|
|
IDT WinChip
Posts: 237
 
Location: USA | Looking forward to your Mirror's Edge LP. Looks like a really beautiful game. I might have to snag a copy.
Well the dual P3 is up and running. Contrary to what seems to be a general consensus that revisions prior to 1.06 cannot run 133Mhz bus, I was able to successfully get this board running at 133Mhz bus thanks to an old favorite Softfsb. While the 133Mhz setting ( jumper and softfsb ) for the ICS9150 generator actually operates at 111Mhz ( very odd number ) turns out using the ICS9250 profile in softfsb and choosing the 124/31Mhz option actually results in 133Mhz bus, or more accurately 132.7. Radeon 7500 seems to be taking the increased AGP clock speed well, though I will be trying some other cards in this system to see what will be an "optimal" video card for this system. Unreal Tournament does a very good job at dividing the workload among both processors. After playing Quake 3 for about 30 minutes, it suddenly kicked back to the desktop so I will have to figure out if it was something with the game as I was running the Quake 1 mod pack at the time, or if it could be a symptom of the radeon 7500 not quite being totally stable with the AGP bus speed.
As far as the Maxx is concerned, W98 only card for both GPU's to be enabled. ATI never could figure out how to make the Windows 2000 driver talk to both GPU's. I had forgotten about that.
And finally, some goodies I picked up for $10.
 (s-l1600.jpg)
Attachments
----------------
 s-l1600.jpg (342KB - 687 downloads)
|
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | Mirror's Edge is worth playing at least once.
The crash-to-desktop could be the video card having problems with the speed, but my 7000 tended to just lock up the machine. Still, that's not a rule, so it is probably worth looking into.
First thing I notice in the picture is the VIVO card, I nearly got one of those when I was initially building the OverDrive but it was overpriced and the S-2000 came up at about the same, so I got that instead. Still, if the whole lot was $10 you probably can't lose. That Slot Athlon alone could probably fetch more than that if it works.
I have good news and bad news regarding the problems I ran into. The bad news is that the recording of Mirror's Edge is on hold indefinitely. The good news is that I stripped the system down to find the problem and I think I have indeed found it;
It would not POST and though the fans spun there was no power LED. I took out the CPU and tested it in another motherboard, the board complains but it does POST so whilst I think the CPU may be unwell, it should start up. At this point I feared the worst and thought my motherboard had finally given in to old age, there is no way I can repair it anymore at this stage. Then I pulled out the RAM and installed it in the test board... The test board will not POST with that RAM installed so there is a good chance that this is the problem.
My only remaining concern is that the motherboard in the Pentium D actually is broken because although I was overclocking, I was not overvolting and the RAM is rated for 800MHz operation. With my overclock it would have been running at around 780MHz at the most (The motherboard only goes up to 666MHz without overclocking the FSB) and even then I wouldn't be surprised if the motherboard's firmware stepped in and started dividing the clock. On the other hand, the system had started crashing on occasion when memory heavy applications were running, as in a full lock up. As luck would have it I am always short on DDR2 anyway and it is currently dirt cheap, as are Pentium D 950 processors, so whilst I said I wouldn't repair the system any more I can't really begrudge £30 as I will use the RAM elsewhere if it doesn't work here and the processor would probably find a home eventually.
It looks like Wednesday / Thursday before the parts get to me, so here's hoping I can get it to run one last time. I don't think I will be overclocking it though. The PLL and VRM are a little unstable and I can't help thinking that perhaps they did something they shouldn't have done when I turned the clock up. It has been higher before and it was fine, but I now think it is simply too worn out to do this reliably.
In the mean time, I will start uploading both Banjo Kazooie and what I have of Mirror's Edge - only the final level is left to record, so I should either have the machine working by then or else have had time to play up to that point on another system to get it finished if I can't get the system to work. Still, if it has indeed died, I would rather it went out doing something useful like this than sat in the closed gathering dust, that would not be a fitting end at all. Unfortunately the RAM I had in the test motherboard really is faulty and whilst it can get the test board off the ground, it is not compatible with the Pentium D motherboard and I can't use the test board because... well... it's shit basically and I wouldn't trust it, it was never stable and that is why it was relegated to testing processors, RAM and hard drives in the first place.
Edited by DXZeff 2017-03-20 7:22 PM
|
|
| |
|
TM Crusoe
Posts: 671
| Some more hardware for my collection of 486 class stuff. This little one..
Might be a bit too expensive. Can't be that much anyway, and it should be industrial grade.
Anyway... 512mb is enough for a 486.
For tinkering, I am well on my way of piecing together the dx2-66 system.
As of now it is resting on the testbench. Perhaps I can source a case in the next month.
I am having that dual channel controller heading my way.
Then I have that DOM heading my way, as well as a 512mb Transcendal CF card.
So... Today I started testing games out. Lotus-III, Dynablaster, Wolf3D and Doom.
Went back to 1990/93 with this machine. It was really surreal "going back".
The parts list.
- Abit AH4T VLB Motherboard.
- Intel 486dx2-66 Write Back.
- 24mb Ram. (probably downgrade to 8mb)
- S3-805-1mb VLB. (Spea-something??)
- Goldstar VLB Controller. (single channel IDE)
- Creative SB16 CT-2890. (Vibra16 with an OPL chip)
- Floppy.
- 32mb CF Card.
- CD-Rom reader.
I was a tiny bit dissatisfied with the way this machine handles Doom.
When the maps get too big, this machine start "coughing", it is not bad as such.
I just feel that a dx2 should at least be able to run Doom fluidly.
So I might install my Amd 486dx2-80 instead, just to see how it goes.
Or I might just decide that Doom is running ok enough as such and keep the dx2-66.
Edited by Brostenen 2017-03-21 10:45 PM
 (Back-To-DX2.jpg)
Attachments
----------------
Back-To-DX2.jpg (291KB - 672 downloads)
|
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | Doom is quite heavy for what it is really and I think a DX2/66 was their recommended minimum in the revised system requirements. I'm glad to know you have it all together now for the most part and I hope you do find a case for it.
I was hoping to post some pictures of my new K6 working as it got here yesterday, but thus far it is not detecting any hard drive I plug into it. Unfortunately I'm not sure I can justify sending it back because of damage to the corner of the motherboard which happened whilst in my possession. It wasn't my fault, the board was on a wide flat surface, but the floors in this house have so much movement that as I walked away the floor sank enough to tilt the shelving, causing the board, PSU and hard drive to topple onto my chair. I was hoping the chair was soft enough but the padding is worn out and the corner of the board is ruined. It looks like it is just a ground plane and it is miles away from the chipset or IDE, yet I can't help wonder if this is what broke it.
On the other hand, the seller is using IDE2 in the photos with an unusually small drive, which makes me suspect there was always a problem and they didn't disclose it. IDE2 does not work for me though as both channels simply serve to lock the system up. CD-ROM drives appear to work to some extent. I also had to put holes in my heatsink because this board clutters the CPU socket, a problem most of them have and one of the reasons I liked the Chaintech a lot, so there are capacitors touching the heatsink and obstructing it at either end, one of them actually touches the clip peg on the socket itself. I guess this kind of thing is why I haven't seen any other TMC boards from this late on. The silkscreen for the front panel doesn't even seem to be correct as the pins for the power switch are miles away from their indicated position in the silkscreen, to the point that it rivals the MSI boards which had it printed backwards. Otherwise, it might be a great board as even the Chaintech had some really questionable quirks (It was actually prone to lock-ups) but I have no way to make it work. My last PCI IDE card broke down a long time ago and I need the slots anyway.
It seems my old suspicions about buying shiny things are once again correct. Almost any time I buy something that looks new it ends up not working properly whereas when I buy beaten up, dusty looking old junk it generally does what it is supposed to do.
I shall keep tampering with it for now, perhaps I have just neglected something simple. In fact, I might change the RAM because the stuff that came with it is crap and won't even run at the speed it is rated for. Unfortunately I don't have much good RAM left now and I'm not sure my ECC sticks will work in this board. It does take EDO, so I could always try with that for now and change it later if things start working. It has been reluctant to POST so I have to wonder really, I've never seen dodgy RAM cause this problem before but I suppose it isn't impossible.
Also, it is worth noting that to work on the Pentium D any further I have to remove it from its current position. This sucks because it weighs several tens of Kilos - about 55 KG or 120 lbs last time I calculated it. It is also a problem because it means moving my network hardware and so the forum might disappear for a while until I plug it back in. Should only take a few minutes unless the switch batteries are flat though.
Edit: I think it's safe to say the K6 is broken at this point. Even with detection disabled and everything disconnected it still takes well over a minute to pass POST after the memory test has completed. I got it to boot from other devices but it either locks up of halts with random errors. I can't be bothered anymore and think I will put some time into simply down-grading the crappy Celeron my old neighbours gave me. I got the board to detect one hard drive but it just throws the heads around for a while and the drive spins down, said drive works on every other board I have plugged it into. I'm not in any mood to fuck around sending this thing back to New Zealand so I think I'll claim a refund and keep it anyway for the CPU because, given that image,
I think the seller knew there was something wrong with it and did not disclose it, so fuck them, it's their problem, they shouldn't have sold me a motherboard that didn't work eh? Wankers.
ALL TESTED, CLEANED AND WORKS WELL!!! My arse.
Edited by DXZeff 2017-03-22 5:58 AM
|
|
| |
|
TM Crusoe
Posts: 671
| What a bummer, that the SS7 board is not working as it should. Not good at all. :-(
Hope it can somehow be fixed. Could it perhaps be the data cable? |
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | I did consider that and tried several cables because I know I have one that doesn't work properly due to wear. I am sure the busted one is marked though and I'm almost certain I had dismantled it for wires to be used in another project ages ago. Even then, the first cable I used was from the 775 test board which, as we know, was working before.
The only progress I have made since that post is that I can get it to detect a drive on Channel 1, but only as Slave and it locks the system up before POST completes. It also seems I cannot use the board with drives over ~8GB, though as it either crashes or throws up error messages when booting anything, from any drive type including floppy and CD, this discovery is of little help to me. I do find this unusual for a board made in 1999/2000, the Chaintech was running a 40GB drive without having any real problems and that drive in itself is unusual because it is a Deathstar which has passed 15 years of use. |
|
| |
|
IDT WinChip
Posts: 237
Location: USA | Have you checked the dram clock jumper+switch and system divider jumpers to make sure they are right for 100mhz bus speed? Are you using an 80 pin ide cable or standard 40? I only ask because I have run into some older boards and hard drives that just piss on 80 pin ide cables.
Edited by waybacktech 2017-03-22 4:01 PM
|
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | I did try that, in fact I noticed it was set wrong because it was clocking the RAM down to 66MHz despite the modules being rated for 100MHz. The system will not POST if the RAM which came with it is set to run at 100MHz hence my concerns about those modules earlier. I also tried various cables, both ancient 40 channel cables and more recent 80 channel cables, but no such luck. I can also report that network booting causes a lock-up or else causes the boot environment to Kernel Panic with a fatal exception. I have no idea how to interpret the cryptic wall of text that Linux puts on the screen but it is probably safe to guess that it basically means the same as the messages I see if I try to start a Windows environment; That something is broken.
The last test I did was to install a much slower CPU, thinking that the one the board came with might be faulty and having no other board in which to test it. By now the system is painfully slow, running a Pentium and some EDO RAM at the lowest speeds the board has to offer, but nothing improves. Essentially this board is behaving exactly like my Jetway 531CF did right before it died.
As I said, I'm past the point of caring. I am sure I can cripple the Celeron enough to make it fill in this gap I have between P1 and K7 type machines. Otherwise, in the mean time I think my time would be better spent going back to playing around with the STPC Atlas as I still haven't imaged that DoC. I think it is one of those things whereby it will either not work, or else if I walk away from it for a few days and come back with a fresh outlook I may have more luck, at the moment I'm just tired of it and find it hard to thing straight because of how much it irritates me.
Edited by DXZeff 2017-03-22 11:21 PM
|
|
| |
|
IDT WinChip
Posts: 237
Location: USA | Boy, I don't know. The behavior you describe is what seems like either bad ram or something seen when overclocking to the breaking point.. Kind of makes me wonder if the L2 cache on the board is any good. Making the assumption that the board was functional before it was shipped, which there is no proof one way or the other, but maybe the spill onto the floor cracked some solder joints. Maybe you can try a bios update, maybe there is something with the bios, i mean, this is a long shot but someone could have modded the bios and fiddled with hidden memory timings... or it's a shit revision and has issues with certain memory types. cpu voltage issue maybe? Maybe power supply? Maybe shoot the cpu socket, ram and slots with some contact cleaner. That's about all I can think of other than the board is just on it's death bed for what ever reason.
edit: My FIC 503+ has jumpers for ram voltage, 3.3 or 5v.. I don't know if your TMC has that or not but that's another idea...
Edited by waybacktech 2017-03-23 12:09 AM
|
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | It is probably worth noting that so far as I remember, the board was not showing any signs of life prior to the drop and wasn't even bringing up POST, though at that point I just assumed something was loose or dirty. It would appear all the timings are exposed in the BIOS but I'll keep this in mind for when I come back to it in a few days, the only problem is that I don't know where my PLCC adapter has gone for my programmer and there is no way on earth I'm attempting to program that chip in the board. I do not have another board using PLCC EEPROMs that are socketed, or at least not ones where it is easily accessible, so hot flashing isn't an option either. I will also try turning caches off (though that isn't going to be much help if that does turn out to be the problem) otherwise, I've exhausted most possible causes by now. |
|
| |
|
IDT WinChip
Posts: 237
Location: USA | Well that's interesting... you don't suppose something or several small somethings conductive might have gotten lodged under the chipset, cpu socket or somewhere else when the guy "cleaned" the board? Maybe you should give it another deep cleaning. |
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | Could have done I suppose as there's no way to know how they cleaned it or what they used to do it. I guess it isn't impossible that too much of certain detergents might be slightly conductive if they left a residue behind. To be honest that is one thing which is definitely accurate about the listing, everything does look exceptionally clean aside from the dusty heatsink.
Meanwhile I have figured out my route of attack for getting that DoC drive image. |
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | Much as I am sure I should just give up, I have made progress. Firstly, I found my PLCC adapter and discovered the BIOS chip was faulty;
So I stole one from a dead Soltek board. Then I learnt that using AGP video cards causes a crash when changing video modes, which is probably what some of the lock-ups were, attempts to display LOGO.SYS or change the resolution of the console when starting an operating system or its installer.
The patched BIOS negates any need to disable the cache as it does not detect any being present on the motherboard, though it is at least stable enough to run SpeedSys and grants me a score of 11.47... Oh, how wonderful.
I did some more research and it turns out that the first version of the board (Mine is V2.00) only supported 66MHz CPUs officially and maxed out at 75MHz, which may explain why I can't get the RAM up to speed and the rest of the system has problems running at the higher clock speeds in general. It also seems TMC did extensive patching to the BIOS in a very short space of time after the board came out and numerous communities also made their own patches.
Scratch that, the cache comes back if I clock down to a 66MHz FSB. I will experiment with latency and the clock adjustment options in the BIOS to see if it comes back. Still, I guess you were right about that being a problem in a way and updating the BIOS certainly fixed the drive detection - though according to what is out there, it should have detected drives up to 32GB just fine whereas mine was topping out at 8GB... Albeit with that faulty BIOS chip which was likely responsible and would definitely explain the occasional no boot (as in, not even POST) that was happening.
I can't exactly say I'm happy, I mean some K6-2 this is when it has to use EDO RAM, can't use AGP video cards and doesn't have working caches. But it sort-of works I guess, which is more than I could say before. I cannot help but feel no matter what I do, that the board will never be a true 100MHz board and will always have problems running at this speed. It is kind of novel for this I suppose, I haven't seen another MVP3 board that wasn't rated for this speed and I doubt there can be that many of them out there.
One thing I must test at some point, purely out of curiosity, is if it has the same aversion to K5 processors as the Chaintech did. For some reason the Chaintech had compatibility issues with them. I wouldn't want to waste a Super 7 board on a K5, but it was something I discovered (and one of the problems I faced) long ago when I did the Socket 7 Showdown video and I have to wonder if this board is the same.
Edit: Cache working, AGP working, but now consistently reports "Memory Test Fail" when starting and in SpeedSys. Definitely leaning towards faulty chipset at this point.
Edited by DXZeff 2017-03-23 11:50 AM
|
|
| |
|
IDT WinChip
Posts: 237
Location: USA | Bad eeprom... didn't see that one!
Does this board have a jumper or switch for sram intel/cyrix burst option? |
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | I don't think I've ever actually ran into a board with a bad EEPROM before, so I definitely didn't see that one coming as they don't usually break for no reason, perhaps it had been tampered with.
No such jumper, but it does have a reset jumper. I had to go out to the shop and noticing that the latency settings weren't sticking I acted on a total whim, took the battery out and also left the CMOS reset jumper in the reset position until I came back...
For whatever reason this appears to have fixed most of the remaining problems aside from one lock-up that happened. Honestly, I'm baffled, perhaps the CMOS RAM is not working as it should? It seems to be mostly behaving now anyway, it got through Quake's timedemo (which it wouldn't even start previously) so it might not be a total lost cause, it is even running with the RAM at 100MHz now. I have to sleep very soon so I can't really test it much further, but once I wake up I suppose I will try to install Windows and see if it can hold itself together.
Seriously though, completely baffled, never seen that happen before. It still has a ways to go before I trust it at all though. I mean, I reset it after updating the BIOS already.
Oh, also, the RAM and CPU for my Pentium D got here, so I shall be cleaning the water block and installing those later. Here's hoping the motherboard in there has enough life left in it to finish Mirror's Edge, but if not, the RAM will just be used elsewhere as I have other DDR2 boards that could use the upgrade and I can play the final level with the Xeon. |
|
| |
|
IDT WinChip
Posts: 237
Location: USA | I've read from time to time that CMOS data stored can become corrupted even though the settings in the bios options seem to be holding on save, and a full clear/reset will clear the corruption. Personally I've never experienced this phenomenon. Closest, and perhaps is the same thing, I don't really know, is those dual bios gigabyte motherboards sometimes will suddenly pop up a message about cmos data being corrupt and it begins restoring from backup bios, which I have had happen on either just completely when it feels like it, or after an overclock.
Glad you figured out a combination of things to get this board working again. At least it won't be a waste of money, I was a bit afraid you might have gotten schlonged without lube on that one there.
I'm not sure I recall you uploading a video on your Pentium D. |
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | I remember hearing all sorts of horror stories about those Dual BIOS boards when they started showing up. Personally, I always kind of liked ECS's solution - Top Hat Flash - though only because it wasn't attached to the board and therefore caused no problems. Otherwise it was much less elegant, less user friendly and I never had to use it.
Indeed, here's hoping it stays working. I have tested the included RAM modules elsewhere, as I was using some of my own on loan from another machine anyway, it turns out they don't work properly so I'll have to keep mine there for now. I have no immediate need for the system they belong to so I can grab some more later, may even have some spare modules at the bottom of a box I have forgotten about.
No video for the Pentium D, the plan was to do it after Mirror's Edge and possibly after the Athlon. It will still get one even if it doesn't work as I have ample video of what it did to still make something passable anyway, especially given it was behind almost everything on both channels until December last year and didn't officially retire from that role until Duke Nukem II was over. Counting everything, including the work it did off of YouTube, the machine has clocked up well over 100,000 operating hours, rendered thousands of videos and completed unfathomable numbers of smaller jobs as it approaches twelve years of near constant running. I have photos somewhere of when it was new, it looks pretty sorry by comparison at this stage and the years have not been kind to it. |
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | Another report: The K6 is still behaving and the last of the problems do indeed seem to have disappeared with the SRAM reset. Still no clue, but whatever, it works so I won't question it too much and instead just hope it stays working.
Pentium D is not playing ball. I am currently researching a few things though as I may have one last (relatively dirty) trick to get it working again... Not sure how long it will last if this works, but it may hold on long enough to complete the game. I may lose the RAID array in the process depending on how badly this affects the board, but if it does I won't lose anything important. Nothing but a few thumbnails and a small cache of music I have backed up elsewhere were really on there outside of what was already moved to the Xeon. I have the board out, I have the iron ready and I will report back later if I get it moving. My test board might have to die in the process, but screw the test board, it was crap anyway.
Don't know if I'm alone in this way of operating; I reach a point whereby something is so broken with no prospect of working, that I don't mind taking drastic and potentially destructive measures in an attempt to make it run again. My reasoning being that as it doesn't work anyway, what is there to lose?
Anyhow, fingers crossed.
Oh, back on the K6, thinking of moving to a Rage 128 now because the image quality on the TNT is pretty bad. Most of the ones I used have been so I guess that was just a thing with them. I seem to remember the Rage being quicker in most of the applications I want to run on it anyway, so the hit in 3D isn't really a problem and I don't suppose anything will stop me changing it at a later time if it becomes one.
Edit: Update...
-Maniacal laughing as I rub my hands together diabolically-
Not completely out of the woods yet though as it is giving me a black screen, but the beep codes sound correct and I get drive activity, implying the system does boot.
Further edit:
I'm going to stop for today and see about maybe getting the K6 into a case.
Edited by DXZeff 2017-03-24 7:38 AM
|
|
| |
|
IDT WinChip
Posts: 237
Location: USA | Rage 128 is a very good chip. If you have or can get an ultra version for cheap, highly recommend it for this k6.
I have the same philosophy when it comes to hardware. If it's already broke, no harm in breaking it more to fix it!
|
|
| |
|
TM Crusoe
Posts: 671
| That is a lot of stuff, you have done with that motherboard. Good work. :-) Rage-128? Hmmm... Yeah. I had one at some point, though I sold it in a complete SS7 machine (P5A 1.03, K6-2-300, 128mb and some 40g HDD). Personally I like TNT1/TNT2, G400 and Voodoo3 better in a K6/K6-II/K6-III system. Though a K6 might actually be underpowered for those GPU's. For a pure K6, I really do not know what is the absolute best to go with. V1+G200 or V1+S3Trio64?
Anyway...
I recieved them antistatic bags that I ordered some time ago. And I have torn down two of my build's. I actually never used them for anything, and I was able to rediscover some cards I had forgotten that I had. I found two AWE64-Value's, and one Matrox G400. So now I have four G400's in my box of cards.
Yeah.... Need to store them ATX cases, so they do not fill up the space in the room.
In the future, I will to rebuild the K6-II-500 (P5A rev 1.04) and the P-II-350 (P2B-S rev 1.02). That however is not before I have bought shelves for the room.
Edited by Brostenen 2017-03-24 10:28 PM
|
|
| |
|
IDT WinChip
Posts: 237
Location: USA | Well... I've been spending money on ebay again. I have to stop looking up stuff, however, since this is a rare bird to find... and the price was a very reasonable $21 for the mb...well... I really couldn't say no. Of course that meant finding some processors, so I settled on a pair of 2MB 700Mhz with heatsink at $15 for the pair, and found a parhelia PCI-X card for $35, which these are normally anywhere from $68 to well over $400 (lol ya) pair of VR's for $17 also. We shall see. I was looking for the 900Mhz Xeons but those were few and WELL over my comfort zone. Not confident this board will run 133bus chips, and as those are only 256K cache, which doesn't seem like any better than a normal P3 to be honest, well this is the combo I have settled for. I am sure there is a resale value to this if I decide it's really not worth it in terms of speed over a normal P3.
Extra thoughts:
As for the P2B-DS build, so far the radeon 7500 is turning out to be the card to beat on this system. The Matrox G450 I tested scored nearly 1/3 of the 7500 in 3Dmark2000, and locked up after that run, so I don't think the G400, even overclocked to Max levels will perform close to the 7500. 9200 scored lower also, but was nearly identical in UTbench. I'm having issues with the version of quake 3 I downloaded, since my CD seems to be unreadable now though I think I still have it installed on my Ppro with the bigfoot drive, have to get it off that drive and then benchmark should work. Geforce2GTS also scored a little lower than the 7500, so I might try my GF3 Prophet III card as it seems as typing this, the system just locked up during 3Dmark2001 run LOL. This agp bus thing might be a real problem. 7500 seemed to tollerate it better than anything else so far.
Interesting message just popped up on my screen from adblock - "This site has been known to show targeted messages to adblock plus users. Do you want Adblock plus to hide targeted messages?" LOL cool. First time. Apparently you've gotten this site on the map paul. Good job! 
Edited by waybacktech 2017-03-26 12:11 AM
 (slot2_intelc440gx.jpg)
 (xeonP3.jpg)
 (MGIParhelia.jpg)
Attachments
----------------
slot2_intelc440gx.jpg (583KB - 673 downloads)
xeonP3.jpg (304KB - 646 downloads)
MGIParhelia.jpg (207KB - 684 downloads)
|
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | @Brostenen; Rage 128 should be good enough, I have found a 32MB 128GL (with TV Out) at the bottom of a box, but don't know if it works yet. I still have my Xpert 2000 if it doesn't which would work fine as a hold-over until I could get something better. My problem with TNT cards is that so many of them have dirty signals, probably because people bought the cheapest crap they could find where the card maker cut back on the components. I had a G400 in a low-end P3 some years ago and it was OK, but I always saw it as more of a 2D multi monitor card than anything else, very good multi monitor support. V3 I think is more of a card for Coppermine type machines in that gap between the K6 and Athlon and the V1 is too slow to outrun Software rendering in my K5, so it wouldn't be a good choice here, V2 might be good if I was doing that. Several of these cards will likely end up in here at some stage however, due to a project which is running extremely late by now.
Overall the Rage always worked for me the best because it is still a fair performer when software rendering is required, even in DOS, the Windows games/applications I want to run are primarily 2D or use DirectDraw, and the all of the 3D games I intend to run are Direct3D. Anyway, I like having MPEG2 decoding to hand, it comes in useful.
Haha, I hope your K6 rebuild goes smoother than mine when you get around to it.
@WaybackTech; Hell yeah, Slot 2. That is almost certainly an Intel motherboard so it should at least work, hopefully, and good job on actually finding the VRMs for it, that's one of the things which puts me off every time I consider putting one together.
I think I said that I still have fond memories of my 7500. I moved to an MX420 once it stopped working and it was never the same, even the 9200 didn't "wow" me in the same way. They were always good cards and the only thing I ever got stable on the K7S5A I was using at the time, a board with a horrific AGP implementation.
This site doesn't display any ads and it certainly shouldn't react to AdBlock. All I can guess is somebody dislikes me enough to file some bogus reports somewhere, or else my old joke script has caused an automated response by AdBlock's lists - the main page used to have a modified "FuckAdBlock" type script which placed a message on the page. Said message congratulated the user for running AdBlock and told them that the page will never show advertisements anyway. The only thing that could be considered an advertisement would be the forum notice at the bottom of the page and whilst I probably could get away with removing it, I think it would be disrespectful to the authors and it is not intrusive anyway.
Edit: Oh, I forgot. My Audio Module is here, but there's a small customs bill for it. The Royal Mail site is acting up so I'll probably just walk to the office and get it myself on Tuesday.
Edited by DXZeff 2017-03-26 6:03 AM
|
|
| |
|
TM Crusoe
Posts: 671
| Well.... The K6-II-500 system was build, using an Asus P5A Revision 1.04. It had an Asus V3800 (TNT2-Pro 32mb), Creative AWE64-Value and 128mb Kingston running at 100. Though the module is a 133 edition, I used it downclocked for super stability. I never had any issues with this specific setup, other than it might not be as fast as a K6-II-500 system can be. It was just a super stable machine.
If you want super VGA signal quality on a TNT2, then eighter Compaq, Asus or Creative are the way to go. Keep to Pro and Ultra models.
Edit:
I can't remember the exact 3D-mark99 score. Though I think it gave me some 3200+ or 3300+ results. I did the testing some 10 to 12 months ago.
Edited by Brostenen 2017-03-26 9:02 AM
|
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | I found that I have a Rage 128 Pro from a gutted out garbage find machine. I may end up using this instead.
My old K6 was ATI, so I'll probably stick to it in the long term anyway aside from that experiment I want to run some day with different cards.
Edit: Forget it. Mounted the board in a case, now it is back to freezing at drive detection or failing to detect them, no boot with 100MHz modules and basically the rubbish it was puking at me with before.
Edited by DXZeff 2017-03-26 10:36 AM
|
|
| |
|
TM Crusoe
Posts: 618
Location: Hull, UK | Mostly working, but with reduced speed and overall, not an adequate replacement for my 5AGM2. But I guess it will have to do.
Edit: If you can't open TIF files, don't worry because it's not spectaculr. I merely viewed the forum from the K6 (yeah, it works in 98SE out of the box) and snapped a picture of it. I had no tools to compress it to formats like JPEG to hand, so I went with what I had.
I feel this is going to be a very temperamental board overall and I'm really not confident it running it at these speeds. Much like the early Socket 7 boards that topped out at ~120MHz before becoming unstable, in part because they were basically modified Socket 5 boards, I can't help feel that TMC's updates from the 66MHz version were a little precarious overall. Still, I don't have to rely on this thing, so it should be fun figuring out what I have to do to keep it working I guess. Surprisingly, it seems very stable once Windows is running, that I did not expect given its behaviour everywhere else.
In case you wondered, the Socket 7 board I reference was also a TMC and it was actually a very good board when running a slower CPU.
Edited by DXZeff 2017-03-26 5:31 PM
Attachments
----------------
k6on.tif (901KB - 690 downloads)
|
|
| |
Recently Played/Bought/Messed with
