HanSolo wrote on 2022-12-05, 18:46:

As an Exe-packer I would recommend aPACK. That gave me better results than UPX on a 4K-Intro 20 years ago. (But things might look different with larger Exe-files)

Have you considered using the higher density format of Fdformat that you also mention in the video? I used that back then for some of my disks.

Ill keep aPack in mind since for some programs the UPX result was broken while the original one was with aPack.
I have considered the 1.72M format, but wanted to maintain maximum floppy drive compatibility since not all drives can handle that well.

henk717 wrote on 2022-12-05, 19:28:

HanSolo wrote on 2022-12-05, 18:46:

As an Exe-packer I would recommend aPACK. That gave me better results than UPX on a 4K-Intro 20 years ago. (But things might look different with larger Exe-files)

Have you considered using the higher density format of Fdformat that you also mention in the video? I used that back then for some of my disks.

Ill keep aPack in mind since for some programs the UPX result was broken while the original one was with aPack.
I have considered the 1.72M format, but wanted to maintain maximum floppy drive compatibility since not all drives can handle that well.

Is that some kind of overburn feature for floppies? Or more bits per track?

PD2JK wrote on 2022-12-05, 20:58:

henk717 wrote on 2022-12-05, 19:28:

HanSolo wrote on 2022-12-05, 18:46:

As an Exe-packer I would recommend aPACK. That gave me better results than UPX on a 4K-Intro 20 years ago. (But things might look different with larger Exe-files)

Have you considered using the higher density format of Fdformat that you also mention in the video? I used that back then for some of my disks.

Ill keep aPack in mind since for some programs the UPX result was broken while the original one was with aPack.
I have considered the 1.72M format, but wanted to maintain maximum floppy drive compatibility since not all drives can handle that well.

Is that some kind of overburn feature for floppies? Or more bits per track?

More sectors (up to 21) per track and more tracks (up to 82).

Since I don't plan on frying eggs I made a bracket for an additional 50mm fan next to cpu cooler, for cooling the voltage regulator.

Screw holes are too small at 4mm, of course. Will be fixed tomorrow.

Since I had to buy an E/EPROM progammer for recover a corrupted BIOS chip, today I successfully installed an XT-IDE rom into a NE2000 compatible ethernet adapter in my IBM PS/ValuePoint.

Finally it can stretch its legs a bit more.

I had to use the 386 binary version of XT-IDE and the ethernet adapter configured for use D000-D7FFF range because the C000-C7FFF is occupied by the video rom.

A001 wrote on 2022-12-07, 18:42:

Since I don't plan on frying eggs I made a bracket for an additional 50mm fan next to cpu cooler, for cooling the voltage regulator.

Screw holes are too small at 4mm, of course. Will be fixed tomorrow.

760.jpg

Nice, I've got the same board I think TC430HX? That regulator gets toasty so the extra fan is a good addition. I think I screwed a 92mm fan over both the CPU heatsink and regulator on mine.

I finally got my Compaq Armada 7350MT back into a functioning state! Compaq hardware is so weird and I kind of like it for that - it needs a special partition on the hard drive just to load the BIOS setup. It's been such a long road to get this thing back to functioning, first I had to:
- rebuild the casing since this thing is 1997 plastic and that didn't hold up well
- find the rather unique power adapter, only used ever by the Armada 73xx and 74xx series
- troubleshoot RAM faults, it refuses to boot with most of my EDO 144 pin SODIMMS
- fix the LCD screen, the connector for inside the LCD panel, for its column driver had popped unplugged = just displays mess - man I was happy when I got it working though, the LCD is a weird 12.1" parallel AA121SB12 and those are not cheap to replace
- reverse engineer the hard drive connector's unique pinout: Finding pinout for Compaq 50-pin HDD connector on Armada 7300 series Eventually I got an Armada 7400 with the caddy to complete that, all working now including DMA!
- found out the LCD has this really weird thing where some sections of the display lose detail after it's been running a while
- battery is not charging, oh well. Maybe I'll fix that, maybe I won't.
- The super unique folding down keyboard is all working
- The wavetable audio is working great! the speakers are quite good on this old thing

So as it stands, the screen gets a bit unusable after an hour or so of usage and I can't go over 16MB of RAM right now but I'm so happy this thing is functioning again 😁

I just "fixed" my IBM 5150. The other day some friends were over and one asked about the old PC so I fired it up and was greeted by the text mode being out of sync with the display. I remembered the location of a game and fired that up and that mode worked fine so I knew it wasn't anything serious.

I had a chance to look at it today and remembered that I had used it to test out an NEC Multisync II I picked up back in the spring, so, of course I had changed the DIP switches on the EGA card! Then I'm pretty sure I left the poor IBM on my desk as a test setup for months. I eventually reassembled it and had totally forgotten that I'd changed anything. Oops.

So, I got out the 36 year old manual for the Everex EGA card, flipped two switches and the system was back in business. It always amazes me just how good the old 5153 display looks when fed 16 colors.

While tinkering with it I found some text files that my daughter (now 9) had typed when she was 3 and when she was 6. It was so cute... haha. This is what retro computing is all about folks. 😁

Thermalwrong wrote on 2022-12-08, 02:50:

A001 wrote on 2022-12-07, 18:42:

Since I don't plan on frying eggs I made a bracket for an additional 50mm fan next to cpu cooler, for cooling the voltage regulator.

Screw holes are too small at 4mm, of course. Will be fixed tomorrow.

760.jpg

Nice, I've got the same board I think TC430HX? That regulator gets toasty so the extra fan is a good addition. I think I screwed a 92mm fan over both the CPU heatsink and regulator on mine.

TC430HX Yes. 92mm is the sensible choice but these 50mm screamers are part pf the experience.

listened to some game midis on my 98 build, sc-55 and klipsch promedia 2.1 THX, nearfield sounds crispy, loud and punchy 😀

Not quite retro, replaced the broken grounded kettle socket on the cheap medium sized board preheater, socket was flopping around in the air due to broken thin plastic ears with quality one made of metal and thick plastic. Had to dremel the opening bit wider. Routinely used for the game console repair using preheater to assist with replacing HDMI port connector so need to be safety sake.

Cheers,

Tested out my recently acquired I-O Data PK-X486S50-3 (286-to-486 upgrade).

I had originally hoped to use it in my Tandy 1000 TL/2, but it won't physically fit on the motherboard due to the location of the 5.25" drive bay cage.

Fortunately it does fit in my Tandy 1000 TL and appears to work with what I'm assuming is a default base clock of 16 MHz.

There are a series of four dip switches on the board, but I have no idea what they do and haven't found any documentation on them. Initially switch 1 was set to on, but the system wouldn't boot. Setting them all to "off" allowed the system to boot.

Finally got around to test boards and one of them is SiS 551x based pentium socket 5 has corrupted bios firmware, saying bios checksum is bad. Need to find time to make new fresh firmware for the P/I-P55SP4.

Update: Board is not doing too well after refreshing the bios, keep saying 8MB no matter SIMMs I use and detects no cache with either 32k x 8 or 64k x 8 chips. Jumpers set correctly. Only thing is CMOS batteries are all flat. I searched my stock of boards with battery coins and two I found is also flat!

Ordered fresh Intel flash chips (NOS) and 10 pack of CR2032 cells.

Really wanted Asus P/I-P55SP4 to be set up with async 512K cache but this board is not going anywhere yet until I straighten out the no cache detected and memory stuck at 8MB no matter SIMMs I use, using pentium 133 CPU. Future plan is put together a slow combination of async, winchip for old DOS machines.

Looks like my compaq computers are winning so far, they simply keep working.

Cheers,

Shponglefan wrote on 2022-12-10, 03:00:

Tested out my recently acquired I-O Data PK-X486S50-3 (286-to-486 upgrade). […]
Show full quote

Tested out my recently acquired I-O Data PK-X486S50-3 (286-to-486 upgrade).

I had originally hoped to use it in my Tandy 1000 TL/2, but it won't physically fit on the motherboard due to the location of the 5.25" drive bay cage.

Fortunately it does fit in my Tandy 1000 TL and appears to work with what I'm assuming is a default base clock of 16 MHz.

There are a series of four dip switches on the board, but I have no idea what they do and haven't found any documentation on them. Initially switch 1 was set to on, but the system wouldn't boot. Setting them all to "off" allowed the system to boot.

Would one of those work on a Tiny-Turbo 286 upgrade card for an 8088 machine? I have one of those in my IBM 5150. The 8Mhz 286 is already a massive improvement and is very stable and compatible, even with the CPU on the end of a ribbon cable (LOL)... but I'm not sure if 486-level CPU performance would provide any real benefit for such an old PC. I feel like I've seen this done before, but I don't know if it involved this exact combination.

Ozzuneoj wrote on 2022-12-11, 00:16:

Would one of those work on a Tiny-Turbo 286 upgrade card for an 8088 machine? I have one of those in my IBM 5150. The 8Mhz 286 is already a massive improvement and is very stable and compatible, even with the CPU on the end of a ribbon cable (LOL)... but I'm not sure if 486-level CPU performance would provide any real benefit for such an old PC. I feel like I've seen this done before, but I don't know if it involved this exact combination.

No idea if it would work with a Tiny Turbo. But in both these setups (5150 or Tandy TL), I expect the underlying XT architecture to be a significant bottleneck.

Shponglefan wrote on 2022-12-11, 02:01:

Ozzuneoj wrote on 2022-12-11, 00:16:

Would one of those work on a Tiny-Turbo 286 upgrade card for an 8088 machine? I have one of those in my IBM 5150. The 8Mhz 286 is already a massive improvement and is very stable and compatible, even with the CPU on the end of a ribbon cable (LOL)... but I'm not sure if 486-level CPU performance would provide any real benefit for such an old PC. I feel like I've seen this done before, but I don't know if it involved this exact combination.

No idea if it would work with a Tiny Turbo. But in both these setups (5150 or Tandy TL), I expect the underlying XT architecture to be a significant bottleneck.

Yeah, in the context of modern computing, it's incredible to think that these machines were capable of anything at all when you think about the bottlenecks and the way the systems were designed. I mean, modern PCI-Express speeds are apparently highly dependent on board construction and trace length (the reasoning behind 4xx-series chipset boards not being able to have Gen4 speeds reliably)... and yet on these PC\XT systems the CPU, RAM and every component run all over the place through incredibly slow interfaces, mostly over an 8bit bus. For goodness sake... I upgraded it by sticking the original CPU in a socket on an ISA card that also has another CPU on it.. and there's a lonnngg ribbon cable that plugs into the CPU socket on the motherboard. How does that improve things and not cause problems? 😀

Or how about today's CPUs having stacked dies to add scads of high speed cache! My 5800X3D has 150x as much cache as my PC 5150 has system RAM... heck it has 5x as much cache as the PC has hard drive space. 🤯 ... and yet, to perform all of it's tasks, the lowly 5150 was given... no cache. None. And may not have even benefited from it if it had any. 🤣

Crazy time to be alive.

The reason for AMD 4XX boards not supporting PCIe 4 was a bit more complicated and mostly due to fabs cost cutting thinking they could get away with using gen 3 traces. Length really was not an issue as the lengths didn't change between the boards what did change was signal integrity due to PCIe 4 requiring more robust traces and higher trace voltages than PCIe 3 did.

There were a few 4XX boards that did support PCIe4 quite well for the first PCIe slot but AMD forced even the compatible 4XX boards to adhere to their BIOS changes in supporting 3.0 only, you could keep using the older BIOS however and keep the working PCIe4.

I spent several hours today working through my pile of Zip Disks, so far they all work and I now have 30 Zip 100 disks and 20 Zip 250 disks ready for my retro rigs. The USB Zip 250 drive works amazingly well and even under DOS it behaves nicely, one of the Zip disks had all the Iomega tools on there so the Zip250 will now behave itself under Win3.11 and Win95 too.

TrashPanda wrote on 2022-12-13, 04:19:

I spent several hours today working through my pile of Zip Disks, so far they all work and I now have 30 Zip 100 disks and 20 Zip 250 disks ready for my retro rigs. The USB Zip 250 drive works amazingly well and even under DOS it behaves nicely, one of the Zip disks had all the Iomega tools on there so the Zip250 will now behave itself under Win3.11 and Win95 too.

that's a lot

back when usb wasnt a thing i used zip 100 to transfer 'big' files between computers using parallel

i still have the drive i bought all those years ago and it still worked when tested a couple of years ago, but i cant see me using it now, i just dont have that many pre usb PCs anymore

gerry wrote on 2022-12-13, 09:24:

that's a lot […]
Show full quote

TrashPanda wrote on 2022-12-13, 04:19:

I spent several hours today working through my pile of Zip Disks, so far they all work and I now have 30 Zip 100 disks and 20 Zip 250 disks ready for my retro rigs. The USB Zip 250 drive works amazingly well and even under DOS it behaves nicely, one of the Zip disks had all the Iomega tools on there so the Zip250 will now behave itself under Win3.11 and Win95 too.

that's a lot

back when usb wasnt a thing i used zip 100 to transfer 'big' files between computers using parallel

i still have the drive i bought all those years ago and it still worked when tested a couple of years ago, but i cant see me using it now, i just dont have that many pre usb PCs anymore

I have both a Zip 100 and 250 in parallel and they both work, I use them for my machines that run DOS, the 250 USB does in a fashion work under DOS but its a bit fiddly getting the DOS USB drivers working right and they are TSR so eat memory that the parallel doesn't.

I could use CDRs but its a waste of media for just copying across files and I hate pissing around setting up local networks under DOS just for file transfers.

Im thinking about getting a Zip 750 with some disks to complete the collection, they are actually much cheaper than either the 100 or 250 which is a bit odd.

First shipment for my "Let's replace broken SB32 SIMM socket" project arrived - Blingy low profile Purple RAM.
In order to test them I had to implement temporary fix. Even managed find matching tan colored nylon thread. 🤣