Errius wrote:

http://aopen.jp/products/housing/h300.html http://aopen.jp/products/housing/h340.html […]
Show full quote

http://aopen.jp/products/housing/h300.html
http://aopen.jp/products/housing/h340.html

The H340A is externally similar to the H300A but has more ventilation holes. It also comes with a beefier (200W) P4-capable PSU.

The FSP145-60SA PSU was produced in two variants. The version in the H300A is microATX and has a projection next to the cable exit. The other version is FlexATX and lacks this projection.

http://aopen.jp/products/power/fsp145-60sa.html

I have zero problems with this H300A case. My system does not overheat and the psu fan is quiet. I have used the system for hours straight already. I think cable management en keeping fans clean (mine are new) helps keeping temperatures low.

Another one is ready 😀
Will post the internals tonight...

Case: AT case with LED display
Motherboard: FreeTech P5F76
CPU: Pentium 133
RAM: 128MB
GPU:S3 Trio64/V2DX + Voodoo1 ATC-2465
Soundcard: SB16 CT2290
Harddrive: 8GB (I think Quantum)
Drives: floppy disk and 48x speed CD-ROM
OS: Windows 95B

As some of you know I finally found my most wanted piece of vintage hardware which was the Cardinal SNAPplus video card. Unfortunately the SNAPplus I found did not come with a piece of software that my SNAPplus I had back in the early 90s came with which was Mathematica's Tempra Pro. Tempra Pro is a hard to find MS-DOS based video graphic overlay suite for use with video capture cards, but I managed to find one sealed copy on Ebay. It finally arrived today and I took a photo of the box before I removed the shrinkwrap and I am glad I did. The shrinkwrap had sat tight on the box for 26 years and when I went to remove it the ink on the box also went with some of it. Now the box has ink missing and some tears. This is why museums remove shrinkwrap off sealed items because it eventually degrades the material underneath. So I opened the box and everything looked mint inside of course except the floppy disks. Another tragedy from age started to unfold. It came with both 5¼" and 3½" disks, but they were packed so tight together in a paper envelope from the factory that the 3½" disks had warped the 5¼" disks beyond repair. I then went to image the 3½" disks with WinImage and three of the four disks imaged without issue, but the fourth disk was having major read errors throughout. I was desperate so I did the old shake and blow method on the disk each time it got a read error and that seemed to limp it along enough to get all the data off the disk. Whew!!! So now I have a good image of the software thank goodness. Shows that just because it's new and shiny doesn't guarantee it will work. I suspect it was stored in a moist basement or humid attic and if it had not been sealed all the disks would have probably been toast. Now I can go ahead and do a full Youtube demo of the SNAPplus and the software it works best with. Should be fun.

Here is the box before I unwrapped it and it no longer looks this good unfortunately.

The warped 5¼" disks. Completely ruined unfortunately.

Lately I've been playing around with some spare CPUs. Since some of you have claimed that a Pentium MMX is overkill for a Win3.x machine, I decided to play with a 200 MHz Pentium 1 CPU (without MMX) to see if it would be suitable for the most intensive programs I run on my multi-purpose DOS/WfW3.11/NT3.51 rig. Unfortunately, the performance decrease is worse than I expected, Firefox 2.0.0.22pre (on Windows NT 3.51) suffers from slow loading times on the P200 yet runs just fine on the P233MMX (I guess that explains why the official system requirements state a "Pentium 233 MHz or higher" as the minimum recommended CPU), Speed Haste (in SVGA mode with 30% background depth) ran like a slideshow on the P200 but ran at a playable rate on the P233MMX. I ended up sticking with the 233 MHz Pentium MMX CPU despite the overkill factor, I'm sorry guys, but I'd rather have an overkill Win3.x machine that can properly run CPU intensive software/games on the Windows NT and DOS sides than have an "era-correct" Win3.x machine that takes the fun out of playing with CPU intensive software/games.

Socket 7 aside, I've successfully tested two out of three of my spare Socket 370 CPUs in an HP e-Vectra, the two spares that were tested include a 700 MHz Coppermine Celeron and a 1 GHz (Coppermine) Pentium III, the third one that couldn't be tested for a reason you most likely know is a 1 GHz Tualeron (that was used in my GX150 not too long ago for the purpose of Tualatin compatibility testing).

Also playing around with a newly-acquired MSI nVidia Geforce4Ti 4200 (64MB AGP) video card to see if it's a better alternative to the Radeon 7200/DDR that was previously in my Dell Optiplex GX150.

Cyrix200+ wrote:

Selling/giving away some stuff I don't want or need. Concentrating on large items now, moving to smaller parts next week. I need to get some space back to work in! 😀

Sold two Pentium 3 systems and a CRT, so I gained some space. Then my brother came and gave me a very very dirty Amiga 500 with a whole bunch of accessories. 🤣 Girlfriend not happy. I should just sell it as is, but I have never had one so I do feel tempted to clean/restore it and play with it a bit.

liqmat wrote:

As some of you know I finally found my most wanted piece of vintage hardware which was the Cardinal SNAPplus video card. Unfortunately the SNAPplus I found did not come with a piece of software that my SNAPplus I had back in the early 90s came with which was Mathematica's Tempra Pro. Tempra Pro is a hard to find MS-DOS based video graphic overlay suite for use with video capture cards, but I managed to find one sealed copy on Ebay. It finally arrived today and I took a photo of the box before I removed the shrinkwrap and I am glad I did. The shrinkwrap had sat tight on the box for 26 years and when I went to remove it the ink on the box also went with some of it. Now the box has ink missing and some tears. This is why museums remove shrinkwrap off sealed items because it eventually degrades the material underneath. So I opened the box and everything looked mint inside of course except the floppy disks. Another tragedy from age started to unfold. It came with both 5¼" and 3½" disks, but they were packed so tight together in a paper envelope from the factory that the 3½" disks had warped the 5¼" disks beyond repair. I then went to image the 3½" disks with WinImage and three of the four disks imaged without issue, but the fourth disk was having major read errors throughout. I was desperate so I did the old shake and blow method on the disk each time it got a read error and that seemed to limp it along enough to get all the data off the disk. Whew!!! So now I have a good image of the software thank goodness. Shows that just because it's new and shiny doesn't guarantee it will work. I suspect it was stored in a moist basement or humid attic and if it had not been sealed all the disks would have probably been toast. Now I can go ahead and do a full Youtube demo of the SNAPplus and the software it works best with. Should be fun.

Awesome, glad you managed to get a working copy off the disks. I'm really curious about this kind of old AV equipment & software, I was always under the impression that the Video Toaster was the only thing of its kind but I guess there were bound to be some alternatives around. Please do link the video when you post it up. Funny to think Mathematica wrote software for video production, but it does make some sense.

Also that box art is amazing, shame yours got damaged. Is there a high quality scan up somewhere?

There's a boxed copy of Mathematica for 68k Macs sitting on a dusty shelf in my office, maybe I should go "rescue" it.

Murugan wrote:

Case: AT case with LED display Motherboard: FreeTech P5F76 CPU: Pentium 133 RAM: 128MB GPU:S3 Trio64/V2DX + Voodoo1 ATC-2465 Sou […]
Show full quote

Case: AT case with LED display
Motherboard: FreeTech P5F76
CPU: Pentium 133
RAM: 128MB
GPU:S3 Trio64/V2DX + Voodoo1 ATC-2465
Soundcard: SB16 CT2290
Harddrive: 8GB (I think Quantum)
Drives: floppy disk and 48x speed CD-ROM
OS: Windows 95B

Great P133 build, pretty close to mine though mine is DOS only.. My one question is, why stuff 128MB onto a board that can only cache 64MB? That is rather counter-productive, you are better off using Win95 disk caching than over 64MB system memory..

TBH I forgot about that even though I read about it a few weeks ago. I'm still pretty new in building retro rigs.
Guess I had the memory and I wanted it to the max 😀
Since my goal for now is to get them all to a working state, I'll leave it for now but when all my builds are ready, I'll evaluate them again to see what I can do better.

Thanks for the reminder though since I still have 2 W95 build to finish. Or maybe I'll make it DOS only like you.

xjas wrote:

Awesome, glad you managed to get a working copy off the disks. I'm really curious about this kind of old AV equipment & software, I was always under the impression that the Video Toaster was the only thing of its kind but I guess there were bound to be some alternatives around. Please do link the video when you post it up. Funny to think Mathematica wrote software for video production, but it does make some sense.

Also that box art is amazing, shame yours got damaged. Is there a high quality scan up somewhere?

There's a boxed copy of Mathematica for 68k Macs sitting on a dusty shelf in my office, maybe I should go "rescue" it.

Newtek's Video Toaster was by far the supreme ruler of the era in video editing and overlay. Hands down, no question. I was a big Amiga nut from 1985 until 1992. Of course right when I needed a video overlay card I had already sold all my Amiga gear and moved back to the PC in 1993 so the SNAPplus was a waaaay cheaper solution for me and did the job. Newtek's products, though, were far superior, but more complex and definitely more expensive.

The Cardinal SNAPplus was one of many PC video editing/overlay cards, but I liked Cardinal's software package and it was very competitively priced even at $1200. Remember this card was basically a computer on a card and is why you could do this kind of work on the minimal system requirement which states is a 286 in the manual. I ran mine on a 386 SX/25 with 16MB. One card that comes to mind that was in the same competitive space at the time was the TrueVision VideoVGA. This article shows most of the contenders on the market back then starting on page 250 and the Cardinal SNAPplus isn't even the best performer, but IIRC it had the best balance of price and performance that I needed.

https://books.google.com/books?id=qwZH3rQBuOk … napplus&f=false

By the way, another sealed copy of Tempra Pro popped up on Ebay from a different seller right after I grabbed mine, but at almost double the price. At that price, nope. I also did a test run in DOSbox of Tempra Pro and Tempra Show. I managed to get them to load, but if the CPU cycles are too high the software crashes with a divide by zero error. So I probably can't demo this on my Tualatin 1.4GHz system running DOS, but I am going to test it today. If I have to buy a 386 or 486 to demo this and at the prices those types of systems go for I will probably not be posting a video on this. I do have a few mint AOpen Socket 370 AX34 Pro II boards with 1GHz CPUs to trade or sell, but I still don't think that will get me to a 486 box. Maybe I'll get lucky locally and find one cheap. We'll see.

https://www.ebay.com/itm/Mathematica-Tempra-P … hwAAOSw2Yta3gIq

Well, good news is Tempra Pro and Show work fine on the Tualatin 1.4. I read an article about Divide by Zero errors on older programs in DOS and one of the solutions is to disable the L2 cache. Once I did that in the BIOS Tempra worked perfect. So the next step is... will that 16-bit ISA slot work with an old SVGA card on the new Chinese Socket 370 board I bought? If it does I should be good to go on the video and not have to invest in the absurdly expensive 386/486 systems out there.

Update: So remember that factory new Chinese Socket 370 Tualatin board with one ISA slot I bought recently? Yeah, ISA video cards work perfectly on that thing . It also turns out I have to disable the L1 cache and not the L2 cache to get around those Divide by Zero errors in the Tempra software. So it looks like I am good to go on doing that SNAPplus demo on the Tualatin machine. Saves me a lot of time and money. I will definitely be getting another one of those boards from Alibaba. It's my Swiss Army knife motherboard at this point. I also do all my CP/M 360K disk writes on that thing as well.

Testing... testing... testing...

Have made good progress today - my two P3-1000EB CPUs work, but not on my Tyan Tsunami for some reason (it's perfectly happy with a P3-600E, so it's not CuMine incompatibility, and the P3-500 Katmai is also likes draws more current, so that's not it). The work perfectly OK on my MS-6168 though.

A while back I picked up a 44p IDE DOM-like flash module with 1GB, which is a big upgrade from the 32MB my HP Transmeta Crusoe-based thin client was shipped with. Not so 'thin' anymore. Next question is which OS to run on it. BeOS would have been perfect, but no support for the sound (Via AC'97 codec), so contemplating OS/2 Warp 4. Just need to figure out how to get a CD or USB stick with OS/2 bootable media, as my install CD isn't bootable.

Also gone through a big pile of sound cards. The ISA cards all worked (nothing special: OPTi, AD, Crystal, Yamaha and ESS SBPro2 compatibles, none with wavetable or any such thing), but the PCI cards were a mess. Two SB Vibra 128, a Yamaha DS-XG and an Aztech AZF3328 did work, but an Aureal Vortex 8820 and Terratec Aureon Space (Via Envy) didn't even get detected and a Trident 4DWave-NX card caused BSODs at boot in two separate builds with both the drivers I could find for it. The only one with visible damage was the Auureal - one pin on the chip seemed bent but turned out to be completely broken. Oh well, can't win'em all, goes into the 'try soldering this stuff' box.

And in other news, my son (5) managed to run his first game today (Worms, the 1995 original), but immediately complained about lack of sound. So we picked a sound card and he installed it - a nice little ESS688 card (I did the jumpers and BIOS resource settings though). He likes Worms a lot more now 😉

Been hunting down my original 1990 installation of TEMPRA in my offline backups. This is a PNG conversion of the TARGA TGA file as it appears on the TEMPRA box art.

Tested a bunch of IO cards and a 286 motherboard. I picked up SimIsle and a few other titles (Encarta 1996, an IOMEGA driver disk, and two serial mice) at a local-ish computer store. Had I known about that place a few years ago, I'd have been in AT case heaven.

I'm getting frustrated trying to get the onboard SATA channels of an AMD880 motherboard to be recognized during Windows 7 installation. I tried and failed to resolve this with some files on a USB drive. I then check Asus' support page for the motherboard. They helpfully list the drivers, but neglect to offer a download link for them.
I decide I'll just find the original CD that came with the motherboard.

So I go to grab my Ancient Box of Disks, conveniently located on an upper shelf of the Big Messy Closet.

Disaster. The entire box tumbles over and every disk in the box is scattered through that corner of the Messy Closet.
Spend the next hour rummaging around, trying to gather all the floppies and CDs in the vain hope that I won't lose any.

In the bottom corner of the closet, I find a box. It's in my way. The box is innocently labeled as containing some "old drives". This is from the quaint old days when all my "old drives" could fit in one little box.

Inside... huh? I lost this stuff ages ago. My 3 oldest hard drives... a 40MB Type 17 Conner (OEM from our 386SX), a 240MB Conner, and a 1.6GB WD (which had bad sectors in it's first year - piece of crap). And some other junk I've already forgotten about. It has been so many years since I misplaced those hard drives that I wondered if I had thrown them out. Nope.. they've just been hiding.
The Conners were both perfectly good last I used them. They were in tattered old ESD bags. I carefully moved them to new 3M resealable ESD shielding bags and stored them with the rest of my hard drives - finally where they belong. And back to cleaning up all the disks.

By the way, the driver CD I was looking for was not in the box I had reached for. An hour after cleaning all this mess, I found it somewhere else. And it doesn't work. Windows 7 still refuses to see the SSD plugged into the SATA ports.

Not sure whether to be happy or pissed. I can't get Windows 7 to install, but hey, I found some ancient hard drives in my closet.

====
Edit: Problem solved. Windows 7, just like older versions, is confused when it doesn't understand what's already on the target drive, and believes the best way to respond is to act as if the drive doesn't exist. After "cleaning" the drive, now it's recognized.
This is rather maddening behavior when I have never suggested that it needed to concern itself with anything that was stored there. Just do a clean install please, Windows, and don't act stupid.
====
Edit: Another problem solved. Turns out Modern-Windows will also refuse to install if the target drive isn't configured as the first drive in the boot order in your BIOS. https://tomrichards.net/2014/01/how-not-to-in … erating-system/
Windows is doing overtime finding cryptic excuses not to install itself.

I saw an ISA riser that looked like it would fit in a Dell 386SX board I have which I didn't have the riser for, so I bought it. I put it in there with an old SMC network card that only has BNC and DB-15 ports on it (and jumpers to enable the boot ROM!) with a freshly-programmed XT-IDE EPROM to try out. It wouldn't detect a WD31600 for some reason but it worked with Quantum CR6 and a Samsung 40GB. The board boots at only 8MHz and I couldn't figure out how to switch it to 33MHz (I thought I did one time in the past, but can't remember how)

One night I found myself wondering whether the 286 AMI BIOS uses INSW/OUTSW instructions (which are faster but absent on 8086) for disk I/O. So I ran it through NDISASM and indeed there is a routine like this:

10000A49D  B90001            mov cx,0x100
20000A4A0  BAF001            mov dx,0x1f0
30000A4A3  FA                cli
40000A4A4  FC                cld
50000A4A5  F36D              rep insw
60000A4A7  FB                sti

Another thing I was wondering about was timing for 8-bit ISA cards in AT-class systems. I had observed via benchmarks that 8-bit cards have low throughput comparable to what would be expected with the 4.77MHz 8088 bus. I found this informative page about the ISA bus: http://www.hardwarebook.info/ISA

The AT approximates the longer access time of the XT by using additional wait states. So a memory access on an 8-bit card plugged into the 6MHz AT will take 6 cycles (1uS). But it is based on the CPU clock, so when I overclock from 6 to 8MHz even 8-bit cards get a little bit faster. Cards can use more waitstates with the IOCHRDY signal, or fewer waitstates with the NOWS signal (although this signal wasn't specified for the original PC). So the 6-cycle timing for 8-bit cards is a "default," whereas for 16-bit cards it is 3 cycles. The webpage implies that 16-bit cards can also use NOWS to get a shorter access time. I wonder if any do? (8MB/s over ISA?)

Did a little bit of time-shifting for my systems; my Pentium system now covers 1993-1995 DOS (previously 1992-1994 DOS) and my K6-2 system now covers 1996-1998 Win9x (previously 1995-1997 Win9x). This follows upgrading my P3 system to a GF2GTS and shifting it to 1999-2000 (previously 1998-1999). I am now MUCH more comfortable with the hardware and their respective period correctness. This requires a lot of game uninstalls and reinstalls of course, so that is exactly what I have been up to. I've started removing 1992 games from and installing 1995 games to the DOS PC today. In the next few weeks I will do the same for 1995/1998 games for the K6-2 system.

Powered on a IBM 5154 monitor I got for free. It took about 5 seconds before the whole thing blew up. Loads of smoke! Ran out with it since I feared it was on fire. Seems one of the capacitors blew in the power supply which is a common problem with these. Gave me a good scare!

Attempted to play the Jimmy Neutron game again last night with the GeForce2 MX400, and still nothing. I believe it's because there's no shadowing whatsoever (especially the BIOS doesn't have the video BIOS shadowing option since it's a PCChips M871G v1.5 and that must've gotten revoked back when Pentium 4 systems were manufactured).

I have a board from a parted HP Pavilion 7955, but that gives me resource conflicts and my GPUs never worked with that board, except for the TNT2 it came with. Going to attempt to install my GeForce 6200 AGP card and see what happens.

I finally bought a Promise Fasttrak TX2300 controller for my Tualatinator. Hopefully the card will be easy to install and will detect the two 300GB WD Velociraptor hard drives, as I had no luck at getting them working on a Silicon Image SIL3112 card before it croaked for good (it detected the HDDs but refused to write to them and caused BSODs and slowdowns) . Strangely, the Silicon Image SIL0680 ATA133 card successfully detected the hard drive hooked to a SATA to IDE adapter and gave quite a boost to the system.

I'm also sniping an auction for a Kingston 512MB PC133 SDRAM stick that will end tomorrow morning. Will this stick be finally error free, and above all, will it be able to run at CAS2 Turbo speed?

That is the question...

xjas wrote:

Is there a high quality scan up somewhere?

Here. I scanned in the glossy cover of the Tempra Show 2.0 manual. Same exact graphic as the box. 600dpi in 24bit color. It's 30MB so Google drive it is.

https://drive.google.com/open?id=1B4qwHE9xB42 … _mHRJ-Gv35vGWhP