I had that motherboard! Got it in 1999 when my Amptron motherboard died. I still remember that Friday very well when my dad drove me and my broken P200MMX to the computer shop and we were even invited to the workshop while the shop owner swapped in the GA-5AA.
Funny timing. I do have the exact same motherboard to repair. Southbridge brim is cracked on the sides and shorted. As it is the chip handling the poweron, it doesn't turn on. I'll tackle the repair once I get hold of a microscope. Thanks for the remotivational video :)
I fully agreed a friend gave me this motherboard, I repaired it, but those X O X O (lol) confused me quite a bit, they made me be very careful when changing settings to try different CPUs. I sent theretroweb some better images. the second one and the last are mine, good those helped Necroware on this vdeo
@@adagioleopard6415 Yup, familiar with that convention, but I was giving them the benefit of the doubt on this one. I can see where they were going with X. But zero for "closed"? That makes no sense to me.
@@necro_ware A year ago a friend gave me this motherboard, exactly this revission, which he had stored for 20 years. The capacitors were almost dead, but except a couple, hadn't leaked yet, so I replaced most of them. But I got into trouble with one that when I soldered it, the tin hid what the hole really looked like and as it was a little harder to get the older one out from the board (those big ground plates around the power section eat the heat away ) the pad was slightly damaged. I forgot to take detailed pictures before starting the job. I had to look for all the possible ones on the web. Then someone sent me some detailed ones. In the end, everything ended well. I even upgraded the BIOS (to F6 version, not the last one F7b BETA as you ) to fix some RAM compatibility issues. I got some help on Vogons, look there for "Help with a Gigabyte GA-5AA rev 2.2". I also sent theretroweb some better images. the second one and the last are mine. good my pictures helped you.
@@mikes989 No worries, I can see all your comments. Seems like YT has issues to refresh the view properly, but your comments about the board which you got from a friend and the pictures on TRW are perfectly visible ;) Thank you.
Ahh... Super Socket 7. Of all the pcs I built, it was the platform I built more of than any other... Those were Super Fun days for me... Nice, as always...
I upgraded my mother's computer with one of these boards in about 2004. It ran Window XP pretty well, though less well as time went on with the bloat that came with the service packs.
There is always something special about SuperSocket 7 with its fancy AMD processors, having a win95 logo and stuff. And is certantly a good board. Gigabyte, ALI chipset. Nice find and good job reviving it.
Super socket 7 is my favorite. My first pc was a IBM with the SIS530 chipset. I miss that pc. I had it with 3 256 MB ram sticks and a K6-2 at 550mhz. Windows 98 was rock solid on that rig.
Thank you so much for this video! The other day I gave up as I tried to get a mainboard PCChips M559 up and running. It started to post, but stopped instantly with a code D3D1. Of course I did not insert memory modules, since it should indicate the missing modules later. But as I saw your video, I instantly tried it again and it worked perfectly fine. Cool! Otherwise I had scrapped this great board.
Very cool motherboard. It has connections for ATX and AT power supplies. This board reminded me of my FIC VA503+ that I had with AMD K6II500. Congratulations on the video and success to the channel. Hugs from Brazil.
This board is absolutely wonderful. Runs my entire hoard of S7/SS7 processors and the most stable board I have. The only disappointment is that the Turbo connector toggles the Turbo LED but doesn't actually have turbo functionality.
I currently build a K6-3 system and have one of those boards where the CPU sits behind the PCI slots. But luckily its not a problem for me. The AWE64 and Voodoo3 are my longest cards and just sneak left/right to the CPU cooler corners :) I also use the SIMMconn module for AWE which extends its length but that is much higher then the cpu cooler so no problem as well. I had a Super7 PC back in the days with a TNT2 and i remember playing Quake 3 in 32-bit was absolutely fine and smooth. I still love this platform so much. I built everything from 286 to AM3+/Core-4xxx but Super7 is my absolutely favorite.
BIOS probably tries to update hardware configuration in ESCD at the end of the POST which is stored in Flash ROM page if I remember correctly. If motherboard can not write to the ROM chip (wrong voltage, unsupported internal configuration) it will hang or print an error about ESCD.
Beautiful gem of hardware. Seen once the Zida \ Tomato mobo with super socket 7 in 1998 and it was able to do its job. Some older games do not like NTFS and it is better sometimes to use something like a test board for aged software running. Thank you a lot
I am glad to see you are not of the mind re-cap for the sake of re-capping old hardware. There are a bunch of creators that annoy me by saying they are future proofing the hardware by replacing components where there is not sign of faults.
Normally i don't really like that especially when the caps some of these people put in instead are sort of garbage. I figure if the original ones lasted for 35 years with little measurable degradation, they can do that much again, i mean some of the Japanese caps in 80s computers seem indefinitely stable. But there's a very strong counter argument by James Lewis who is a former Kemet (capacitor manufacturer) executive and indeed himself a classic computer enthusiast. He mapped out and preserved numerous 80s computer schematics and does occasional classic computer repairs. In his view these are fundamentally decaying parts, and after 20 years, it's time for them to go. After all, we don't test them all the way to the forming voltage, which a new capacitor is validated to, so the gradual reduction of safe operating voltage is likely to be creeping in, and the caps can just fail suddenly someday soon. Given modern caps are both smaller in size and lower ESR, when replacing, you might as well up the voltage rating significantly, to arrive at similar dimensions and ESR as original and give them more life headroom, if you need to or choose to replace, in 80s computers in particular. Also if you notice one cap being dead, all caps of the same series and under similar load in a given product have to be taken out, just regardless of whether you subscribe to recap good or recap bad side of the fence, because as a rule, measurable degradation has already set in and you'll see it when you characterise them. There's also any number of products where the experience shows that the caps have to go, you don't keep SMD aluminium electrolytics of the 90s in most products if you don't have to, don't keep ones in Game Gear absolutely not, etc. If you know there's bad caps, i'd hate seeing them being kept for originality or whatnot.
Greetings! Good job resurrecting this board and its not undead! Ha! By the way, i stumbled upon boxes with all kinds of old (motherboard) connectors from power connectors to PCI and whatnot. Its all "used" / removed from boards i guess. I have absolutely no use for it, as i am more interested in other equipment. I just wanted the boxes so ... if you want, i could sent you the stuff for free (if you indeed live in germany). So i don't need to throw it away. Best wishes from near Berlin!
Moin! Es ist tatsächlich Deutschland :) Ersatzteile sind immer willkommen und in der Beschreibung des Kanals ist eine Email hinterlegt für solche Fälle. LG
Yeah, I'm just too much used to it. As I got my first 386, Norton Commander was already the "Standard". Without the autocompletion, which we have today in the command line, the prompt back then was a horror to use for navigation, so nobody did. Later I switched to Dos Navigator and use it still on my retro machines, but VC is a lot compacter and faster especially on 286 and XT, that's why I use it as a compromise on my CF cards in the workshop. One fits all so to speak.
@@worroSfOretsevraH I'm one of those crazy people :) To compare terminal today with the command line in DOS is like comparing a cart from ancient Rome with a modern car. Both have wheels, but somewhat very different.
The line "Windows 7 confirmed working on this motherboard..." caught my attention... somebody installed and got Windows 7 to run on a computer maxing out at 133MHz on the FSB and like 450 MHz on a processor (that I remember being obtainable)... I want to see Win7 boot on whatever configuration they got working on that board.
Nice video, as usual. May I ask - how this board handles powering of AGP? I suppose 3,3V are generated just by quite weak on board regulator, right? I have similar board (Gigabyte GA-5AX) in ATX form factor and even this board has AGP connected to regulator. Luckily in latest revision there is jumper to switch to 3,3V in ATX power connector instead. With on board regulator used, there is risk with more power hungry AGP cards to be damaged. I was wondering why even ATX only super socket 7 boards does not simply use 3,3V from ATX power suply and only reason I found is support of very old (IDT) 3,52V CPUs.
I have the ATX version (GA-5AX rev.4), mine has been fully recapped, but unfortunately it's not the most stable board in the world, especially with AGP cards. In the end I much prefer my Slot1 i440BX system.
This channel is damn fun! I maybe selfish and probably am, I don't want you to grow, lol. I've seen several channels devolve into bs that doesn't represent its continents. I know I'm selfish, but thank you! I love this space. God bless.
I loved SS7 boards. I had around 30 Socket 7 and Super Socket 7 boards at my job from 2000-2003 timeframe. Almost all of them got updated from P54, P55 and 6x86 chips to K6-2 400 and 450Mhz chips. Barely any of them had fsb frequencies above 100Mhz but those were my favorites. There was 1 board(I can't remember the model sadly) that I got to run stable with a 3x133mhz setup, but it didn't make a huge difference versus 4x100mhz. I'd still like to get my hands on a motherboard that supports higher fsb clocks though as I do have a K6-2/533 in my collection that runs fine at 550mhz that I would love to experiment with it. I still kick myself for never picking up a k6-3 back in the day though.
Nice looking board! My criticism with the layout would be the floppy/ide cables would have to stretch over the cpu, which would block natural venting to the back of the case. Maybe they thought ducting upward would be ok as it's going to the PSU, often the only bit of airflow for lower end PCs? I would still prefer ribbon cables towards the front of the case though.
I have similar board that behaves on POST exactly the same. No signs of life without ram gave me a headache one day. I also had to downgrade BIOS to make it work with CF card.
This board looks similar to my Asus P5A, same chipset, and the P5A also specifies a 2.0v option for the CPU but the implementation is bugged and setting 2.0v gives 0v and no POST. For the P5A, you've gotta jumper a pin across the voltage control chip so you can access the lower half of the voltage table - where the range is something like 1.6v to 2.0v, and THAT modified 2.0v setting actually works, enabling the K6-2+ ... or your other option is to just overvolt your K6-2+ to 2.1v. But I did the mod and it was a success.
The best BabyAT SS7 board by far. Should work with all Socket 5-SS7 CPU regardless of the BIOS version. Really annoyed that mine had a mosfet short circuit right after I had just gotten it working again. Ended up buying a Rev 1.0 Jetway J-542B (which uses the same VRM as the J-542A) as a stand-in until I can fix it, but due to all voltages being 0.4V higher than set I can't run my modded K6-2+ 570.
@necroware, any recommendations where to buy the BIOS? Also will be great a video regarding BIOS types and possible replacements. Tanks a lot for your great work!!!
@atheatos made a nice overview a while back. The chips you can buy on Ebay and the ROM images on the retro web project. ruclips.net/video/bQM4e7AwI0U/видео.html
I had a similar Super Socket 7 board when I was younger, though I unfortunately only had it for a few days since that was the board that taught me the hard way to always double-check pin connections and jumper settings. I don't remember exactly what I did but I know I fried a perfectly good AMD K6-2 and also damaged the board itself. it was pretty devastating at the time, so it feels especially nice to see one of them being repaired all these years later! maybe one of these days I'll get another Super Socket 7 board to play around with, but in the meantime I have a 486 board to work on instead (it's my first retro board repair and I'm really enjoying it so far)
I have this motherboard with severe stability issues. DOS runs great with Tillamook 120x3.5 for a nice 420 MHz but Windows 95/98 just won't run properly. Should look into recapping or if other repairs might be needed.
It rather started end of 90's. From around 2001 it got just more visible, because Athlon and P4 put much more stress on those caps and they started to fail a lot faster.
What is the brand of the caps you used? I checked LCSC for the reference and they seem to be... ROQANG? Is that a known good brand? I want to do some repairs but I'm always held back when it comes to buying capacitors online.
precisely because of the winbond's vpp requirement, is why it cannot write to nvram, if you look on post, with the sst chip it said update ok, with the winbond it said no nvram, the nvram is used for things like escd, pci/pnp configuration and such. it (pci/pnp) might work without that but I just wanted to share.
I'm surprised that it doesn't try to write ESCD data to the flash. I have socket 7 boards that display "Cannot write ESCD data" or even lock up on POST if the BIOS cannot write to the flash chip. Maybe AMI behaves different than Award in this regard.
Thank you your post! In case of a screwdriver damage at the edge of the chipset pcb what are the chance for chip survival? One via is definitely gone. DO this mini pcb holding the chip routing traces between the top and bottom layers?
@@necro_ware Thank you your quick reply. You are lucky not having 4 dash on the post card at the start. Usually I have this situation. In unlucky case the chipset manual is not reacheable, there is not much to do. In lucky case the chipset manual explains everything EXCEPT what and in which sequence happens when the chip is getting power. If somebody knows a method in general what steps should be taken in order to determine the root cause of the 4 dash sign seen on the postcard than please share with us. Let it assume we closed out the obivous reasons like broken traces, oscillator, corrosion, power rail, etc. What left is the chipset: what they do in the first one-two seconds after the switch on? How can we debug these steps? I believe the socket 7 chipsets work the same way regardless of the producer. Some has extra features but the first two second should be almost the same for these chips. This two second event sequence is somewhat burried well. I hope I am blind and somebody is able to show me light.
Sharing one experience - Updating BIOS using that Uniflash DOS command may not verify the content using T48 Universal Programmer with same BIN file! I faced issue twice.
In an upcoming video, can you explain how the 16bit part of an isa slot works? I have a 486 mainboard that none of the 16 bit parts work but the 8 bit parts do and it has me stumped!
Works basically in the same way as the lower 8-bits. Check the buffers/tranceivers on the data and address bus. There is usually LS245 nearby. I think I already mentioned it in some of my videos, but don't ask which one :)
Hi. I have two Asus P2B-DS boards to repair. One does not start the power supply at all, the other one starts the power supply but generally does not work. Are you able to guide me in repairing it?
Hi, for such cases I suggest to ask on vogons or other forums in your area. Usually the community is very helpful and has a lot of ideas. Much better, than to ask one random guy.
Have to disagree with the board layout, for the time (typical white box build) and probably today the IDE/Floppy ribbon cable connections location wouldn't help with airflow to the CPU cooler and the said cable runs messy with the standard ribbon cables that would have been supplied with these MBs. Think I too have experienced that ROM/CF issue, but I just thought it was a compatibility issue with the CF card/adapter as it worked with a standard IDE drive... makes me wonder.
You have a modern perspective on it. That wasn't an issue back then. This CPU can run for couple of minutes even without a cooler at all :) Airflow issues first appeared with later generations of CPUs. Back then PSUs had only one 80mm fan on the back and no opening inside towards the CPU, so the ribbon cables wouldn't cover anything. For graphics cards however it just started to be an issue, which often was solved with an in-slot fan between the expansion cards.
Hi,Necro you should enable all IDE in the bios even if you don't use them,I noticed it stopped booting after I disabled one in BIOS on a GA 6VEML Socket 370 ,I disabled the secondary IDE and it got stuck on detecting IDE drives,I went back to BIOS enabled it and it worked fine.
That could be a BIOS bug or an issue with the controller. I deactivate often what I don't need to increase boot speed and reduce resource usage, which was very important in DOS.
@@necro_wareIf you still have the BIOS could you please check if it makes a difference? I'm using the latest BIOS also and it might be a beta one. My mobo doesn't have a Agp and I managed to squeeze some performance overall by disable the onboard sound, adding a pci one but that one also cripples the 66 FSB Celeron coppermine 600,using a S3 Savage4 PCI and a Voodoo 2 3000 with 512MB of SDRAM, also I disabled the USB and Lan, Modem and then at IDE got stuck with the boot
Your little component tester complains about not being calibrated. Just gotta tie all the pins together, let it check itself, then put in a capacitor in the value range it asks.
I saw the information on inet that Ali Alladin 5 chipset don't working with Riva TNT card, can you check this? And my Acorp mobo with Ali Alladin 5 + chipset don't detect and work with Trident 3DImage 9750 PCI video card at all. But AGP variant of this card detect and work well.
Back in the days I had this problem on my Aladin 5 and TNT, random crashes. When internet was not a thing, my only solution was to swith to full Intel platform (440BX and P2-400) instead of AMD and Ali chipset (Asus motherboard and K6-2-400). 😞 Than, no probem at all!! 😀
I have asus p2-99b board ( AT , 440ZX) that doesn't works with AT power supplies ( tried couple of 100% working ones) but it works with ATX psu. What it might be ?
First of all double check all jumpers. If everything's fine check the 3.3V regulator. ATX has 3.3V supply and doesn't rely on the regulator on board. AT doesn't provide 3.3V and an additional regulator on the board has to do the job.
@@necro_ware 3.3v regulator was my first thought and i checked it with multimeter. This mb uses it anyway(atx/at mode) and there is even jumper for overclocking 3.3-3.45-3.6v.
I don't know if Intel CPUs FPU were so so sooo much better than others, but, Quake was heavily optimized even in assembler for the Pentium FPU architecture, of course it performs better on the targeted CPUs. K6-2 added 3dnow instructions which Quake was never optimized for, for example. I mean, it is not a fair comparison. Best wishes.
With all respect to K6-2 and its outstanding integer performance, unfortunately it really lost every competition against Intel when it was about FPU. Leaving alone 3DNow!, since Quake was also not optimized for MMX, Intels Pentium FPU was ahead of every competition and it was beaten first with Athlon.
The super socket to rule all sockets! Really, is there any single type of socket that can run as many different kinds, makes and types? Not just the Intel, Cyrix, AMD Pentium-level, but also Rise and some others I can't recall now. As for AMD, the K5, K6, K6-II, K6-III and so on. Edit: I love the repair but I was kind of scared when I saw the 2.7V core voltage... if the switches were the other way round, did that poor P5-166 get that 5.5x multiplier? People would do lots of nasty things for a 366Mhz Pentium, but... it wasn't on the table. Or at least not without smoke and fire.
P5 cannot get 5,5x multiplier, Pentiums used only BF0 and BF1 pins, so only 1,5x, 2x, 2,5x and 3x multipliers were possible. And 3,5x for Pentium MMX which was the same as 1,5x. Higher multipliers using BF2 pin were possible for AMD only.
The recipe for electrolytic capacitors was stolen at the time, but the stolen recipe was falsified as a honey pot. That was the rumor I heard at the time.
Yeah, back then Linux was everything else, but mature. We had to choose wisely which hardware to use if we wanted to go with Linux. This is even still valid today in some rare cases, but is nothing compared to that days.
I didn't mess with ALI back then, but I did have a VIA chipset board, and that kicked off my loathing of VIA. I would say that Intel chipsets are what kept me from even considering an AMD CPU, because the chipset options on Team Green (at the time) were all awful. It wasn't until the late Pentium III era when I gave VIA another chance, because those Epia mini-ITX boards with the C3 CPUs were just too cute. And they're actually pretty alright!
I personally wouldn't be so daring to plug in a CPU with caps in that condition, after all the output of a switch mode voltage regulator can only be stable with output caps vagely to spec, and the prior unloaded voltage test also tells you as much. Indeed i wouldn't even call it unloaded, since the caps in that condition also exhibit leakage, equivalent parallel resistance, so the output was already loaded down by the caps and was still jumping all over the place. Things are potentially worse with ripple than a multimeter is capable of indicating.
![[#2024MAMA] G-DRAGON - 무제(Untitled, 2014)+POWER+HOME SWEET HOME+뱅뱅뱅+FANTASTIC BABY | Mnet 241123 방송](http://i.ytimg.com/vi/Ox29z5Nf1Uk/mqdefault.jpg)
I had that motherboard! Got it in 1999 when my Amptron motherboard died. I still remember that Friday very well when my dad drove me and my broken P200MMX to the computer shop and we were even invited to the workshop while the shop owner swapped in the GA-5AA.
The CPU can probably run at 100*2.5 on that board.
Yep those are the memories you'll never forget.
Sweet 90's
Always joy when a vintage board comes back to life 🙂
A Bits und Bolts and Necroware video in the same day!!! Woo hoo!!! I love this board. I want to see how far you can push it!
Funny timing. I do have the exact same motherboard to repair. Southbridge brim is cracked on the sides and shorted. As it is the chip handling the poweron, it doesn't turn on. I'll tackle the repair once I get hold of a microscope. Thanks for the remotivational video :)
Theres a special place in hell for the guy who used X and O as 0 and 1.
I fully agreed
a friend gave me this motherboard, I repaired it, but those X O X O (lol) confused me quite a bit, they made me be very careful when changing settings to try different CPUs.
I sent theretroweb some better images. the second one and the last are mine, good those helped Necroware on this vdeo
absolutely :)
Yeah, what were they thinking? X is nothing, okay... sure. But zero? That's your choice? Which one is supposed to be "no jumper" here?!
@@nickwallette6201 x in electronics means "Dont care"
Why the hell would youbkame that zero?
@@adagioleopard6415 Yup, familiar with that convention, but I was giving them the benefit of the doubt on this one. I can see where they were going with X. But zero for "closed"? That makes no sense to me.
Nice board. It would be great to see it upgraded to 1M of cache. There is no specific reason for it, but just because you can.
Yeah, may be in the future. I have to find 1MB SRAM chips, they are less common.
@@necro_ware A year ago a friend gave me this motherboard, exactly this revission, which he had stored for 20 years. The capacitors were almost dead, but except a couple, hadn't leaked yet, so I replaced most of them. But I got into trouble with one that when I soldered it, the tin hid what the hole really looked like and as it was a little harder to get the older one out from the board (those big ground plates around the power section eat the heat away ) the pad was slightly damaged. I forgot to take detailed pictures before starting the job. I had to look for all the possible ones on the web. Then someone sent me some detailed ones. In the end, everything ended well. I even upgraded the BIOS (to F6 version, not the last one F7b BETA as you ) to fix some RAM compatibility issues.
I got some help on Vogons, look there for "Help with a Gigabyte GA-5AA rev 2.2".
I also sent theretroweb some better images. the second one and the last are mine. good my pictures helped you.
@@necro_ware Necroware.. my comments are not appearing..... why?
Luckily my board has two of these 512M chips ^^
@@mikes989 No worries, I can see all your comments. Seems like YT has issues to refresh the view properly, but your comments about the board which you got from a friend and the pictures on TRW are perfectly visible ;) Thank you.
There weren't many BAT Super Socket 7 Mobo's on the market at the time.
Thanks for sharing and greetings from Germany
Love this board, been using one with a K6-3+ 450mhz every week since 2015 for retro gaming. Solid as a rock.
Ahh... Super Socket 7. Of all the pcs I built, it was the platform I built more of than any other... Those were Super Fun days for me...
Nice, as always...
Exactly what I need for today. Thanks a lot.
I upgraded my mother's computer with one of these boards in about 2004. It ran Window XP pretty well, though less well as time went on with the bloat that came with the service packs.
this board is fantastic. also works flawlessly with linux distributions of its era
Excellent work, I appreciate the attention to detail shown by replacing the RAM module lever. That would definitely bug me!
5:39 until this point I honestly thought I was the idiot not understanding X and 0 on this board. Thanks for pushing me up!
Love it! A Super Socket 7 board is on my wish list for retro systems.
There is always something special about SuperSocket 7 with its fancy AMD processors, having a win95 logo and stuff. And is certantly a good board. Gigabyte, ALI chipset. Nice find and good job reviving it.
Brilliant work, amazing video Thx!
I like that this board is so small. Nice work!
This reminds me that i have to get an EEPROM chip for my P5A-B to get it up and running once again.
Well done on the board as it looks like new! Thank you for sharing your journey with this one 👍
Great video as always sir.
Super socket 7 is my favorite. My first pc was a IBM with the SIS530 chipset. I miss that pc. I had it with 3 256 MB ram sticks and a K6-2 at 550mhz. Windows 98 was rock solid on that rig.
Amazing! SS7 is truly the most flexible DOS platform.
Thank you so much for this video! The other day I gave up as I tried to get a mainboard PCChips M559 up and running. It started to post, but stopped instantly with a code D3D1. Of course I did not insert memory modules, since it should indicate the missing modules later. But as I saw your video, I instantly tried it again and it worked perfectly fine. Cool! Otherwise I had scrapped this great board.
Very cool motherboard. It has connections for ATX and AT power supplies. This board reminded me of my FIC VA503+ that I had with AMD K6II500. Congratulations on the video and success to the channel. Hugs from Brazil.
This board is absolutely wonderful. Runs my entire hoard of S7/SS7 processors and the most stable board I have. The only disappointment is that the Turbo connector toggles the Turbo LED but doesn't actually have turbo functionality.
Yep, the board is great. Still have 4 of these. all in Rev. 2.2. But I think the Epox EP-MVP3G5 is a little better. Thanks for the video
Vos vidéos sont très satisfaisantes; c'est toujours un plaisir de les regardés 😍❤️👍
I currently build a K6-3 system and have one of those boards where the CPU sits behind the PCI slots. But luckily its not a problem for me. The AWE64 and Voodoo3 are my longest cards and just sneak left/right to the CPU cooler corners :) I also use the SIMMconn module for AWE which extends its length but that is much higher then the cpu cooler so no problem as well. I had a Super7 PC back in the days with a TNT2 and i remember playing Quake 3 in 32-bit was absolutely fine and smooth. I still love this platform so much. I built everything from 286 to AM3+/Core-4xxx but Super7 is my absolutely favorite.
I'm loving your channel!!!!
Doy like porque se que sera un video 10/10
Thanks for the video. Hope to see more.
BIOS probably tries to update hardware configuration in ESCD at the end of the POST which is stored in Flash ROM page if I remember correctly. If motherboard can not write to the ROM chip (wrong voltage, unsupported internal configuration) it will hang or print an error about ESCD.
Yeah, probably that's it.
Love these videos.
The designer of this board was clearly a fan of Tic-Tac-Toe!
FYI, disabling the IDE autodetection for secondary master & slave does not disable the secondary controller :-p
I know, it's just for faster booting and I set it to auto, because I have multiple CF cards with various test setups.
What EEPROM chips do you use for these older boards?
Can you provide various makes/models that are compatible?
Thanks.
Beautiful gem of hardware. Seen once the Zida \ Tomato mobo with super socket 7 in 1998 and it was able to do its job. Some older games do not like NTFS and it is better sometimes to use something like a test board for aged software running. Thank you a lot
I love those tiny Tomato boards. They're just adorable.
Love this mobo , Remind me 1998 build my k6-2 450 mhz
I am glad to see you are not of the mind re-cap for the sake of re-capping old hardware. There are a bunch of creators that annoy me by saying they are future proofing the hardware by replacing components where there is not sign of faults.
Normally i don't really like that especially when the caps some of these people put in instead are sort of garbage. I figure if the original ones lasted for 35 years with little measurable degradation, they can do that much again, i mean some of the Japanese caps in 80s computers seem indefinitely stable.
But there's a very strong counter argument by James Lewis who is a former Kemet (capacitor manufacturer) executive and indeed himself a classic computer enthusiast. He mapped out and preserved numerous 80s computer schematics and does occasional classic computer repairs. In his view these are fundamentally decaying parts, and after 20 years, it's time for them to go. After all, we don't test them all the way to the forming voltage, which a new capacitor is validated to, so the gradual reduction of safe operating voltage is likely to be creeping in, and the caps can just fail suddenly someday soon.
Given modern caps are both smaller in size and lower ESR, when replacing, you might as well up the voltage rating significantly, to arrive at similar dimensions and ESR as original and give them more life headroom, if you need to or choose to replace, in 80s computers in particular.
Also if you notice one cap being dead, all caps of the same series and under similar load in a given product have to be taken out, just regardless of whether you subscribe to recap good or recap bad side of the fence, because as a rule, measurable degradation has already set in and you'll see it when you characterise them. There's also any number of products where the experience shows that the caps have to go, you don't keep SMD aluminium electrolytics of the 90s in most products if you don't have to, don't keep ones in Game Gear absolutely not, etc. If you know there's bad caps, i'd hate seeing them being kept for originality or whatnot.
Greetings! Good job resurrecting this board and its not undead! Ha! By the way, i stumbled upon boxes with all kinds of old (motherboard) connectors from power connectors to PCI and whatnot. Its all "used" / removed from boards i guess. I have absolutely no use for it, as i am more interested in other equipment. I just wanted the boxes so ... if you want, i could sent you the stuff for free (if you indeed live in germany). So i don't need to throw it away.
Best wishes from near Berlin!
Moin! Es ist tatsächlich Deutschland :) Ersatzteile sind immer willkommen und in der Beschreibung des Kanals ist eine Email hinterlegt für solche Fälle. LG
Still love how you use VC;) When most people like to use commands and waste time never got that;)
Yeah, I'm just too much used to it. As I got my first 386, Norton Commander was already the "Standard". Without the autocompletion, which we have today in the command line, the prompt back then was a horror to use for navigation, so nobody did. Later I switched to Dos Navigator and use it still on my retro machines, but VC is a lot compacter and faster especially on 286 and XT, that's why I use it as a compromise on my CF cards in the workshop. One fits all so to speak.
@@necro_ware I use VC too, always used it after NC :) It works also good in 80x50 mode which i normally prefer.
@@necro_ware I still don't understand the Linux people. They all use bare terminal for everything. Crazy!
@@worroSfOretsevraH I'm one of those crazy people :) To compare terminal today with the command line in DOS is like comparing a cart from ancient Rome with a modern car. Both have wheels, but somewhat very different.
@@necro_ware Why not use Midnight Commander, or something similar than?
Thanks for the another great video!
The line "Windows 7 confirmed working on this motherboard..." caught my attention... somebody installed and got Windows 7 to run on a computer maxing out at 133MHz on the FSB and like 450 MHz on a processor (that I remember being obtainable)... I want to see Win7 boot on whatever configuration they got working on that board.
I'd love to see a video showing how far this board can be pushed.
Thank you, I'll put it on my todo list.
Nice video, as usual. May I ask - how this board handles powering of AGP? I suppose 3,3V are generated just by quite weak on board regulator, right? I have similar board (Gigabyte GA-5AX) in ATX form factor and even this board has AGP connected to regulator. Luckily in latest revision there is jumper to switch to 3,3V in ATX power connector instead.
With on board regulator used, there is risk with more power hungry AGP cards to be damaged. I was wondering why even ATX only super socket 7 boards does not simply use 3,3V from ATX power suply and only reason I found is support of very old (IDT) 3,52V CPUs.
Nice video! Thanks.
I have the ATX version (GA-5AX rev.4), mine has been fully recapped, but unfortunately it's not the most stable board in the world, especially with AGP cards. In the end I much prefer my Slot1 i440BX system.
This channel is damn fun! I maybe selfish and probably am, I don't want you to grow, lol. I've seen several channels devolve into bs that doesn't represent its continents. I know I'm selfish, but thank you! I love this space. God bless.
I'm kind of surprised that he has so few subscribers, with the quality of content he deserves many more.
I loved SS7 boards. I had around 30 Socket 7 and Super Socket 7 boards at my job from 2000-2003 timeframe. Almost all of them got updated from P54, P55 and 6x86 chips to K6-2 400 and 450Mhz chips. Barely any of them had fsb frequencies above 100Mhz but those were my favorites. There was 1 board(I can't remember the model sadly) that I got to run stable with a 3x133mhz setup, but it didn't make a huge difference versus 4x100mhz. I'd still like to get my hands on a motherboard that supports higher fsb clocks though as I do have a K6-2/533 in my collection that runs fine at 550mhz that I would love to experiment with it. I still kick myself for never picking up a k6-3 back in the day though.
Good job well done your amazing
Nice looking board! My criticism with the layout would be the floppy/ide cables would have to stretch over the cpu, which would block natural venting to the back of the case. Maybe they thought ducting upward would be ok as it's going to the PSU, often the only bit of airflow for lower end PCs? I would still prefer ribbon cables towards the front of the case though.
I have similar board that behaves on POST exactly the same. No signs of life without ram gave me a headache one day. I also had to downgrade BIOS to make it work with CF card.
This board looks similar to my Asus P5A, same chipset, and the P5A also specifies a 2.0v option for the CPU but the implementation is bugged and setting 2.0v gives 0v and no POST. For the P5A, you've gotta jumper a pin across the voltage control chip so you can access the lower half of the voltage table - where the range is something like 1.6v to 2.0v, and THAT modified 2.0v setting actually works, enabling the K6-2+ ... or your other option is to just overvolt your K6-2+ to 2.1v. But I did the mod and it was a success.
The go to board in the day, v2 had ata133
I have one like this with an SMD bit (capacitor?) that looks blown. I am scared to turn it on
if the capacitor is not shorted in most cases it will even work.
The best BabyAT SS7 board by far. Should work with all Socket 5-SS7 CPU regardless of the BIOS version. Really annoyed that mine had a mosfet short circuit right after I had just gotten it working again.
Ended up buying a Rev 1.0 Jetway J-542B (which uses the same VRM as the J-542A) as a stand-in until I can fix it, but due to all voltages being 0.4V higher than set I can't run my modded K6-2+ 570.
Thanks for another great video. Do you have a link to where you buy your EPROMs/EEPROMs from?
Ebay and usual suspects.
Nice 👍 what eprom number do you use for pc mainboards?
@necroware, any recommendations where to buy the BIOS?
Also will be great a video regarding BIOS types and possible replacements.
Tanks a lot for your great work!!!
@atheatos made a nice overview a while back. The chips you can buy on Ebay and the ROM images on the retro web project.
ruclips.net/video/bQM4e7AwI0U/видео.html
I had a similar Super Socket 7 board when I was younger, though I unfortunately only had it for a few days since that was the board that taught me the hard way to always double-check pin connections and jumper settings. I don't remember exactly what I did but I know I fried a perfectly good AMD K6-2 and also damaged the board itself. it was pretty devastating at the time, so it feels especially nice to see one of them being repaired all these years later!
maybe one of these days I'll get another Super Socket 7 board to play around with, but in the meantime I have a 486 board to work on instead (it's my first retro board repair and I'm really enjoying it so far)
I find it strange the voltage switches for Auto and 2.8V are the same. 🙂
Yes, I was confused in the same way. Don't know why they did that.
I would love to see how/if this board would work at 133fsb 2.5x multiplier. The bump in fsb was the best performance boost at this time.
As a big fan of baby-AT this is perfect for me, but that poor slot 1 board in baby-AT formfactor broke my heart 😢 ❤ 😂
Strange that it didn't come with a battery holder, didn't notice an empty socket for a Dallas chip, just one for the EEPROM.
Nice board....
I have this motherboard with severe stability issues. DOS runs great with Tillamook 120x3.5 for a nice 420 MHz but Windows 95/98 just won't run properly. Should look into recapping or if other repairs might be needed.
Capacitor plague in 1999? Didn't it start slightly later, approx. 2001..2002?
It rather started end of 90's. From around 2001 it got just more visible, because Athlon and P4 put much more stress on those caps and they started to fail a lot faster.
What is the brand of the caps you used? I checked LCSC for the reference and they seem to be... ROQANG? Is that a known good brand? I want to do some repairs but I'm always held back when it comes to buying capacitors online.
They (C3299) were quite reliable so far and very inexpensive, but if you want to spend more go for Panasonic.
Finally a self explanatory error code -> d3d 🙃
Both of my K6-3+ cpus are 1.6V parts so it is difficult to find a motherboard where you dont need to have over voltage of 2.0V with them.
First day at work,. april 2000, in PC service, I've build a PC with k6 on that mobo :)
I have the same board and can confirm X=OFF and O=ON
precisely because of the winbond's vpp requirement, is why it cannot write to nvram, if you look on post, with the sst chip it said update ok, with the winbond it said no nvram, the nvram is used for things like escd, pci/pnp configuration and such. it (pci/pnp) might work without that but I just wanted to share.
I'm surprised that it doesn't try to write ESCD data to the flash. I have socket 7 boards that display "Cannot write ESCD data" or even lock up on POST if the BIOS cannot write to the flash chip. Maybe AMI behaves different than Award in this regard.
Thank you your post! In case of a screwdriver damage at the edge of the chipset pcb what are the chance for chip survival? One via is definitely gone. DO this mini pcb holding the chip routing traces between the top and bottom layers?
Yes, those are usually very fragile multi layer PCBs. So far I unfortunately had no luck trying to restore it.
@@necro_ware Thank you your quick reply. You are lucky not having 4 dash on the post card at the start. Usually I have this situation. In unlucky case the chipset manual is not reacheable, there is not much to do. In lucky case the chipset manual explains everything EXCEPT what and in which sequence happens when the chip is getting power. If somebody knows a method in general what steps should be taken in order to determine the root cause of the 4 dash sign seen on the postcard than please share with us. Let it assume we closed out the obivous reasons like broken traces, oscillator, corrosion, power rail, etc. What left is the chipset: what they do in the first one-two seconds after the switch on? How can we debug these steps? I believe the socket 7 chipsets work the same way regardless of the producer. Some has extra features but the first two second should be almost the same for these chips. This two second event sequence is somewhat burried well. I hope I am blind and somebody is able to show me light.
Try to run it at max FSB! Always wanted to see a K6-2+ at 133mhz FSB and how it would compare to other CPUs
Hello friend, a question, I have a Foxconn CN15235 motherboard saved, it is old with the INTEL05 E2168 processor.
❤❤❤❤❤❤❤❤❤❤❤❤❤
Imnsho, replacing caps on a 20-40yo mobo is like changing the motor oil, its probably ok but its a easy/cheap service to do.
Sharing one experience -
Updating BIOS using that Uniflash DOS command may not verify the content using T48 Universal Programmer with same BIN file!
I faced issue twice.
In an upcoming video, can you explain how the 16bit part of an isa slot works? I have a 486 mainboard that none of the 16 bit parts work but the 8 bit parts do and it has me stumped!
Works basically in the same way as the lower 8-bits. Check the buffers/tranceivers on the data and address bus. There is usually LS245 nearby. I think I already mentioned it in some of my videos, but don't ask which one :)
Hi. I have two Asus P2B-DS boards to repair. One does not start the power supply at all, the other one starts the power supply but generally does not work. Are you able to guide me in repairing it?
Hi, for such cases I suggest to ask on vogons or other forums in your area. Usually the community is very helpful and has a lot of ideas. Much better, than to ask one random guy.
Have to disagree with the board layout, for the time (typical white box build) and probably today the IDE/Floppy ribbon cable connections location wouldn't help with airflow to the CPU cooler and the said cable runs messy with the standard ribbon cables that would have been supplied with these MBs.
Think I too have experienced that ROM/CF issue, but I just thought it was a compatibility issue with the CF card/adapter as it worked with a standard IDE drive... makes me wonder.
You have a modern perspective on it. That wasn't an issue back then. This CPU can run for couple of minutes even without a cooler at all :) Airflow issues first appeared with later generations of CPUs. Back then PSUs had only one 80mm fan on the back and no opening inside towards the CPU, so the ribbon cables wouldn't cover anything. For graphics cards however it just started to be an issue, which often was solved with an in-slot fan between the expansion cards.
Hi,Necro you should enable all IDE in the bios even if you don't use them,I noticed it stopped booting after I disabled one in BIOS on a GA 6VEML Socket 370 ,I disabled the secondary IDE and it got stuck on detecting IDE drives,I went back to BIOS enabled it and it worked fine.
That could be a BIOS bug or an issue with the controller. I deactivate often what I don't need to increase boot speed and reduce resource usage, which was very important in DOS.
@@necro_wareIf you still have the BIOS could you please check if it makes a difference? I'm using the latest BIOS also and it might be a beta one. My mobo doesn't have a Agp and I managed to squeeze some performance overall by disable the onboard sound, adding a pci one but that one also cripples the 66 FSB Celeron coppermine 600,using a S3 Savage4 PCI and a Voodoo 2 3000 with 512MB of SDRAM, also I disabled the USB and Lan, Modem and then at IDE got stuck with the boot
But why you did't try MMX on 100Mhz FSB?)
Your little component tester complains about not being calibrated. Just gotta tie all the pins together, let it check itself, then put in a capacitor in the value range it asks.
I saw the information on inet that Ali Alladin 5 chipset don't working with Riva TNT card, can you check this? And my Acorp mobo with Ali Alladin 5 + chipset don't detect and work with Trident 3DImage 9750 PCI video card at all. But AGP variant of this card detect and work well.
I think, I don't have any Riva TNT, but I have TNT2, which I could test.
Back in the days I had this problem on my Aladin 5 and TNT, random crashes.
When internet was not a thing, my only solution was to swith to full Intel platform (440BX and P2-400) instead of AMD and Ali chipset (Asus motherboard and K6-2-400). 😞
Than, no probem at all!! 😀
I have asus p2-99b board ( AT , 440ZX) that doesn't works with AT power supplies ( tried couple of 100% working ones) but it works with ATX psu. What it might be ?
First of all double check all jumpers. If everything's fine check the 3.3V regulator. ATX has 3.3V supply and doesn't rely on the regulator on board. AT doesn't provide 3.3V and an additional regulator on the board has to do the job.
@@necro_ware 3.3v regulator was my first thought and i checked it with multimeter. This mb uses it anyway(atx/at mode) and there is even jumper for overclocking 3.3-3.45-3.6v.
I don't know if Intel CPUs FPU were so so sooo much better than others, but, Quake was heavily optimized even in assembler for the Pentium FPU architecture, of course it performs better on the targeted CPUs.
K6-2 added 3dnow instructions which Quake was never optimized for, for example.
I mean, it is not a fair comparison.
Best wishes.
With all respect to K6-2 and its outstanding integer performance, unfortunately it really lost every competition against Intel when it was about FPU. Leaving alone 3DNow!, since Quake was also not optimized for MMX, Intels Pentium FPU was ahead of every competition and it was beaten first with Athlon.
@@necro_ware as far as I know, MMX is only for integer ops
The super socket to rule all sockets! Really, is there any single type of socket that can run as many different kinds, makes and types? Not just the Intel, Cyrix, AMD Pentium-level, but also Rise and some others I can't recall now. As for AMD, the K5, K6, K6-II, K6-III and so on. Edit: I love the repair but I was kind of scared when I saw the 2.7V core voltage... if the switches were the other way round, did that poor P5-166 get that 5.5x multiplier? People would do lots of nasty things for a 366Mhz Pentium, but... it wasn't on the table. Or at least not without smoke and fire.
P5 cannot get 5,5x multiplier, Pentiums used only BF0 and BF1 pins, so only 1,5x, 2x, 2,5x and 3x multipliers were possible. And 3,5x for Pentium MMX which was the same as 1,5x. Higher multipliers using BF2 pin were possible for AMD only.
quake slow on a geforce 4???
That's software rendering. The graphics card has nothing to do except showing the image.
@@necro_ware ohh ok.. it was software mode...
The recipe for electrolytic capacitors was stolen at the time, but the stolen recipe was falsified as a honey pot. That was the rumor I heard at the time.
My gosh! There is a hidden motherboard bolt behind the CPU.
I am hopping you have a AGP Radeon HD 3850 to give the old Socket 7 to the max? LOL
Yeah :) I even do have it, however I believe it needs at least AGP 2.0 to run and wouldn't even fit into the slot on this board.
@@necro_ware ya the key is closer to the locking tab. I wonder what the most powerful AGP 2.0 / universal AGP compatible card is?
@@necro_ware Even if it was possible, the AGP voltage regulator on this board has limited power output.
@@xsc1000 Yes, this is what I meant. AGP 2.0 has lower voltage and higher current specification.
Maybe the Eeprom suffered from bitrot due to cosmic rays
Ali Aladin Chip, problems with Linux i remember, spezial Driver was needet for IDE Channels
Yeah, back then Linux was everything else, but mature. We had to choose wisely which hardware to use if we wanted to go with Linux. This is even still valid today in some rare cases, but is nothing compared to that days.
I didn't mess with ALI back then, but I did have a VIA chipset board, and that kicked off my loathing of VIA. I would say that Intel chipsets are what kept me from even considering an AMD CPU, because the chipset options on Team Green (at the time) were all awful.
It wasn't until the late Pentium III era when I gave VIA another chance, because those Epia mini-ITX boards with the C3 CPUs were just too cute. And they're actually pretty alright!
19:47 - after change of BIOS-Chip it tells no NVRAM
Yeah, already found another SST chip and now NVRAM is also fine. First one seems to be bad.
What is great with super Soket 7, is to overclock a Pentium 200 MMX to 250 MHz with 100 MHz FSB, to outperform P-II 233
I personally wouldn't be so daring to plug in a CPU with caps in that condition, after all the output of a switch mode voltage regulator can only be stable with output caps vagely to spec, and the prior unloaded voltage test also tells you as much. Indeed i wouldn't even call it unloaded, since the caps in that condition also exhibit leakage, equivalent parallel resistance, so the output was already loaded down by the caps and was still jumping all over the place. Things are potentially worse with ripple than a multimeter is capable of indicating.
That's true, but a little bit of risk makes it a bit more spicy ;)