltage rails as close to the chips as possible and look for noise on the power rails. If theres no noise, the caps arent the problem.
Aluminium electrolytics are "formed" - the forming current grows an oxide layer on one of the plates which is actually the dielectric. Forming thicker oxide gives you a higher breakdown voltage rating, but decreases capacitance. If the capacitor is left a long time with no voltage applied, the caustic electrolyte etches the oxide layer - capacitance increases, but you lose voltage rating.
Something new I learned today.
I've seen exactly that on a socket A board that wasn't turned on for 15+ years before I took out the caps and tested them. They were Nichicon 5V 3300uF ones and looked fine, but they measured ~10000-12000 uF each (good ESR). I thought "thats strange" and put them back into the board. When I turned the board on for the first time, all the caps swell and vented before the POST was even done, live and right in front of my eyes in a matter of seconds. I thought of degradation and increased suface area myself, but nobody wanted to believe me ("no, its excessive voltage loss that simple testers detect as increased capacitance"). But as you see in the vid, that cap had only 2% voltage loss.
I have the same motherboard in my file server ( and I did a partial recap last year ), the peculiar thing about this motherboard is that it only supports older low density ddr3 memory , with more "modern" high density ddr3 ram it doesn't post, so Adam might have actually fixed the board but he used the "wrong" memory sticks to test it.
Not Adam ;-) Graham
yup, same with dualcore 1156 cpu's, doesn't like 4 Gigabyte singlesided RAM. The 4GB must be doublesided.
I have 2 out of 4 775 mainboard that works till this day. 3 of them are G41 chipset, one has both ddr2 and ddr3 (combo m/b)
plus the primary channel on these older motherboards is the one closest to the cpu
I wanted to point out that it may be worth checking to see if the bios reset jumper is not on as this will sometimes result in a blank screen at boot (no bios screen output).
I was recently given an old socket 478 PC that I made into a retro Windows 98 machine. It had blown caps so I swapped them out. Yours certainly came out easier than mine. My board had a huge ground plane that was sucking the heat away. Also managed to get a short somewhere by connecting both legs to said ground plane, but reflowing and removing excess sorted that. Got it working, but you are right it is not worth the repair. I just did it for pleasure.
I did an old graphics card not so long back and I would suggest that you can't always rely on the silkscreen to be accurate with the polarity. I hadn't taken a picture and just put them in the way the board suggested. You quickly know when you've done it wrong. All three caps violently turned themselves inside out. What fun.
I have a socket a board with bloated caps, it worked ok at first, it then stopped working. I am going to replace the bad ones. And while it may not be financially worth it to recap boards like these, I think it’s still worth doing.
i work at a repair shop, we dont do motherboard repairs but i want to so i do it as a hobby (or so im learning about it). I really hope what im doing is fine by just finding a confusing question and going down a rabbit hole until one thing can make sense, and then i guess meander my way back up... but I find it much better to watch more mirror-like, or unsuccessful attempts at something that would more closely relate to my own thoughts. Bc usually what they do works and you do the same and it doesnt. Im running on a tangent here, i gotta watch all your vids to hopefully understand somethin in the future. if anyone has good insight of how to get started and gradually get into MO repairs id be glad
Nice video, one detail though: that symbol you drew is a battery. Polarized capacitors have one bar thicker or curved.
yes, bugged me that did!
Years ago, I had to fix a LGA775 motherboard ASRock 775i65G R2.0 with some bloated caps. It was not booting up. The top caps near cpu seem fine but not the ones around agp slot. It was painful to replace the 3 caps because of the limited space. I unsoldered the legs and put some used ones from dead boards and the board started to work fine again.
P.S. the memory sticks that came with the board look like high density ones ( low density memory sticks usually have chips on both sides )
Definitely glad you decided to show this video! It was fun to watch and failure IS a learning experience just like success! Be well! 👍👍
Glad you've released the video anyway. You can't always win. It was still entertaining and maybe it'll teach someone who wants to get into board repair to not be discouraged, if something doesn't work out.
Couple of points: 1. great video nonetheless. It's nice to see more "failed" attempts, just for the entertainment factor and to help other techs realize failure is perfectly acceptable, other than giving the same predictable success each time. 2. the symbol for a polarized cap is with a curved bar (like an umbrella) for the negative leg. 3. my 2 cents on the MB: possibly failed clock generator (need scope to check it) or BGA gone bad on NorthBridge (given the memory error code, since the NB is the memory controller, if that's anything to go by). Could try taking the heatsink off the NB and blasting it with hot air just for funzies - nothing to lose at this point. I once got one going by doing this and it did much the same thing: gave me a code on the PCI tester, but nothing else. Cheers
Vloss is related to ability of that capacitor to retain voltage level according to it's rated voltage, sometimes you measure capacitance and is right on point but cap won't be able to hold its voltage and that's shown by vloss BUT the way those Chinese testers find that I usually have doubts because caps have wide range of voltage ratings ie. vloss of 10% on 250V cap and on 6.3V cap will have very different implication considering the way those testers work (not sure but at least I believe they treat all caps the same and not test accordingly to theirs range of voltage ratings), although you can use to compare 2 or more caps of the same ratings to predicts which one has bad vloss and so pick out which is bad. Good video share I always learn a lot from you brother
I was needing to test a few capacitors and totally forgot I had a tester similar to that, thanks for reminding me. VLOSS = Initial capacitor voltage loss immediately after a load pulse due to the energy required to complete the switchover. It's an arbitrary figure. As far as the cap with the exceedingly high capacitance (1200 uf), In a capacitive circuit, when capacitance increases, the capacitive reactance XC decreases which leads to an increase in circuit current. So if you knew where you pulled that one from you may find you have a short or faulty component. Always test all components purchased from ebay prior to use, Thank You for the video...
26:07 Why isn't your tester calibrated? Short the 3 test points together and follow the on-screen instructions. Takes about 30 seconds to do it then you will get more accurate readings. Bear in mind it will ask for a small value non-polarised ceramic capacitor at one point and you need to clamp that in then press the button to continue. Use a common 104 cap (0.1uF). This tester is based on a German open-source design using an Atmel ATMEGA328 microcontroller and it actually a pretty nice piece of test gear for the price.
I think the most important thing with repairing things (regardless of its age and what it is that youre repairing) is that you have absolutely no idea if it will work or not even after you've done the best repairs that you can, especially when it comes to diagnosing electrical goods. There could be a hundred other things wrong with it and its usually the smallest most obscure thing that is never spoke about that plays the tinest of roles that kills the entire system when it goes down or prevents it ever from being powered on again.
Nothing ventured, Nothing gained.
Chalk it down to a learning experience. Do or DO NOT, there is no TRY
Yup. Diagnostics are often more a matter of "fix it, then try to sell that solution to the client" rather than "figure out what's wrong and offer to fix it for a price".
@@Adamant_IT yeah, sometimes you don’t know if something is fixable until you fix it
This was a timely video. I have exactly the same problem with power electrolytics on an old Intel MOBO. The system was behaving really oddly with intermittent boot errors and repeat corruptions of the boot partition on the SSD! Had a look at the board and sure enough I have bulging caps. Think I will replace them and see if that fixes the problem.
It's a valid strat! Just didn't work for me this time.
These sort of board repair vids are very interesting, well to me anyway! We can all 'fix' stuff by exchanging a dead part, but to actually repair the dead part is the skill!
Take the memory out... please continue part 2.. I would, do BIOS and then remove the Memory cards, clean them a bit, or replace.. if nothing then give up... Oh! was hoping for a part 2, ah well I still enjoy watching, and learning. Thank you Graham.
Halfway through the video I suddenly realised this board looks awefully familiar, turns out I have the exact same board lying around 😂 Graham, do you need a favour? 😉
Great video. This was not a failure. I learned more about caps, than I already knew, especially, ESR.
I know you have heard this before, but failures teach more than success does. Good job - was worth watching this video.
I always think it's nice to see the unsuccessful repair attempts as well. Still teaches you something.
To do a "Big Clive"... I'll add this expression to my vocabulary 😂
well done Graham I really enjoyed this video, I am trying to learn about this stuff, I am 70 years of age in bed with me bad back and find your videos very relaxing, keep up the good work, cheers from Ireland
Thanks for this video. I'm very new to doing my own repairs and have not had much luck with most board repairs. It makes me feel better knowing it's not just me that wastes time and money on attempting a repair.
I repaired my old gaming board by replacing a few Capacitors. It worked. Until a few days later, I decided to hook up a harddrive and set up windows 7. The PC refuses to post. I believe something bad had happened to the bios. I cleared the cmos through the jumper, removed the battery, re-seated everything. There are no beep codes, and the button when you press once doesn't turn off the machine. It turns on but doesn't display anything. One thing I noticed which was very strange before it got confused... Was that during booting from the DVD-ROM, the display would go blank, the power button would do nothing even if you held it down, LEDs will still be lit and the CPU fan will still be spinning. Now i need to find a programmer tool that supports that specific bios chip.
The chip is by PMC.
Pm49Fl00T-33JC
Another thing that I think is the cause is that it was saving settings to the bios and the same thing happened whilst it was saving those settings to the bios
I recon the bios has got corrupted
It is possible, for example, that particular capacitor that was leaking (3.1 Ohm), as displayed, may have damaged some other chip, by allowing voltage higher than the gate, could handle, and basically cooked, the logical aspect of the chip, and thus the reason it will not Post. And, agree with comments below, sometimes, seeing a repair fail, is interesting, as one that works.
Great as always Graham, while you may not have been successful I still learnt something. I always thought you were supposed to mount everything flush to the board
Depends on the components. Flush to the board is nice and neat, but certain components like electrolytics, crystal oscillators, anything that's not 'solid state' likes to have a tiny gap so it can move or expand ever so slightly.
There are lots of electronics "out there" that have a small space between the component and the printed circuit board. There are also a multitude of electronics with no space between the components and the board.
Thank you very much for this explanation, I find it better than that given in universities. My understanding of the capacitor behavior was a little blurred, now I understand better.
I would highly recommend using a solder slurper (sucker). They will make your caps and other part removal very easy. I started in electrics in 1961, from the old tube TV's to commercial transceivers to CB's. I really like your repair show. You know more about computer repair than most . Note: replacing and re-soldering is easier as well. Most of the old solder is removed.
The cap testing really showed the failure. What you also could have tried is leaving out the graphics card ( although it gave you an E 1 ) and put the monitor on the internal graphics.
The problem with all the 5V rail caps being faulty means that there also was no surge buffer in case the 5V rail or the 5VSB of the PSU went wonky. With the amount of components on those era boards running directly off of 5V that's a concern.
I have seen some dead PSUs beeing started up with a heat from a hairdryer, can you read any differencee between a failed cap and a hot failed cap?
Also about the motherboard, the ripple produced from a regulator with a failed capacitor can cause damage to the thing they are powering, guess how I found out.
That's OK. I'll watch anyway, because I like listening to your voice.
Cheers for trying to fix it Graham, I’ll try and swing by at some point 👍
One advice, not from me but from a french UA-camr (well Peertubeer now), when he have a failing device (he does mainly old hi-fi/vhs and playstation stuff) that doesn't work properly with bad cap, he heat the cap with a hair dryer to make them "work" temporary and see if there is some life, usually, it work really well, if this advice can help you :)
You're video are great to watch, i'm an IT student, last year to do, and i really want to do the same as you because i love this, and your video help me to learn a lot for board repair
There *can* be a downside to low ESR capacitors. Some negative feedback loops (some voltage regulators for example, not all, depends on the device in question) depend on the ESR zero in their frequency compensation scheme for some phase boost. These systems might become unstable or suffer poorer step response with low ESR capacitors (due to decreased phase margin).
Interesting. This is advanced stuff. Had never thought about whether having low ESR caps could ever be a 'bad thing'. Operational amplifiers came to mind when you mentioned 'negative feedback loops', (which I recollect only require resistors and ceramic caps)...so this looks like a subject I need to learn more about...fancy stuff
@@minilab9030 Any system that regulates something typically employs negative feedback (When the output is too small it has to be increased and vice versa). Delays (phase shift) in the feedback loop will cause instability if they add up to -360° (because the feedback becomes positive at that point). Therefore such systems have to be designed so that the loop gain is lower then unity at the frequency where the phase shift reaches -360° (in reality even sooner then that, you want some margin, called "phase/gain margin"). These techniques are called "frequency compensation".
Because the LC filter in a switching voltage regulator causes a -180° phase shift all by itself, it's properties are very important to the compensation scheme. The capacitor's ESR introduces a zero in the transfer function that can boost the phase back to -90° at higher frequencies. Changing ESR moves this zero.
This applies to op-amps aswell, but most well-known opamps are already internally compensated (using a dominant pole, basically they are slowed down to ensure stability). So as long as you only use local feedback and don't do anything "stupid" (such as adding gain in the loop) they should be stable without much effort.
Been there, done that, sorry to see you having the same experience. It sucks when you invest time and usually quality caps that you'll surely need later somewhere... all for nothing. At least this time people learned plenty of stuff. Thanks!
Might be harder to use them, but you can salvage those caps for later. Just with short legs...
The video definitely was worth sharing anyways. Learned quite a few new things about caps :)
Nice recap video Graham,appreciate it!
you should check
1.5v ramvcc, ram slot pin no. 51,54,57,60,62 (or on ram vrm coil)
0.75v ram vtt ram slot pin no. 120, 240
3.3v vddspd ram slot pin no. 236
bios
cpu
cpu socket
gmch
e1 mostly a bios problem in this motherboard
Let's "TRY TO" FIX Computers should be its own category. Always worth watching your videos regardless of the outcome. :)
Whether this video is a flop or not I still learn some useful stuff and love watching it anyways.
My guess would be that most likely there is a broken trace on the circuit board. Components usually don't fail unless there has been a power surge or lightning hit.
Welcome to the complicated world computer motherboards.
Here anything and everything can fail without any apparent reason. Hell, even a game that was designed to run on a specific graphics card can kill it. These circuits use a combination of hardware and software to work. There are very tight tolerances even on the arrangement of glass fibre bundles on the board. Then there is a very specific timing and sequence in which the board is powered up to running state (S0 state). Then there is BIOS which will go corrupt just by sitting idle for few years. Any deviation in any of the above from its designed parameter will result in board not able to be brought up. And I didn't even mention about dry solder joint or a broken trace yet.
well, that looks like the chipset is on the way out. Reflowing the northbridge or something like that seems a bit excessive though.
Do you have a list of the brand/type of test equipment you use in this video?
Did you power on without memory installed? If you have error beeps, suggests EC chip issue. If no beeps bin it :-)
Nice to see the odd failure sometimes failure is more informative to learning :-)
Great video pal carnt win em all 🙂
This video reminded me that: I have also changed 6 caps (the others were fine and I was bored) of my ASUS A8N SLI Premium board and it was resurrected.
Informative video as always. 😊👍
Do you have a link for that transistor/cap/component tester? Is it from aliexpress? Thanks!
Should be available from most places, ebay, aliexpress, banggood, etc. Here's an example eBay link: www.ebay.co.uk/itm/234284775298
Only difference is that I've modified mine with a USB recharged lipo, where the base unit is 9v powered.
It’s the chipset, it needs to be reflowed or replaced , I recall well back in 2006 I got three motherboards fixed through gigabytes official repair centere and all of them had the same symptoms!
The board to me is getting enough stable power as it stays on. Normally, you'd get beeps aswell if it was ram or at least something with no ram in at all. I think the board is just bricked. Still, intresting video :)
Answering your questions: Can capacitance go up on a failed capacitor? Answer: Yes if your tester works out the value based on charge time. If the capacitor is electrically leaky it will take longer to charge giving a false reading on the tester. What is V Loss? Answer: Loss of charge over a given time in percentage. Another indication of it being electronically leaky.
I just,as in 15 minutes ago, replaced a 1155 H61 Foxconn motherboard with a 1155 B75 board for sata 3 and USB 3. NVMe slot included, I only run one lower end game, World of Warplanes so it is running quite nice for very little cash.
Did you try resetting CMOS, and using different RAM ?
There are many reasons people want their old computers to be revived. I for one am trying to revive my old P4 864 PERL motherboard after almost 14 years of its demise. I am no motherboard expert, but with lots of effort and time I could bring up a totally dead board back to a stage where it beeps when RAM is removed. If I could revive it, I will be able to run my old DOS and windows games and software in it.
Vloss is the amount of voltage the meter has measured that the capacitor has lost, which is 35%
I don't know where all the techs are.
But I can't tell you how many of these Asus 1155 boards i saw fail like this.
It's an Asus Plague with this.
I went through your attempts as well with some of them before realizing this is an Asus plague with many of their 1155 boards, matx especially.
Had a very similar ASUS LGA 1155 board fail this same exact way. I just replaced it but was hoping to see what the issue might have been. Good try anyway.
I'm rewatching all your video's lately. Need something to think about now as I want to move on I bet I never do, lol.
Thank you for posting it!
I'm hooked, you have taught me a lot by your videos!
One question though, why don't you ever measure the power over the mosfets/buck converters and various components surrounding them, like you do on laptops. Is it because the error codes lets you narrow down to the right ballpark
Again, thank you for making electronics available for us, curious people:-)
Måns
Mainly because if you have "fan spin" where the system has turned on, it's probably not a power issue. It could still be a rail that's working well enough for the system to think it's OK, but is low or full of ripple, but yea, we're getting to rabbit holes now.
@@Adamant_IT It's not that unreasonable (but will get you into a rabbit hole for sure), I had old motherboard where the fan turned on no problem but the CPU was stone cold. I never bothered to test it further though considering it was an old Athlon x64.
are you sure that the capacitors where putted in the right polarity ? becouse the minus is white strip also paint in board and needs to corespond with the white line of the capacitor
There are also surface mount capacitors smd ones all different packag size names
goodday sir did you test the mosfets
Add some picture hooks for wall mounting?
When you replace caps, is it best to replace them with the same type & voltage/farad ratings? Thanks
Voltage rating is the _max_ the cap can handle, so it needs to be equal or higher than the original.
Capacitance (Farads) should be the same. Caps have a big tolerance on them though (20% is common) and more often than not, just being in the correct order of magnitude is good enough for a power rail.
@@Adamant_IT ok thank you 🤓
is the cmos battery good and not shorted?
1) The labor and shipping cost is way more expensive than the capacitors themselves; therefore, using solid electrolyte capacitors to replace the faulty ones.
2) The main reason for the capacitors to fail - the work load and environment temperature are reaching the design margin; therefore, buying low ESR solid electrolyte capacitors and increase the spec (e.g. 6.3V 820 uF --> 9~12V, 1000~2000 uF) should be able to prevent the similar failure in the long run.
3) Using proper soldier removing tool.
4) Dead CPU won't turn on cooling fan.
5) Buy a PCI/PCIe diagnostic card for PC.
If you have a failure like this a CPU, PSU, SSD and MEM are all OK and recapping doesn't fix then what other components could be at fault. Could it be support chips like the thing under the blue heat sink?
It was probably never the capacitors, i've owned this exact board, and it had heavy use for most of those years, sold it recently and still worked like new, i also own some slightly earlier boards that have run big overclocks for many years, those boards are still fine to this day, i think the capacitor problems go back a few years before this boards time
My uncle has one, LGA 775 but not the exact same one. It's still working fine even though some capacitors are bulging and even leaking / the insides are coming out of the cross top.
I've warned him though, that the motherboard is in the verge of dying.
Don’t you need some kind of ventilation when working with leaded solder or? I really would like a good cheap alternative to that toxic stuff.
Liked the video, but was wondering why you didn't buy a new Asus P5g41t-m Lx V2 Motherboard from China ($39.94US).
Good video, I have questions how can you know is the bios chip in laptops because I see a loots of similar chips in the borad
There's a lot of small cues that give it away, but the main one is the shape. Most BIOS chips are either WSON8 (flat rectangle) or SOP8 (square, 4 legs each side) The only other SOP8 packages you're likely to see are mosfets, and you can spot those because they'll be part of a power circuit (and have capacitors, inductors, etc nearby).
The last cue is just to check what's written on the chip. There's only a couple of common brands for BIOS ROM chips, such as MXIC, Microchip, Winbond, and they tend to have that brand name written on the chip (unlike mosfets, which only have a part number) so if you see a chip out on its own, and it has winbond written on it, that's your guy. Keep in mind that a lot of laptops have more than one BIOS chip as well though.
@@Adamant_IT thank you very much
most likely a borked northbridge, not really worthwhile to replace that...maybe you can try reflowing it? nothing to lose at this point
sorry, it didn't work out but I got plenty out of it thankyou very much. very interesting.
I have that problem where I don't know when to give up but I really want to know what is wrong with the board! How about testing for chipset / rail voltages voltages? If all looks good there then it's cool to stop... pretty please!?
No plans to come back to this for now, but yea, my next guess is Northbridge issues. Likely a dead chip or back solder ball though. It was producing some heat, so it's getting power.
I repaired two Capacitors on a board, plugged the board in, a few seconds later another one blew up.
Agree w/below poster; good vid.
for the first clean the both dim slot with alcohol or thinner and then put in memory again turn on power motherboard if still error accour then Please check with selector multimeter countinuity on dim voltage, usualy the both dim voltage are broken connection, so just put wire on pin 3.3 or 3.7 volt on dim1 slot in the back motherboard to dim2 , its work with me, or , may be can help
Surface mount 'solid state' electrolytics fail more often then you might think. Whereas through-hole electrolytics sometimes offer visual clues to their failure (bulged domes and leaking electrolyte), the surface mount solid state electrolytics are trickier to diagnose (because no visual clues present) and replace (because they muist be removed by careful clockwise or anti-clockwise rotation so as not to rip away the tissue-paper delicate copper pads on which they are surface mounted). If the board refuses to work after replacing the through-hole electrolytics, I would absolutely remove (with great care - see Mr Carlson's lab video on surface mount/solid state cap removal) the surface mount/solid state electrolytics, before testing them with an ESR meter. Then replace with an equivalent solid state...or indeed a through-hole....and test the board again. I much prefer repairing electronics which do NOT have surface mount electrolytics, because two many times, I have used pliers to twist off solid state caps, only to discover I have ripped away the delicate copper pads x 2 connecting them to the board. Repairing this, in a two-layer boards, is a fucking nightmare....hours of avoidable labour...hence enormous care is required when removing them, and of course, they cannot be tested for ESR in circuit, which means you HAVE to remove them to test for ESR, and cannot simply remove 'one leg' before testing, because there is not enough space to remove only one leg before testing.
Done a few full board re-caps in my day, all completely successful, except that one time where it's never the CPU, except where it's the CPU 🤣 No joke, after a re-cap, still had a toasty CPU, but the entire thing started with a PSU that came 1 step from a firecracker so it had it coming. New PSU+new CPU+ re-cap and it was back on the road.
Only thing I was curious with your Asus board here is if it acted the same way with or without any mem modules. I have seen a post issue like that previously, and being an extremely early ddr3 board, it's QVL would be quite limited and without post, it would be impossible to tell what version of BIOS was on there, but it looks like you went to the trouble of pulling the chip completely and reflashing it, so I'm just assuming you used the very latest firmware available for that Mobo.... Too bad this repair was a bust, but I used to love doing re-cap jobs back in the day.. I still enjoy picking up a 75in 4 year old 4k smart Vizio TV for free off Craigslist and re-capping the PSU for under $2 just to cure the itch, and of course to hoard 75in TVs and put them in every room in and outside my house that I spent less than $10 on all of them total🤣
Totally love your channel, learned so much from you Graham! Thanks for all you do and keep up the fun good work!
Thanks for the video =)
Have you tried cleaning the PCI-E slot that could be a reason why you don't get a display on the GPU because of corroded pins in the slot fixed it with contact cleaner of WD-40 and a wire brush that could fit in between the gap
Hi, I'm struggling to debug this problem on my laptop.
As soon as I unplug the charger, the CPU speed drops to 0.48GHz (BD PROCHOT triggered)
To make it work again up to its speed (2.2G) I need to unplug charger AND battery, and power up again with either battery or charger. This happens with Linux too, so it's definitely an hardware problem, the BIOS battery is ok. Do you have any suggestions?
Thanks!
😂🤣😆I know, Graham "It's never the CPU" Memory error, is it possible you have two different Memory Ram Sticks inserted? Just double check everything.. I could've sworn, you took one out, and inserted a new one. I would have replaced both.. why check one old, and add a new one.. wouldn't that affect the memory value in some way?
Just asking, trying to figure this out myself, just an amateur.🤗
I think you should have at least checked some canned caps, expecially the one near the parallel port that is suspiciously tilted.
It just drives me nuts how motherboard manufacturers indicate the positive side of caps on the silk screen while every other manufacturer of anything indicate the negative as the cap itself. Whyyyy???
solid state caps can also go bad should always check the ESR of all caps solid state or not
It looks like a failure of the northbridge, or the voltage regulator that powers the RAM
If not BIOS and capacitors it's probably a chipset or a multicontroller. Could be a battery too.
The issue here is not caused by electrolytic capacitors, but rather by low quality capacitors. I have an Hi-Fi system with about 25 years, and all capacitors are electrolytic. There are no bulged capacitors, because they have used high quality capacitors (the unit was manufactured in Japan). These computer motherboards and many other devices (like TVs) have failing capacitors due to the manufacturer cutting costs, by using cheap chinese capacitors.
@Mr Guru Yeah, fixed. Thanks for noticing.
pretty sure he can get another board the same second hand i know i have working one similar
excellent
shows even if the repair did not work you can still learn something.
if the board has sentimental value put it in a frame and hang it on the wall like some of your wall art!!!
I've recently got a faulty motherboard (lga1150) I'm getting ready to post-mortem or possibly repair. Used air compressor to clean out daughter's computer like done several times. Next time she booted it started sporadically turning on and off and once in a while makes it into bios before shutting off again(led fans also got brighter sometimes). With the knowledge I've learned from your videos I hope to come to a definitive cause of death.
Thank you M.r
From my experience with flashing BIOS chips outside the board with a programmer; it sometimes ends up with a different result than the original BIOS of a working board.
Correct, the unique identifiers of the board won't be preserved. Those would need to be taken from the backup and written in manually with a hex editor.
Change the cmos battery, some old boards doesn't boot with emty batteries
Nice video, keep it up, thanks :)
Very fun video. I would really like to see how to fix this issue though :D