Hi! Why recap? because all components have a rated shelf life according the manufacture. May be the 25/30 years old motherboard exceeds that shelf life. Could be working but not in specs. You dont need to change your car oil every 15..000Km, the car wont stop running but you should because manufacturer specs.

@@JorgeCarvalho_web_dev all of the components are passed their rated lifespan, not just the caps. With that logic you’d change all the semiconductors and passives. Motor oil becomes contaminants and ineffective, so I would not say that is an apples to apples analogy.

@@adriansdigitalbasement exactly. Also, in car, you have all your life, and life of passangers. That is not good example, to compare it to board, where you have almost nothing to lose. Particulary on old hardware, where you will for sure not running some sensitive database, or something important, but mostly you use it only for play.

@@adriansdigitalbasement Nop, I will not change all semiconductors. A 150 ohms resistor will be always 150 ohms and we can consider it of one type of semiconductor, it retains its specs until it burns out by external factors. A capacitor due to the way it is produced will change its own original specs over the time.

@@warrax111 It is a good example to compare oil changes to change capacitors. You are not forced to do it, the car will continue to start, like capacitors, you are not forced to change it. Some day the board wont boot (the car will start but will damage the motor) analogy. But I understand that you support Mr Adrian thoughts because he is considered a "star" on the YT world, and what he says is the law. I can not support he's argument or thoughts about the capacitors, so if and when I can, I present my own arguments

Don't be too sad, we can't carry everything anyway, but the best of it is always with us - the memories ;) Have a nice holidays.

Not only did he miss the burnt resistor, but then he incorrectly named it as a "fuse" and moronically he even replaced it with a fuse. WTF?

Probably one of the arguably few that was better was the infamous PCChips board with the fake cache. It really sucks they had to tarnish that board that way. Because the implementations on that board are otherwise simply fantastic. With a bios that can be tuned exceptionally well and supports all the cyrix 586 features with a simple toggle. And with its legendary and hard to find cache chip, it is among the best performing boards for socket 3 ever made.

I am brimming with anticipation. I think I already know the outcome, but I always like how Necro tries a few different and oddball configurations.

I certainly wish you a lot of fun and a successful repair ;)

SiS VIP! oooh, great board. Well worth the effort.

Agreed. THAT IS NOT A FUSE! I have no idea why he keeps saying "the fuse", when there is absolutely no indication, PCB marking (R11) or component marking to make him think that. I also cannot work out where he gets the "5 Watts" from (once again, there is no reason to think that). And why on earth did he replace a 0.01 Ohm resistor with a 3A fuse? WTF??

@@johncoops6897 There is such a thing as a fusible resistor. I'm fairly familiar with this type of circuit and I'm fairly confident that it doesn't make use of a current sense resistor. If it did it would need to be a precision resistor not a 10% tolerance wire wound resistor.

@@felixokeefe for digital electronics of that era 10% tolerance fully acceptable.

@@johncoops6897 Because a fuse is a 0.01 ohm resistor

@@johncoops6897 even a straight wire is technically a fusable resistor, (it will burn like a fuse if you try something like 10A at 5V)

You are right the DX50 was a boy ready to run in the 32 bit era. It was mind-blowing

Что касается частот на VLB -- в одной из моих сборок система стабильно работала на частоте 66 MHz. (Процессор 486DX2-66 работал как "DX2-133". ;) ) С шиной PCI также было не всё просто: в период становления стандарта PCI v. 2.0 встречались MB и видеокарты, способные работать на 66 MHz. Процессоров, работавших на такой частоте мне не встречалось, но попадались экземпляры DX4-133, работавшие как "DX3-180". Правда, переключение частоты с 33 или 40 на 60 приходилось делать уже после старта системы.

Most PCI implementations would run 1/2 at 25mhz on a 50mhz fsb. I have a UMC 486 board that has both pci and VLB. And it will drop the PCI to 20mhz when I put the FSB to 40mhz. Which is a shame. But its VLB implementation also seems to be very slightly lower performing than a pure VLB board I have when replicating configuration as closely as possible. I have found on my VIP board that when the busses are the same speed then the VLB is only very slightly faster using my friends ET4000 cards. The difference seems a little more pronounced on a dx4 than on a dx2 as well. But they are still both within a few percent. And best comparisons I can make with HDD controller's indicates they are identical for the most part. Both using flash and spinning rust. However my VLB board seems to outperform the VIP board I have by a more significant amount.. So I can't call that apples to apples as its probably got other optimising going on. We even did comparisons with L2 an L1 disabled at different stages to see if we could isolate more situations. And the differences seemed consistent.

@@wishusknight3009 В период становления стандарта PCI 2.0 встречались m/b и видеокарты, у которых была документирована работа на 66 MHz. А некоторые бренды и ранее выпускали компьютеры с "заводским разгоном". В таких случаях комплектующие сторонних производитетелй были не всегда совместимы (если не могли работать в разгоне). (Позже - поддержка 66 MHz осталась только на "серверных" PCI-64.)

@@cynic3859 There are many cards that had 66mhz capability. Even though they were 32 bit PCI. Silicon Image SATA cards for example are 32bit - 66mhz. And several video cards were capable as well. Including the VOODOO 3 and ARC Logic 1000PV PCI vga chip

Yes :D I always take a careful look actually, but sometimes I seem to be blind. As I finally saw that thing, I was laughing and sure, that everybody else would've seen it from the beginning.... But I think, this way it was more entertaining for everybody and this is all about that eventually :)

I had an EISA server with a DX50 for a while many years ago. It was quite a neat machine and seemed to have some pretty good performance. It seemed comparable to a dx2/66 in some things.

Yeah it was around S478 and 775 that the counterfeit cap solution from china infested the market, the stuff before that was built like a tank.

Thank you! Wish you a Merry Christmas too!

Merry Christmas to you too!

Das waren damals schon geile und spannende Zeiten. Was habe ich da noch herumgebastelt und ausgetauscht. Heute habe ich einen i5 4670K mit 16 GB seit sieben Jahren. Klar, immer mal wieder SSDs hinzugefügt - letztens sogar eine NVMC in einem PCI-E-Slot (weil ich doch gerne bastele). Aber das Ding reicht mir immer noch für alles (inkl. Videoschnitt). Einzig 360 Grad-Bilder und -Videos lasse ich auf dem Handy rendern, weil die RX570, die CPU und alle Lüfter auf Anschlag gehen und es trotzdem länger als auf meinem bald drei Jahre alten Redmi Note 5 dauert (das dabei nicht einmal handwarm wird). Wenn es nicht bald ARM-CPU-Mainboards gibt, wird der Nachfolger meines i5 ein M1. In diesen sieben Jahren hatte ich damals: 384/40, 468 DX2/66 (@80), An5x86-133, AMD K6-166 (@200 (2x 100 MHz)), AMD K6-2 400 (@500), Athlon 1.300, Athlon XP 1600+ (@1,66 GHz). Wahnsinn, wie viel Leistung zwischen dem 386/40 und dem XP liegen. Mir juckt es nicht in der Hose, wenn ich an einen Mittelklasse Ryzen denke. Schöne CPUs, aber uninteressant für mich. Was ich als PC-Spiele in diesen sieben Jahren (und danach) an Geld für Grafikkarten ausgegeben habe, passt auf keine Kuhhaut. Das waren viel mehr Grafikkarten als Prozessoren. Nach dem XP wechselte ich 2004 von Windows zu OSX (Powerbook 17" G4 1,33), dann zu einem C2D 4300 (@2200 MHz) (Hackintosh) und dann eben erst zum i5 4670K, den ich bisher nur zum Spaß übertaktete, aber sonst im Default laufen lasse. Vom Hackintosh bin ich jetzt ein Jahr weg. Ubuntu bietet alles, was ich benötige. Nur für FinalCut starte ich noch den Hackitosh, ganze selten mal Windows für alte Spiele. Ich zocke nur noch auf einer alten Xbox 360, die Switch benutze ich gar nicht. Sorry für das längliche Posting. Liegt wohl am Jahresende, dass man mehr über die Vergangenheit nachdenkt.

Moin zurück! Ja das ganze scheint jetzt etwas mehr Beachtung zu bekommen, als sonst.

Thank you! And wish you a Merry Christmas as well!

Yes sometimes it needs to be done. Especially on the P4 and Athlon the caps were under very heavy load and tend to go bad a lot faster there.

@@necro_ware I hope it is. Board worked for half an hour and then died on my eyes. Second one randomly forgets settings, crashes on memory count or reports BIOS checksum invalid and starts in BIOS recovery mode. On the first one I thought it was CR2032 or RTC crystal (your video about interrupts inspired me for this investigation), but scope said "it is OK" and fresh battery didn't help either. PDF from ICH5R says everything including CMOS RAM is integrated so I hope it is about bad caps 🙌. Also funny thing that Watchdog randomly shuts down power to that first board. I replaced any other component PSU, CPU, RAM, GPU... Nothing else left ;)

Yeah, I probably threw it out during editting. Some flux is always a good idea.

It wasn't a fuse, and should not have been replaced with a fuse. The incorrect substitution of a resistor with a fuse may be contributing to the CPU voltage being incorrect.

Such power transistors in the linear voltage regulator circuits can get really hot and on this board there is not much of a heat sink except of the copper plane underneath. Due to the stress those transistors are notorious to give up one day. In my opinion this is the issue number two on the 486 mainboards right after the barrel battery. The burned trace behind of the CPU is may be a sign of a wrong CPU insertion, I hope, that is not too complicated to fix with a jumper wire. With only 20 Ohms you certainly should also check out that transistor indeed.

Yes meanwhile I also think, that it is a current sense resistor indeed. However, since I replaced it with a 3A fuse it would limit the current in a hard way :D I guess, I will have to search for a proper replacement...