How about that ... I've never seen those dual drives. Can you insert two floppies at the same time ?
Or do you switch between them ?
It's 2 physical drives in low-profile enclosures squeezed into a single 5.25 drive bay sled. The two drives share a data connection and power but are otherwise completely independent. One is the A: (3.5" by default) and the other is the B: drive. This can be swapped with a jumper in the guts of the thing.
You can insert a floppy in each drive and copy from A: to B: or vice versa.
Just for the heck of it, I swapped the JAZZ16 for a VIBRA16 card.
Now, the JAZZ16 is a pretty decent card with good DOS support and SBPro mixer compatibility. The genuine Yamaha OPL3 synth is nice, but it's still a clone card and I had two major (and 1 minor) issues with it:
1. The MPU-401 emulation would not work. I tried lots of stuff, and even after setting the enable jumper, reserving IRQ 9, and configuring the driver, nothing would recognize it as present and working.
2. Occasionally digital sound would completely flake out and produce noise until restarting the application.
3. You can't disable the CD interface, so those resources are wasted unless you happen across a compatible LMSI drive.
I decided to try something else, the CT2940 VIBRA16. More than just a replacement for the JAZZ16, this card has some interesting features and obstacles:
1. It is a genuine Creative Labs chipset and features a CQM synthesis chip (the CT1978) instead of the OPL3.
2. It's a cost-reduced SoundBlaster 16 popular with OEMs and sold at retail as a value brand.
3. It's a Plug-and-Play card.
4. It's a Gateway card.
This last fact is important to this build on principle. You can tell it's from a Gateway system because of the Gateway part number sticker (SNDCRD005) and the date sticker (10/24/95) on the back.
The fact that this is a Plug-and-Play card means there's no hardware jumper configuration. Instead, the card is meant to be configured dynamically by the operating system or ICU in a system with a Plug-and-Play BIOS. However, the secret to using this card is that none of that is actually necessary, and just like a hardware-configured SoundBlaster, the CT2940 is usable without any memory-resident drivers or management suites. All you need to do is "free" the card's IRQ in BIOS, execute a CTMU.EXE program to get the card going, and use the SB16 standard applications MIXERSET.EXE and DIAGNOSE.EXE to configure and soft-set the card. After that, it's ready to roll on port 220h and IRQ5 like any other SB16.
If instead you install the driver package as packaged, you end up with all kinds of extraneous TSRs in your config files, which slows down your boot and eats up your RAM. Ironically, the CT2940, a Plug-and-Play card, results in lower memory usage than the older JAZZ16.
The only problem I had swapping this card into the system was some CMOS configuration crud that produced benign errors at POST. After clearing NVRAM with the mainboard jumpers, this went away. The card works in DOS, Windows 3.1, and Windows 95 with virtually no hassle.
And with what should be an operational MPU-401 interface (albeit hampered by the infamous "hanging note" bug) I can move on to experiment with off board MIDI synths and emulators.
The output signal is lower but cleaner than the JAZZ, too. Even with the software mixer maxed out, I have to come off Unity gain on my hardware mixer to get a good output level.
The motherboard occupies about half the case. It's interesting to recall how tightly integrated these systems used to be: note the first two RAM slots lie at an angle, which provides clearance for the 3.5" drive cage to be installed immediately above it.
IMG_1474b.jpg
With the drive cage installed you can barely even see the RAM living beneath it.
Those two angled ram slots are labeled "Bank 0" on the board. I don't ever recall seeing the slots furthest from the cpu meant to be populated first. But it makes sense in this case; its convenient for a user to add memory to the other two slots, they're easily accessible.
This is interesting. The PBA number on this Plato board is 541277. According to the manual, it will only work with the 75 and 90MHz Pentiums. But when I put in a P100 and set the jumper for 100MHz, it goes into BIOS, shows the correct speed there, and boots from the floppy just fine.
IMO, that fact changes everything, and makes this board worth reviving; worth spending a little time with the dremel fixing the RTC chip. A 100MHz Pentium system is a very useful machine for software of that period. Kind of the archetypal machine for Win 3.1 really, and maybe even 95.
As good as the slot-load CD-ROM drive looks in the setup, it's a flaky drive. It's a Pioneer DR-501S, nominally a 24X from 1997, and is a later addition to the system; the original shipping drive is long gone. While this drive works okay for installs, it fails pretty rapidly when used for any kind of multimedia access including playing digital music (not CD Audio) from titles like Command & Conquer, Hereos of Might and Magic, and The 11th Hour. Even installing Command & Conquer was an exercise in frustration. Any such access produces stutters and micro-pauses, rendering the running software unusable.
I swapped it for the only other IDE CD-ROM I have access to, a TDK 4800B, a nominally 48X drive with a garish blue caddy, yellowing faceplate, and a motor that sounds like a jet engine at full speed. On the plus side, it succeeds in all of the tasks the Pioneer fails at.
The noise this thing makes is ridiculous, and it benchmarks at barely above 10X performance which makes it even more ridiculous. Luckily, it responds to CDBQ, at least a little bit. CDBQ, if you don't know, is a DOS utility that sends an ATAPI command to a CD drive to set its speed. Not all drives respond or obey. This TDK drive seems to obey in steps -- anything below around speed "1200" (in kb/sec, roughly 8X) results in 4X performance, but telling to operate at speed "1500" (10X) results in not only quiet operation but it benchmarks not far off of its "full-throttle" performance! The only downside is that the speed resets every time the drive is opened, so you have to execute CDBQ every time you swap a disk.
Full-speed, produces too much noise for barely better than 10X performance.
Another option for slowing down a CD-ROM drive is the TEAC driver which supports a "speed" setting. Unfortunately, while the driver detects this drive and loads, it produces sub-4X performance without reducing the noise.
Unless I happen across a 4-8X vintage drive, I'll probably stick with this TDK for a while. Quiet performance at 8X speeds is a good compromise for the software this computer can run.
The AnyKey keyboard is more than a punny name. It was so called because of its ability -- for good or ill -- to reprogram any key on the keyboard except the keys that perform said programming.
This keyboard was filthy when I got it. Not only full of scum and hair, but just covered in a layer of grime. I ran it through the dishwasher and scrubbed every keycap, and now it's as clean as the day it was new.
The thing is the size of a cafeteria tray and features era-standard rubber-dome key switches. It predates the USB era and has a PS/2 connector. It predates Windows keys so has neither. Instead it duplicates the asterisk and pipe/backslash on the lower row. It also duplicates the entire run of 12 function keys along the left side.
It's not the worst keyboard to type on but when you're used to modern boards the oversized enter key and undersized backspace are really irritating, especially since that oversized Enter means the pipe/backslash is in the "wrong" place. When you look at vintage keyboards, that key seems to roam all over the place anyway.
The navigation block is unusual in that it has 8 arrow keys instead of the usual 4. In the middle of the grid is a blank key that is a second space bar (key, really, but who says "space key"?)
Gateway shipped some AnyKey software with these things, but all it does it backup/restore the keyboard's EEPROM. Otherwise, all programming is self-contained to the keyboard, which requires no TSRs or drivers or anything special to make happen. You just use the "Program Macro" and/or "Remap" keys to make AnyKey work like AnyOtherKey or store an arbitrary macro of keypress sequences. It also hosts its own programmable repeat rate.
The macro programming is the feature that makes sense of the otherwise unnecessary extra dozen function keys. These are by default duplicates of the regular top-row function keys but once remapped or programmed, they are their own thing -- programming F1 to perform some arcane macro sequence does not affect the other F1.
Of course, the AnyKey was also an office prank waiting to happen, and I fielded more than one tech support call about it. The magic reset sequence is "CTRL-ALT-SUSPEND MACRO" which resets all the keys to their unprogrammed and unmapped default state.
Edit: I didn't think I had a "before" picture but I just remembered that I do!
I've also attached the last BIOS released for the board.
Can I flash it over gateway bios? 1.00.12.AX1T
I also have an image of the original disk for the gateway system, if anyone's interested.
My system is from 1995, with 75mhz cpu and vibra16 card
Installed a socketed Dallas chip today: bought from digikey
Should work, yes. The Intel and Gateway boards only differ in their BIOS. The flasher will stop when it detects any incompatibilities anyway, so there's not much that can go wrong. Just let it make a backup of the current BIOS when it asks for it in case you need to revert.
Should work, yes. The Intel and Gateway boards only differ in their BIOS. The flasher will stop when it detects any incompatibilities anyway, so there's not much that can go wrong. Just let it make a backup of the current BIOS when it asks for it in case you need to revert.
What is the driving purpose behind updating the bios? Do you have release notes for the newer intel version? It won’t change the HDD support. What are you after?
