brassicGamer wrote:

Wow, so it's not just a fantasy lineup - you have these machines for real? Would love to see a pic of the actual articles too. The attention to detail is incredible, like the line of light leaking in through the case of the far-right IBM that's visible through the grille.

Yeah these are my machines 😊 This doesn't even scratch the surface, although most of the other machines are 'just' generic PCs... amongst a few iconic machines too. Which I'll model too...

I'll take some photos of my little work area soonish, ended up cleaning the house, as one does at 11pm on a Friday haha

The machines in that render still need a lot of extra detail added, and the light you can see through that PS/2 probably isn't normally visible because of the hardware inside. I might model parts of the internals, or I might just block that off so light doesn't leak through like the real thing...

Edit: thought I'd throw up a link to a large render I did too. Also, if people want this discussion moved out of this thread, please say and I'll do that 😀

https://www.dropbox.com/s/27mdgiijjycy3t3/201 … 51701a.png?dl=0

And nobody gives a damn about glass spheres on weaker hardware 😁 - Nah, nice modelling, I could never get a grasp on that stuff.

Yesterday I cleaned up my turntable and corrected the tracking force, then probably gave collectors a heart attack by dropping one of my favorite albums from 1968 on there. There are a few times things distort a little and I think I need to change the stylus soon as I've been abusing it for years, it has long surpassed its rated lifespan anyway (Last changed about 10 years ago). Here's a recording of it in action; (Edit - I removed the link)

If you like these song, buy it legit as they'll probably sound better to you without hum, pops and scratches anyway. I shall remove the link after 24 hours to stop potential pirates. The hum comes from the input on the computer mostly, it is still audible if I remove the cable so there's literally nothing I can do about it. I don't doubt the music center does add a little of its own being 70s electronics, I don't really care, this thing sounds more awesome than anything I've ever tried to replace it with. I think something is a bit wonky on the low power output as it doesn't always sound as bassy as it did, probably a passive component that has gone a bit dry or some such and I need to go in and replace some bulbs and fix the VU meters anyway, so I'll check it out when I get to doing that after my move.

My new PAS16 arrived too, so I've been testing that. Had a few issues setting the DMA up - seems I changed some setting when trying to fix the old card and forgot to put it back - but that was my fault and the card seems to work fine. Better than expected it outputs Line Level AND has the Speaker header, so now I don't need to build a resistor pack to hook it into the device I want to eventually hook it into.

Oh HighTreason, I think I've watched all your videos on YouTube haha 😊

I especially enjoy your 486 videos... but that's because I have a silly obsession with them!

Good work on the turntable, glad to see they are being kept in service. They just don't make them like they used to!

Yeah, I probably have an unhealthy compulsion towards the 486 too and I recognize your name from my comments section.

I'd post a picture of the turntable, but my camera is still out of action for now. I do have a solution to the problem and will fix it soon. May as well save the turntable until its big fix though, there's nothing really wrong with it and even some of the dead bulbs are actually loose and not blown - to my knowledge it's probably 45-48 years old and that is literally the only issue, it wears out belts from time to time and a bulb blew, other bulbs have come loose and two of the VU needles are stuck, the volume pot is a bit scratch and needs cleaning out, otherwise it is still kicking ass and goes loud enough to give the street a headache. I don't know what Sanyo did, but I have a Panasonic which claims to have an amp twice as powerful, yet it can out-drive that thing by miles and quickly beats it into obscurity when the dial is cranked, it literally drowns it, especially in the low-end when the Loudness is turned on.

I think the thermal paste is even holding up, it still felt pasty last time I went by the heatsink anyway. I don't know what it is, but there's something about the way 70s and some 80s audio gear sounds that just hasn't been paralleled since, presumably the people making them actually listened to music and tested them out instead of just poking a scope in and saying "Meh, it matches the spec sheet, ship it!" like the probably do now. It even has a smell to it to the point you can almost figure out when it came from on that alone.

Edit: New discovery!
My new PAS is wired differently to my old ones, so I tried re-wiring one of my old dead ones the same as the new working one and... It sort-of works now, but PCM from the PAS side plays too fast. Oddly, the previous owner of the new one said it did that when there was no -5V from the PSU, which I have. This old card, as you may know, had a blown amp which would have used that circuit, so I would bet some component went with it. Nonetheless, I get sound and it is perfect with the FM/SB side only there's no amp and the resistors I used are the wrong value so it is quiet. It looks like with a little probing I might end today with TWO working PAS16 cards instead of just one, which isn't something I will complain about as I needed a second one for another machine anyway. So double thanks to the dude who sold me the new one, you might have just saved my old one too. I should add; Don't worry, I'm not going to mess with the new card at all, that's installed in the machine and I'm going to leave it alone, it has served only as a visual reference in this experiment.

Not exactly retro (the activity in it self), yet me and my woman went thrifting today. Nothing of interrest, the usual clothes that did not fit her, and what did fit her was of no interrest to her. Me. I found some games that I allready have. Might go to a different store next time, 20 minutes ride further down the road.

I bought one those test bench cases to play about with my old hardware. It really useful to have easy access to change things out.
This is what I got up to today. Installed a socket 7 board and did some benching.

For anyone interested, it's an Aerocool Strike-X Air test bench case.

The last few days I've spent installing a voltage regulator into my 486 board, so I can run an Am5x86-133 in write-back cache mode.

My board is a VL/I 486SV2G rev 1.7.
I was using the excellent instructions from mAJORD: https://www.dropbox.com/s/bjjvk2jtebps2xj/rev … 205x86.doc?dl=0

My board normally only supports 5V CPU's. I was running a Kingston Turbochip, which is an Am5x86-133 wth a voltage regulator on a single chip; but this only runs in write-through cache mode.

Here's a pic of the board before starting work:

Here's after soldering in the new components (plus a sacrificial test CPU 😜 ) :

I also had to upgrade to the latest beta BIOS. Side note for anyone with a Genius G540 programmer.... it won't work on win 10. I installed Virtualbox plus the extension pack, and run up a win XP VM. I setup a USB1.1 controller (2 or 3 don't seem to work properly) on the VM and passed thru the programmer - this seemed to work fine.

After mucking about with jumpers, and confirming 3.45V on the VCC pins at the CPU socket, success!

Here's some benchmark results:
Before upgrade:

3dbench 76.9
pcpbench vgamode 21.8
Speedsys CPU 48.49
Speedsys L1 84.29 MB/s
Speedsys L2 53.06 MB/s
Speedsys Memory 31.39 MB/s
Norton sysinfo 198
Doom demo3 3863 2815 (48.03 fps)

After upgrade:

3dbench 83.3
pcpbench vgamode 22.3
Speedsys CPU 50.25
Speedsys L1 110.67 MB/s
Speedsys L2 42.41 MB/s
Speedsys Memory 25.28 MB/s
Norton sysinfo 288
Doom demo3 3863 2510 (53.8 fps)

A decent performance boost all round... interesting that L2 and Memory bandwidth in speedsys actually drop after the upgrade?? Any ideas what might be going on here?

And here's a shot of the machine:

I bought the case off a fellow member at OCAU... I had retrobrighted it when I got it but the yellow has come back even worse than before... I couldn't really be bothered doing it again.

The specs of the rig are:

Mobo: VL/i 486SV2G (with 1MB 15ns Cache)
RAM: 16MB
CPU: Am5x86-P75 (X5-133 W16BGC)
Video: Hercules Dynamite VLB (ET4000W32/p) with 2MB RAM
Controller: Promise EIDE2000 VLB
HDD: Seagate Medalist 2.1GB
CDROM: Sony CDU55E
Sound:
-SB16 CT2230
-Roland RAP10
-Gravis Ultrasound classic 3.4 with 1MB RAM

CelGen wrote:

Attach power supply cables to J20 for 5V power supply or J21 for a 3.3v power supply. (optional for OEM customers only)

So....I'm guessing there's an OEM power supply that handles 3.3v externally for some reason instead of regulation happening on the board itself. Meaning I could of had this machine built and running years ago.
Guess what I"m building this weekend? 😁

Yep. a typical AT power supply will work.

Supermicro suggests at least a 300W power supply. All I got is a 150 and a 200. I might still end up transplanting the guts of a modern ATX power supply into the old enclosure and grafting in the old AT cable harness.

CelGen wrote:

Supermicro suggests at least a 300W power supply. All I got is a 150 and a 200. I might still end up transplanting the guts of a modern ATX power supply into the old enclosure and grafting in the old AT cable harness.

Will this not work then?
http://www.ebay.com/itm/20P-ATX-to-2-Port-6Pi … VEAAOSwNSxVX3Xf

Going through my DOS library and testing for CPU speed sensitivity for the Vogons Wiki page. Added 5 games so far out of 10 games tested. Might get more done later today. 😀

While you're at it, could you add OMF 2097 (runs too fast) and Red Storm Rising (Adlib detection fail). Also, Quest for Glory III has scripting bugs with higher speed CPUs (third party patches exist). These can all be addressed with cache manipulation.

gdjacobs wrote:

While you're at it, could you add OMF 2097 (runs too fast) and Red Storm Rising (Adlib detection fail). Also, Quest for Glory III has scripting bugs with higher speed CPUs (third party patches exist). These can all be addressed with cache manipulation.

Or, you know.. actually running it on a slower computer.

while you're at it, I've spent a lot of time figuring out Jazz Jackrabbit I for MSDOS. The CPU error doesn't require a patch, it's just that it only works up to Pentium P54c cpu's. Pentium-MMX is the beginning of where it gets the "Runtime error 200" error. It doesn't need patching, just run it on an older chip and it's fine. Although, running it on a Pentium chip @ 166 mhz and it's in-game events get "too fast" and have timing issues trying to play it. 486 @ 133 is best.

I'm only adding games I have, since I can test at what CPU speed they start to go wonky. Already added QFG3, but the situation I saw was General MIDI getting garbled at anything higher than 16Mhz 486. I suppose I can retest with Adlib/SB music to try to flush out the scripting bugs.

I believe Jazz Jackrabbit is already on the list (someone else put it in): http://www.vogonswiki.com/index.php/List_of_C … sensitive_games

kithylin wrote:

I've spent a lot of time figuring out Jazz Jackrabbit I for MSDOS. The CPU error doesn't require a patch, it's just that it only works up to Pentium P54c cpu's. Pentium-MMX is the beginning of where it gets the "Runtime error 200" error. It doesn't need patching, just run it on an older chip and it's fine. Although, running it on a Pentium chip @ 166 mhz and it's in-game events get "too fast" and have timing issues trying to play it. 486 @ 133 is best.

I see what you're saying now. Although a 133Mhz 486 is not ideal to put for maximum suitable CPU (how many people have one of those?). Since you timing issues start at Pentium 166, if I put Pentium 133 as max suitable CPU, would that feel right to you?

Zapp3012 wrote:

For anyone interested, it's an Aerocool Strike-X Air test bench case.

Does it transform into a truck?

clueless1 wrote:

I'm only adding games I have, since I can test at what CPU speed they start to go wonky. Already added QFG3, but the situation I saw was General MIDI getting garbled at anything higher than 16Mhz 486. I suppose I can retest with Adlib/SB music to try to flush out the scripting bugs.

I believe Jazz Jackrabbit is already on the list (someone else put it in): http://www.vogonswiki.com/index.php/List_of_C … sensitive_games

I noticed Jazz in that list. Didn't know it was written in Pascal.

Epic released One Must Fall as freeware in 1999, so you can test at your leisure. Get it from one of our local MIDI aficionados, Rich Nagel.
http://jasonwilliams400com.startlogic.com/sno … t_Fall_2097.htm

gdjacobs wrote:

I noticed Jazz in that list. Didn't know it was written in Pascal. […]
Show full quote

clueless1 wrote:

I'm only adding games I have, since I can test at what CPU speed they start to go wonky. Already added QFG3, but the situation I saw was General MIDI getting garbled at anything higher than 16Mhz 486. I suppose I can retest with Adlib/SB music to try to flush out the scripting bugs.

I believe Jazz Jackrabbit is already on the list (someone else put it in): http://www.vogonswiki.com/index.php/List_of_C … sensitive_games

I noticed Jazz in that list. Didn't know it was written in Pascal.

Epic released One Must Fall as freeware in 1999, so you can test at your leisure. Get it from one of our local MIDI aficionados, Rich Nagel.
http://jasonwilliams400com.startlogic.com/sno … t_Fall_2097.htm

Nice...thanks. Always good to add one to the collection. Going to check it out in a minute.

edit: Interesting that under sound setup they offer a "Pentium" option, yet when I run at Pentium 120 speeds, it's way too fast. My next slowest speed step is 25Mhz 486, and it seems to play fine at that speed. I'm not really in a place now to pop the lid off my rig and change jumpers, but that's the only way to get a slower Pentium speed out of my rig. I'd wager it starts getting too fast somewhere between P75 and P90?

Today I removed a few of barrel batteries from 386 and 486 boards I had put away over 18 months ago. My procrastination cost me one 486 board which isn't too bad, it looks too far gone to revive. I removed any left over battery acid on the other boards and cleaned up with vinegar.

brostenen wrote:

CelGen wrote:

Supermicro suggests at least a 300W power supply. All I got is a 150 and a 200. I might still end up transplanting the guts of a modern ATX power supply into the old enclosure and grafting in the old AT cable harness.

Will this not work then?
http://www.ebay.com/itm/20P-ATX-to-2-Port-6Pi … VEAAOSwNSxVX3Xf

It would, but not without some hacking. An ATX power supply will not fit in this case without modification. Seeing how expensive high wattage AT power supplies are it wouldn't hurt to do the transplant anyways.
Also, it seems my board is of a later revision. The manual and its diagrams and wiring info within indicate just the AT power connenctor is all you need however there's signs the ground and 5V pins have overheated more than once on this board. I guess 300w is a lot for this style of socket. While not mentioned anywhere, my board has an additional six pin header which gives you three more 5V and ground lines to the machine, presumably for the much heavier load. I guess while I'm doing the PSU conversion I can also splice in a little more extra cabling for this and eliminate the overheating issues.

So I carried on working on the machine. I needed better heatsinks and fans. I dug through my stockpile and found two mismatched heatsinks and fans. One of the fans needed a resistor or two to quiet it down. Both heatsinks were extremely tight as I had ceiling tolerances plus the VRM's are way too close to the CPU's. One regulator had a piece of tape placed over it to prevent it shorting out while the other heatsink had to be trimmed before it would fit at all.

I'm also completely out of CR2032 cell holders. Being desperate to get this thing up I scrounged up a good battery and holder normally intended for a mac. It will work with a diode added in however these cells are well known in to past to leak so I don't know if I want to keep this or hold out for more coin cell holders.

One of the cool things about this case is there's a volume knob. There' no onboard amp or accommodation for a loopback from a sound card but there's still the PC speaker. I might build a little adapter board so I can add a volume control for it. 😎

The "crtfix" patch must fix more than just the runtime 200 error in Jazz Jackrabbit; I've played the patched version of the registered game on Pentium 4 machines with no visible problems whatsoever.
Also works perfectly (with sound and music) on my Pentium MMX 166 machine.