ASUS P2B Restoration: Board #5 - Replace 25-year-old capacitors?
HTML-код
- Опубликовано: 15 июн 2024
- The board in this video is in almost pristine condition. The ASUS hardware monitoring IC and ISA slots are free from corrosion, the back of the board is immaculate, and the serial ports are clear from rust. Since there is no need for much restoration, this project allows me to answer a question I have had for many years: Do we need to replace dated capacitors on motherboards even though they seem in good condition? I am sure there is no easy answer, but today, we will answer it for my six P2B motherboards from ASUS.
▬▬▬▬ PCBWay ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
Sponsored by PCBWay: 🔸pcbway.com/g/lF253I
PCBWay provides full feature custom PCB prototyping services, 3D printing, and much more. Check them out using my referral link above and get a 5 USD welcome bonus when you sign up as a new customer!
Full-Color PCB Printing:
🔸www.pcbway.com/blog/News/Unlo...
My contribution to the Community Project Space at PCBWay:
🔸www.pcbway.com/project/member...
▬▬▬▬ Affiliate Links ▬▬▬▬▬▬▬▬▬▬▬▬▬
Cordless Rotary Tool - Engraving pen
Amazon US: 🔸amzn.to/4a0ZgSs
Amazon UK: 🔸amzn.to/3JEIltO
Amazon DE: 🔸amzn.to/4ddf5Ik
AliExpress: 🔸s.click.aliexpress.com/e/_DkI...
Titanium Alloy Tweezers Professional Repair
🔸s.click.aliexpress.com/e/_Ddj...
Copper wire different diameter (0.1mm and 0.2mm)
AliExpress: 🔸s.click.aliexpress.com/e/_DeN...
AliExpress: Mechanic low-melt solder:
🔸s.click.aliexpress.com/e/_DDQ...
🔸s.click.aliexpress.com/e/_DEX...
AliExpress: Mechanic Solder Mask
🔸s.click.aliexpress.com/e/_Ddy...
AliExpress: Desoldering Station / Desoldering Pump
🔸s.click.aliexpress.com/e/_DCV...
🔸s.click.aliexpress.com/e/_Dmm...
🔸s.click.aliexpress.com/e/_Dln...
Flux for easy soldering
🔸s.click.aliexpress.com/e/_DFk...
🔸s.click.aliexpress.com/e/_Dlv...
TOWOT Solder Wick from Amazon:
🔸US: amzn.to/3SyHrDB
🔸DE: amzn.to/3HFpCgJ
🔸UK: amzn.to/3OeIVRZ
7-45X Continuous Zoom Stereo Trinocular Microscope
🔸www.aliexpress.com/item/32880...
Solder Mask from Amazon
US:
🔸amzn.to/43K8KOy (6pcs)
🔸amzn.to/470ISRw (4pcs + UV light)
DE 6pcs:
🔸amzn.to/3Dr49WU
🔸amzn.to/43Av6lC
🔸amzn.to/44XIbGB
FNIRSI HS-01 Soldering Iron
🔸s.click.aliexpress.com/e/_DE4...
T2S Plus Thermal Camera:
Amazon US:🔸amzn.to/3VV6zaU
Xinfrared:🔸www.xinfrared.com/products/t2...
Get 10% OFF with code: t2splus10
▬▬▬▬ Support Bits und Bolts ▬▬▬▬▬▬▬▬▬▬
Patreon: 🔸 / bitsundbolts
Bitcoin: bc1q2dvf9flf78lam33qjj9w9xgvfxdgre7g9x3rmw
Twitter/X: 🔸 / bitsundbolts
Facebook: 🔸 / bitsundbolts
Website: 🔸bitsundbolts.com/
▬▬▬▬ Timestamps ▬▬▬▬▬▬▬▬▬▬▬▬▬▬
00:00 Intro
01:41 AGP trouble
02:03 PCBWay
02:50 Fixing AGP slot
05:00 High CPU TDP
05:42 First test
06:22 Power benchmarks
09:20 Remove old capacitors
10:17 Clear holes from solder
11:31 Testing old capacitors
12:54 Power control chip and MOSFET
14:36 Install new capacitors
15:05 Testing new capacitors
15:55 Conclusion 
Развлечения
Demonstration of the negative case is critically important and often overlooked. The recap test was not a waste.
You're right. At least you know that there may not be the urgent need to blindly remove all caps from the board. I'm aware that my equipment is not necessarily accurate. However, it should provide at least some insights. Thanks for watching!
That unreal intro is so epic! I wish this would get a remake in UE5 intro included. lol
Nice video. Some people here in brazil are against recap, because most of the time you are changing good capacitor for cheap capacitor.
I tried this time to get better caps, but I have the same issue. I don't know if the caps I get are of good quality. So, I'll skip recapping in the future if it's not entirely necessary.
@@bitsundbolts in this case the caps you used (the gold and black ones with the (Y) vent are actually a top brand called Nippon chemi-con so those are great)
The green ones with the (X) vent are cheap crapy caps and 100% worse than a rubycon
Is it possible to use solid capacitors to replace the old ones?
@@fetazihsp Yes it is possible. But most solid caps are designed for surface mount, so they have no wires.
@@bitsundbolts When I need good caps, I buy them at the big distributors like Digikey, LCSC and Mouser. Than in the selection I pick Kemet, Kyocera AVX, Nippon Chemi-Con, Nichicon, Matsushita, Panasonic (which is Matsushita), TDK, Vishay, Samwha, NIC and Sanyo
There will be some other companies I forgot in the list but normaly capacitors from these companies last a long time. I've seen 40 year old caps from them still being fine.
On the otherhand caps from capxon for example are called crapxon for a reason. Some of the abouve mentioned companies also making crappy lines of caps. I think they made them in reaction of the chinese to compete with them.
Sanyo and Rubycon are japanese so they were not affected by the plague, the affected ones were from Taiwan mostly.
Rubycom MBZ and MCZ and one of the series from Sanyo were affected by the capacitor plague
@@andrewsawesome Are you sure? As far as I know the original formula (the good one) was from Rubycon and when some engineers left it they copied it wrong and was used among taiwanese capacitor manufacturers.
@@drgusman The stolen formula is a story, which is very convenient for some industry players. Sometimes high-end manufacturers make mistakes when trying to outpace each other on specific capacity or claimed operating limits.
I mean there's probably a kernel of truth to it; but would it surprise you if a fragile chemistry which is so susceptible to being formulated wrong, also gets upset by some unforeseen material in the capacitor itself, something outgassing from the rubber bung, a degrading coating on the inside of the casing or a stray alloy element; or slow ingress of outside material diffusing through or around the rubber bung or a microscopic crack in the top vent? Or egress of electrolyte components through these? Things that aren't covered by accelerated endurance testing?
Every time the new technology generation rolls around, you have to be somewhat cautious of spec claims. Sometimes things are exactly as good as they claim and better; sometimes they are not.
@@SianaGearz No, of course it will not surprise me, and I'm sure that there are batches of bad capacitors from mostly every single manufacturer out there :)
I just was referring to what was called the "capacitor plague", I lived it myself, in that time I was working as computer technician and it was a true nightmare, I remember specially the mainboards that came with the green caps, very similar to the ones on the video but they were Chinon if I recall it right, those were mostly guaranteed to fail xD
I have motherboards from that time with japanese capacitors gone bad (Chemicon and Nichicon).
OMG the old caps still in great condition
Looks like they made high quality components back then!
The old caps are the following:
Sanyo CA series 6.3v 1000uf
Rubycon YXG series 6.3v 1000uf
I’m not surprised they are close to their original specs
The rubycon are perfect still no surprise there.
Recapping this board is not necessary
0.39ohm is quite high should be around 0.08.
@@thomasandrews9355 I would have to check the datasheets but you are very likely correct that is too high
Considering there measured capacitance they are probably in need for a reform
Edit: Ok just checked
Rubycon YXG 6.3v 1000uf should be at 0.08 so yes you are correct
Sanyo CA 6.3v 1000uf should be at 0.15 so those are reading correctly
Like I said they all look fine and at least they are in the need of a reform
@@garfieldbenb3784 it’s probably fine these aren’t p4s. I use the yxg and similar when restoring old Mac’s so I’m familiar with them.
@@thomasandrews9355I know it’s fine for a board like this
I recap my socket 478 boards with rubycon MBZ/MCZ or polymers for the output side of the VRMs and the input I use MBZ or Panasonic FM
@@garfieldbenb3784 absolutely!
Capacitors which have a high internal resistance will heat up when used in a power supply circuit requiring low ESR. This can be easily seen using infrared imaging it just feeling with a finger.
@3:32 got off my chair for some Schuhplattler 🤣
My preferred method for removing through hole capacitors is similar but involves using a large flat soldering tip to heat both leads simultaneously. The capacitor will often fall out without further help.
For clearing the holes I tip the board on its edge. Heat the hole from one side with the soldering iron and apply the de-soldering iron the the other side.
Some guy told that the most PCB friendly way to remove caps is destroying them with pliers, then disorder small wires is much more safe.
Haha, how many hands do you have 😂. I think I wouldn't be able to work on both sides at the same time. Interesting method with the large tip. I don't have one that's big enough - that's why it probably never crossed my mind. Thanks for the hint!
I've seen people use pliers and twist off caps. Unfortunately, then I cannot test them. If you're looking for a fault, then this method is not ideal.
@@bitsundbolts Just the usual two. It helps to use something to hold the PCB. An L shape PCB holding rack would work well. I often use an big alu heatsink from a scrapped audio amplifier. Also you need a desoldering station with vacuum pump. I use a JBC station at work :D. At home I have a ZD-8915.
Cap replacement is definitely required on *some* boards - I got a nice QDI i815 based board that was completely unstable - you could get to DOS, but even trying to start Windows setup would crash the system - after recapping it works like a champ!
I suppose the age old adage stands - 'if it ain't broke, don't fix it...'.
Agree! If the board works, no need to recap.
QDI used crappy capacitors. Otherwise QDI boards were excellent otherwise, and work extremely well with good capacitors. On some QDI boards the capacitors will look fine, but they are defective, mostly 6.3V 1000uF ones
@@michvod Its not just problem of QDI, but most board from the time of caps plague are affected. I have i815 board from MSi and all caps have to be changed - and most of them were visible leaky.
@@xsc1000 yeah MSI. I had some P35 775 boards MSI boards (from 2009) that had bad caps (around the CPU socket). Also I recapped lots of Abit boards. Asus and Gigabyte were better, but still occasionally had bad caps.
Not a waste of time from my perspective, it's always fun to watch detailed soldering. 5 down, 1 to go.
Yes! What a journey!
Awesome again! I am not surprised by the cap results. I have done similar experiments with Commodore 64 and Amiga 2000 mainboards and have found basically the same result. Some caps just don't need to be replaced and may last 40 years or more. The recapping just to recap fad is totally unnecessary in many cases, Especially true for through hole. SMD caps tend to be another story it seems and not counting the early 2000s to early 2010s as I have seen a lot of failures in this era.
nice method to deal with those nasty ground planes!
And thanks for covering the caps: I'm sure someone will still comment "you must replace all the capacitors" :)
Depends on your iron and whether or not the solder is contaminated with capacitor leakage. For many, hot air is better specifically for the reasons he states! Many irons simply aren’t responsive enough to saturate the ground plane and heat the plates thru-hole all the way with the component lead. You pretty much need a cartridge type iron to count on having the performance needed for this method.
Capacitance and esr change at different voltages. I'd like to see how they'd handle themselves in a high current buck converter. 12v > 5v buck with a measurement of input and output voltage and current and the output ripple. Do this at lower current and higher current. You'd probably see the ESR and ESL start showing.
The capacitor plague is my reasoning behind an automatic re-cap. Before the plague I would try and re-form capacitors that went out of spec and had really good luck with them. The plague will cause a capacitor to test and run within spec but will degrade after use. In storage they seem to survive without issue. It's when you actually start making them work. For me it was a lot more difficult to test and replace when necessary, reassemble the machine, have it pop, then disassemble test and replace. I used to do a lot more than mainboards. For TV, VCR, power supplies, etc. it was a lot more work for little benefit.
Rocking an ALR Business VEISA 386 from 1990. All original components 😎👍
The mid to late 90s was a very narrow window of good capacitors in radial and axial types. Before 1996, those types of capacitors were afflicted with a different problem, corrosive electrolyte eating the seals out the bottom of the capacitor and leaking all over the logic board. Nichicon capacitors are especially bad about it, they're a plague in Apple computer power supplies and analog boards.
SMD capacitors in the 80s, 90s and early 2000s were even worse, those all need to be replaced, from any manufacturer. They also had seal failures, but since the plastic foot can wick the leaked electrolyte, it makes it very difficult to detect until its too late.
Also, if you find any purple Sanyo polymer capacitors, get rid of them. They don't physically leak, but they become electrically leaky, go high ESR and the rated capacitance plummets to nothing.
Certainly surprising to me those caps are in such good condition. Clearly a recap wasn't needed at this stage. Given how prevalent bad caps were clearly Asus used good quality ones back then 🙂
Yeah but the thing is caps will still degrade regardless how good they are due to the acidic materials inside of them at some point in time they'll just eat themselves it's usually about 20 to 30 years
The only exception might be solid capacitors that don't use electrolytic.
Though those can potentially explode violently.
Ah nearly weekend and time again for this entertaining series. Good stuff bub!
i got a cheaper more budget slot 1 mobo and even that one has very good quality capacitors! they didn't cheap out on that kind of stuff before the plague! great video as always, looking forward to the last one.
Not surprised, but an interesting experiment.
Interesting how to diagnose caps without disordering ! They are all in parallel, ESR meter won't help I think
well elec repair and polka, thats a first. I think you have me sold on the chipquick solder vs hot air.
Well, I was told I and other youtubers reuse the same music - I tried something different those time 😅.
I do like low-melt solder - it's like magic ✨
As none of my Atari STs caps are bad, nothing surprises me as long as the caps are of a respectable brand
I'm keeping most of the electrolytics in my ancient 1949 TV, the set works fine with them, their parameters measure decent enough, and there are no visible signs of damage. The P2B's capacitors are only 1/3 as old so unless they are known problematic series I'd expect them to have several decades left before failing from old age.
For some quick tests: measure ripple voltage with an oscilloscope and measure capacitor temperature with your thermal camera. If both look good they're not about to have a bulging-type failure. They could still leak though.
Thanks for confirming - it if looks good leave it - especially if its not booting - good looking caps are usually not the problem
Yes, unless you have a good reason to suspect bad caps, just leave them. There are many other devices that can go bad before caps.
@@bitsundbolts You can check voltage ripple by oscilloscope. If its OK, caps are OK too.
@@xsc1000 I second this. A ripple check with an oscilloscope is the best way to find functionally bad electrolytic capacitors that still may look fine. I suppose an in-circuit ESR could also work but that won't give any idea as to what is acceptable in the specific application.
Great video!
Wasn't surprised with the results of the capacitor change :-).
I guess the quality standards were a bit different back then - maybe it wasn't the case that you had to cheap out on every corner to remain competitive.
On some of my boards I've ended up replacing all except 10uf caps due to constantly having to pull boards back out of a build and replace newly failed caps. My Jetway 994AN-L is the next candidate due another dead 1000uf cap - so far the only one that has failed with the electrolyte itself venting (and after significantly reduced stress in the form of various Via C3 and my SiS 315).
I've seen some similar Sanyo cap (1200uf however) fail on my Asus A7V266-E (low ESR 2200UF as a temporary standin), while all of the Rubycon caps are perfectly fine. And I've got the motherboards in 2 Xbox to recap due to nearly all Nichicon caps having failed (had the same ones fail on an early Intel LGA775 board)
@BitUndBolts
I was a bit surprised at the large caps, but given the healthy state of the caps, I would have left them in and used the new ones on the next board, keeping the extra set spare. That's just me though. Well done on that APG port pin!!
Yeah, that board didn't need recapping. At least I learnt something. I have enough caps for many board repairs.
@@bitsundbolts
Of course, it never hurts to replace caps either. 🙂
That inductor does not have an iron core, it would be to lossy with the high frequency used in the power circuit, if it was iron it would be even hotter.
It instead is made with a ferrite core of some formulation, modern ferrites are even better with even lower losses making then cooler with the same power.
If nothing else, redoing the caps will have been better for ripple on the voltage lines going into the CPU. it's not going to immediately change the temperature of components around the CPU, but it will still be a far cleaner signal. ESR isn't everything, but how fast the capacitor is to switching definitely is. In a PIII system where you're getting 5v in to generate the voltages for the CPU, it makes not a whole lot of sense unless you're overclocking or going at higher frequencies. for a Pentium 4 system of any kind, however, it's really everything. The plague caps had very good switching characteristics when they were new, but as time went on and they got hot, they get unstable and vent from the poor chemical composition.
Anyone running a Pentium 4 system with a high-end (for the board) CPU should be looking to upgrade the capacitors to polymers, following the rules for replacing electrolytics with poly as needed given how much faster they are. the caps on the input side of the VRM section can still be electrolytic, but need to have fast recovery time (higher-end general use caps and most power supply rated caps i.e. Nichicon PS Series, now discontinued as of last year but still attainable) in order to satisfy the needs of those power hungry CPUs. if you're overclocking, this is especially significant as the VRM section needs to be able to provide clean power to the CPU to supplement running the CPU out of bounds of the deemed safe running speed.
I have a GX400 I recently put a socket 423 to 478 adapter in with a 2.8GHz Northwood to max the system out. When I did that, I also did my homework and upgraded the VRM card to one with a better design that was only used in the Precision 330. But, I haven't had to really worry too much about capacitors too much on the motherboard itself, as my particular revision has Sanyo oscon poly caps on the output, and some nicer sanyos on the input. Given this is still from the cap plague though, and to minimize issues in the future, I'm still replacing every capacitor on the board as needed for the sake of completeness and caution.
I have a few boards and I can tell the Pentium 4 are the ones that give more problems, like killing CPUs and memories. The motherboard itself would still work, but the other components were killed after some time working. I believe those caps are the culprit combined with poor power delivery circuits.
I agree with you, is not just a question of consuming more power, but it is actually dangerous for the hardware if the caps are not good. Even the voltage could be right when measured with a multi-meter you wouldn't know how bad the ripple is. For that you need an oscilloscope.
3:50
I think Big Top Polka would have been better here :P
IMO recapping on motherboards is only needed if there are clear physical signs of failure such as bulging or leaking or you're seeing stability issues that aren't related to excessive heat. Otherwise I would leave the caps as they were.
The capacitor plague era was awful to go through! Every PC I built or owned from the late Pentium 2 Slot 1 era all the way forward to an early Core 2 Quad machine (and everything in-between on both the Intel & AMD side) ended up with motherboards with leaking or bulging caps on it after a couple of years.
However, I have 4 more recent PCs, the oldest being an Ivy Bridge i7, a Haswell Xeon desktop, a Xeon W 2135 based workstation running TrueNAS 24/7 and the newest being a 3 year old B550 AM4 board with a 5700x3D recently installed in it. None of those machines have had any capacitor or stability issues that I have seen. The Ivy Bridge desktop is one I purchased in 2012 and still gets regular use as living room game emu box.
Not really surprised about the capacitors, since they were not from the troublesome period.
I think recapping depends on who made the original caps and I would only go with the good well known brands for replacement (rubycon, nichicon, etc...) I know for rare hardware I would not want to chance killing it because a crappy cap decided to take out the rest of the board. Also I saw people saying you can replace one type cap for another which is not a good idea as the electrical characteristics of different types of caps are not the same. There is a reason besides cost that boards are designed the way they are, so unless you really know what you are doing, like for like is the way to go.
Try an ECS Athlon 64 board. The caps on those are party poppers.
As you've hinted, one way bad caps can exhibit their failure is by excessive vrm transistor heat.
I recapped a working but leaking Socket A board and the vrms dropped easily 20*c.
Yes, those caps were obviously dead in that case, but it is definitely a sign to look for as you did.
I'm sure that I'll get a similar board like your socket A one day and hope to be able to reproduce this behaviour! You just need to get it in the right state. If the caps are too leaky, it probably triggers the overcurrent protection in the power supply.
So you mean IR camera can show a bad cap without disordering? interesting...
@@alexloktionoff6833 IR camera + ripple on CPU power line measured by oscilloscope. IR camera can find one bad even others are still OK.
Sanyo and Rubycon are good brands. I wouldn't be bothered about those unless there's real symptoms to make you think they're dying, at least at this board's age.
It's usually just the random offbrand ones that are worth changing as a preventative measure.
That said, electrolytics are like people - they all die eventually, some just take a lot longer than others 🤣
Very much in the camp of, if isn't broke, don't change it. If some caps have failed, would opt to change all caps of the same brand and value but otherwise there is enough landfill as it is :)
Sometimes you just know the capacitors are bad even if there are no visible signs. Because you already dealt with a similar circuit from the same era equipped with similar caps.
Sometimes you have a strong reason to suspect that the caps are good as new. To me this board was a strong candidate, so i wasn't exactly surprised.
Usually i just pull one cap from each series used, one that i expect to experience more ripple current or more heat, and just check that, and if it looks good, back in it goes. If it's even a little dubious, all of them get replaced.
Unfortunately how good the caps are is generally a hindsight assessment. You don't know what you're putting in today, how it's going to fare. I figure something that has been around for 20-30-40 years with little degradation can probably do the same again.
But some very smart people disagree. James (The Bald Engineer) ex Kemet executive, also a classic computer enthusiast, would yeet them all, simply because they are fundamentally wear items and have degradation modes that occur both in storage and in use, and may not fare as well e.g. with voltage spikes, so a sudden failure in the future is a possibility. It's not like our ESR meters test them all the way to the forming voltage, indeed they only do a small-voltage test.
Since P2 series of Asus motherboards originate mostly from 1998 I kind of expected them to perform fine in capacitor departament.
That bent contact though, just looking at it made me breath heavily. I didn't think such good results will be possible.
Works the same as pci-e slots do. Don't need to ask me how I know. :)
Oh oh - I have a feeling what that could be!
A recap wasn't needed, but doing so you might have extended the life of this board by another 25+ years. The caps will eventually dry out or leak, it's their nature. It may take 50+ years, but they will fail some day. You just gave it more time :) if you are preserving that hardware, you did well. If you don't count on having these boards around for the next decades... Well... You wasted your time 😂
Consider using an ESR meter that can test the caps in circuit. An oscilloscope might also be a good tool for testing the ripple
ERS meter is good for testing caps, but gives you false value if just few caps are bad. Because they are connected in paralel.
Nice job on those caps removal... Are you using solder without roisin core?
Haha - nice that you asked this question. I will dedicate a minute in one of my future videos to the solder I am using. As a matter of fact - it is a perfect match for the boards I am working on. I found a date stamp on my "Balver Zinn" Röhrenlot (made in Germany, 26.08.1998) - I guess this is a bit of fun trivia about the stuff I use... But it does have a rosin core. If you wouldn't ask, I would never look at my spool of solder in detail. Thank you.
Back in 2005 I recapped a "plague" era P4 using caps from some old P2 board.. Symptom was freezing while running and refusing to restart for a few hours.. as of last month it's still going strong in daily use (runs old photo editing software for a small business). I think 19 years for a "free" repair isn't bad. Don't be too hasty to remove 1990's caps unless you actually have good reason. I don't think new ones are anything like as good.
Absolutely agree with what you say! I rarely saw caps on PII and older boards failed. I think the quality was a lot better before the plague.
@@bitsundbolts Yep if anything the ones you put in will fail soon and you'll just end up putting the originals back on.
amazing that these caps were ok after this long, tbh. i found last week the abit bp6, the one with dual celeron support, it's pretty known as a board, and it doesn't start at all, dead. i heard that the caps were a big issue with this board, you think it's possible that changing all the caps would bring it to life, BuB? i think it's the most valuable board from my so called board collection :) gr8 video, as usual
You could try to measure the resistance between both leads of one of the caps. If it is low, it might indicate an issue, but to be sure, take one of the caps out that are located next to the CPU socket and measure them if you can. It could very well be that the caps are responsible for the board not working. How do they look? Any swelling or leakage? Where did you find that board? Maybe there are deep scratches on the back of the board. Or contact of through-hole components are touching.
@@bitsundbolts i found it at the scrapyard, visually, it looks ok, and the caps look perfect, no swelling and no leaks at all, which is odd since the forums that i read said the caps where a big issue here, saw a video of TechTangent where he replaced all the caps. thanks for the info, BuB. when you'll get back to Germany...I could send it to you, if you'll have the time :)
I probably will be in Germany only in December. Regarding the caps, it could be that the caps look fine, but they still dried up. I also get most of my items at the scrapyard now. Usually, they always have some fault due to the rough handling there. Maybe one of the voltage regulators popped off the board 😅
Good to know that OK-looking caps can apparently reasonably be expected to actually be OK, as long as they were not subject to the plague. 👍🏻 This board looks like it didn't have a particularly hard life either, but with an unknown history/origin, it's always hard to tell.
Looks like I should get myself some low melt solder too. I also prefer the wiggle method for removing old caps, but clearing the holes often ends up being a major hassle. Thanks for including links to all the stuff you use (some of them no longer work though). Also, I'm curious what this low melt solder is made of, I can't tell with absolute certainty from the info they provide on AliExpress, does it say Sn42Bi50Cu8 on the label?
Yeah, I know that some of the links to not work in every country. It depends from which location you're accessing the stores as it seems. Unfortunately, I don't know how to fix that. I should write a URL resolver where I can update links and it would update below all my videos instead of posting the URLs directly... Regarding the solder, you are spot on: Sn42Bi50Cu8. 42% Tin, 50% Bismuth, 0% copper.
@@bitsundbolts Oooh, besides the "Page Not Found" title and the big "Sorry, the page you requested can not be found", there's a smaller "unavailable in your location" message too. Typically Chinese. 😄 I just saw the bigger message first and immediately concluded that the AliExpress listings had just gone stale, as they often do. No need to fix anything in that case. Providing multiple links is already more than anyone can ask for and it works in my case, too. I just wasn't aware they link to similar products - labeling them as "alternative listing" may help to clarify that though.
A URL resolver would indeed be an elegant solution though, esp. for keeping affiliate links up to date. 😉 I have one in place for other purposes too, super simple (quick & dirty PHP switch/case in my ...case) and super handy for all kinds of situations where you want to retain control beyond the time of posting a link. 👍
If I were to give hardware to someone not equipped to handle board repair then I would replace the capacitors on principle. If they were to go into my own projects then I wouldn't necessarily do it unless needed. I just think it's about understanding who will have the board and whether or not they can catch them failing and then being able to handle the repair.
Good point. Although replacing caps is easier than replacing SMD components (in my opinion), you still need the right equipment.
Instead of soldering leg by leg try a line of liquid solder (solder paste) and hot air.
You might want to add some kind of note or label showing the date you recapped... That way the next person 10 years from now will know 😉
caps if they dont show signs of leaking shouldn't be replaced in my book. but everyone is different. but you always have to be careful with counterfeit caps. they did sneak into everything back then
A good electronics technician repairs a problem. An excellent electronics technician will repair the problem and prevent future problems.
Electrolytic capacitors have exceeded their useful life. If they did not fail, they will surely fail in the short term.
They all have to be replaced.
Most of my boards are still okay and don´t needed nor need a recap. Had to repair the Socket A Gigabyte Board of my Mother in Law, some Fujitsu Boards with caps from OST and also Grafikcards with bulged caps. Most of them had to sustain long operating hours and heat. If the cap is bulged or has a brown top, then replace. Otherwise not.
Yeah, I have a few of those candidates as well. Socket A and P4 boards seem to always be affected by bad caps.
I've seen some enthusiasts using solid capacitors in these old boards, most of them did for better overclocking. Do you think this could be useful?
I did the same for the ASUS P4B in an older video. I'm not sure if there will be a difference on those boards. I doubt it, but I don't know. At least for my use case, electrolytic caps will be fine.
11:32 One might say this is a P2B… no cap 😆
I like how unsupported CPUs boot. They don't now. You have to flash before they boot.
Yeah, pretty cool! These days, if there's no microcode (AM4 was my experience 😅), you're out of luck!
@@bitsundbolts It's not just AM4. AM5 will also require an update and all intel chips as well. It's less of an issue on intel because every other generation is a new board anyway.
I had an agp slot that was damaged and I could not find anyone who sold agp slots, I also did not have donor board.
welp, at least you can use de-solder capps for other project.
My ASUS P2-99 have the same result. Caps are in good condition and no need to recap.😂
Think about it:
They weren’t reformulating electrolytic capacitors in 1999 for no reason. There has to be a reason to make them worse. The “only stole half the formula without the stabilizer” part only explains the Taiwan/China caps that were failing within a year, not the name-brand ones that fail in several years to a decade when their old formula did not. Why were they reformulating to a worse formula BEFORE half the formula was stolen?
It’s because of RoHS. Mandatory environmental regulations have put far more electronics into the ground than there ever would have been without Pb-Free and RoHS initiatives.
I have a feeling that your opinion is unpopular and disregarded by many. However, I 💯 agree with you! More often than not, intervention from regulators causes adverse effects! Look around these days, it's everywhere! God knows what interests were followed with all those regulations. The easiest way to get to the root cause is: follow the money - and you'll see who benefited from it and why...
its not _needed_, but the closer a device gets to 30 years old, the less inclined i am to keep the caps 'original' unless its micas that test okay or similar. electrolytics? yeah, likely going to get replaced. i don't care how 'fine' they look, and without specific test gear, you're not going to get the readings showing they're faulty when they start going out of spec until they've already failed entirely.
those cheaper LCR testers don't give you the info you need to make a proper determination. some of the higher end ones do, and gear made for the purpose (like the carlson capacitor tester) will let you know exactly how bad a cap is, but most people don't have that sort of gear.
any ideas on how to upgrade the power delivery system so it does not get as hot?
It's actually not that hot. Those chips stay below sixty degrees - which is quite good. You could add small heatsinks and have a fan blow over them if you want to reduce the temperature.
@@bitsundbolts but heatsinks are not fun to watch being glued on :P and maybe with a new power system it could overclock the cpus higher and make them a bit more stable
Sanyo caps are nightmare on some xbox 360 motherboards. Thought they were terrible quality.
The xbox 360 is a lot newer than this board. Maybe Sanyo doesn't have the same quality as in the '90.
Do a poly-mod for no.6 😆
All SMD electrolyte caps from 90's and 00's require recapping. No exceptions. I need to dig my wii out of storage and recap that. It will leak and kill it. It was a launch day wii. They are definitely affected. Other than that, I think it's a good idea. Required? probably not, but still a good idea.
it's less about are the values correct and more about they could still go bad and leak later on. I'd say at least 20 years old is a good candidate for recapping. The older it is, more so.
I'm sure every cap will fail at some point. I guess it will be the new owners of those P2Bs who will need to recap them if they want to keep them for a long time.
@@bitsundbolts I kind of feel like 80's should be automatic. And 90's should be encouraged. And then it will be more likely to survive that issue. Then again most people using these ancient boards will be willing to figure out what happenhed.
Low melt solder and wick to clean up after a slow removal of caps? I believe you have a desoldering gun/station, use it?
To me, changing caps on an old board is like an oil change in the car, you dont wait until it goes bad. Its the difference between doing service and repair.
I tried with the desoldering station. Mine isn't good enough to melt the solder. I would still have to rework with the iron. There is copper on both sides of the board, and probably also in the inner layers.
@@bitsundbolts I have a semi cheap chinese desoldering gun, have it set to 350-400 degrees when working on lead free.
I'll give it another try. I don't do much with the desoldering gun. Could be that I need to spend some time to find the right settings
@@bitsundbolts add fresh solder and it works on most things 🤣
I have used mine to clean pads when i was out of wick 🫣
For me - you didn't nothing wrong. Now you have fresh capacitors for next 25 years :)
PCI POSTCODE card all "-" means no CPU? raplaced thermal paste on an AMD Athlon on a GA-7ZX and no post after, the pci post debug just stays at ----, no beepcodes either. so I wonder if it's dead mobo or dead cpu?
Usually, the post sequence starts with C0 and then C1. I think I've seen boards also all "-" when no memory is installed. Hard to tell if it's the board, ram, or CPU - the only way to know is to test each component in a known good system.
@@bitsundbolts if my CPU died, will it damage a working motherboard? or other way, if the motherboard died, will it damage a working cpu?
I have another set of motherboard+cpu coming that should be compatible. but afriad I will break more things so I'll probably just switch the working set into the PC case (fujitsu siemens scaleo with transparent blue front)
@@GizmoTheGreen Permanent damage is unlikely unless a the power rail is way out of spec or a rail gets shorted to a lower-voltage data line. Since the motherboard is in charge of power delivery it's much more likely for a bad mobo to kill a CPU then the other way around.
10:20 - Just use toothpick maaan.
At least it does not look like the caps with fish oil... lord those are nasty.
Oh no! What have you done! You've replaced old reliable Rubycon YXG and Sanyo (now Suncon) HC with generic green Chinese capacitors and fake (or real) Chemicon KZJ series. KZJ is a known defective series! Also please stop using red Wurth Elektronic capacitors which are just rebadged Taiwanese caps.