• Please review our updated Terms and Rules here

BOOK8088 and serial cards - working @ 9600

Phil_G said:
That will be my first job when the thing arrives Sergey.
Presumably I use this image: bios-book8088-xtide-v20.rom ?
Tracking says the Book8088 has arrived in the UK! Are 82C450 uart's supported please?
Click to expand...
Yes, that's the image for V20 CPU. The only difference with the other image is that it uses XT-IDE BIOS extension optimized for V20/80186+, which, according to some testing 50% faster...
I guess, you meant 16C450 UART. BIOS doesn't care about the UART type. It will support 8250, 16450, 16550, 16750, whatever... as long as it is 8250 compatible. The BIOS code doesn't use any advanced features of the newer UARTs (like FIFO).
 
sergey said:
There is schematics for this thing. Two top CPLDs on your photo are CGA logic. Of course CGA also uses the 6845 CRTC chip, EPROM for the character generator, and SRAM (the narrow 28-pin DIP) for video memory.
You could try pulling these CPLDs out, and perhaps other CGA related chips too, if you want to remove the CGA functionality.
XT-IDE is implemented using the SMD ICs on the top right of the board by the Compact Flash socket. The smaller PLCC package is actually an 8237 DMAC, not related to XT-IDE.
Click to expand...

Hmmm, where's the XT-IDE? Cause, I pulled the chip with the red arrow shown below - and the whole system still seems to work just fine. booted off CF card, audio all worked, all same software, etc.

1693097706687.png

I tried pulled those larger CPLDs - no matter which one I pull, or all of them, the system doesn't give the normal IBM PC BEEEEP BEEP BEEP at startup, and on the LCD (TFT?) it just says "VGA" as if it looking for a VGA input signal. I didn't pull the CGA chips on the side though yet.

I yanked out this SMD near the CF like you said, trying to disable the XT-IDE... and the darn thing just still booting up with XT-IDE :D




And - using the full size 5150 to reconfig the NIC is a good idea. Just lazy to open it and swap cards out, maybe in a few weeks after VCFMW.
 
voidstar78 said:
Cause, I pulled the chip with the red arrow shown below - and the whole system still seems to work just fine. booted off CF card, audio all worked, all same software, etc.
Click to expand...

You pulled out the 82C37 DMA controller, which likely has zero to do with the XTIDE. It’s… interesting, that the BIOS doesn’t complain that it’s gone, although strictly speaking nothing in the base configuration of this thing needs it. The only thing DMA does in a base IBM 5150/60/near clone is provide DRAM refresh cycles every 15-ish(?) microseconds, and this computer uses SRAM. I would have assumed that for maximum cycle compatibility the BIOS would have set up the DMA controller and timer to do these cycles anyway, but maybe not? This might explain why apparently this computer fails the Check-It system board check that looks at DMA 0?

This does raise the question in my mind: has anyone actually succesfully driven a DMA using card with this laptop? Maybe the DMA controller is a doorstop.

Looking at the schematic you are very much barking up the wrong tree. The way to kill the internal CF support is you need to flash a BIOS chip that doesn’t have the XTIDE code on it, and then simply don’t plug in a CF card. You can’t disable the decoding for the CF port without reprogramming the IO CPLD, but it looks like the port is unbuffered, so if you leave the card out nothing will reply to the decode.
 
voidstar78 said:
I tried pulled those larger CPLDs - no matter which one I pull, or all of them, the system doesn't give the normal IBM PC BEEEEP BEEP BEEP at startup, and on the LCD (TFT?) it just says "VGA" as if it looking for a VGA input signal.
Click to expand...

What do you expect the internal LCD to do with the driver chip pulled out?
 
Well, like if no MDA or CGA card is inserted, it should beep, do memory check, and disk drive (or hard drive) should spin up like it is trying to boot [ actually not 100% sure on that, can't recall ever trying to start a 5150 with no video card at all ]
EDIT: i.e. if it wasn't even going to TRY to boot, I was gonna hook up the external CGA card and CGA/VGA adapter (I don't have an actual CGA monitor anymore)


and... I bent a pin on that little SMD, it was really hard to pull out. Not too bad, think I can fix it, but if it's benign, maybe I won't bother.

I don't have any flash programmers, but maybe it's time to look for one.
 
Last edited:
Still no luck on GLX.COM tried /oSK and /oAB and just giving a filename. It just squeals some noise.

PETSCII Robot still plays its audio and music on the Adlib very well.

Bummer can't use this thing as a little MOD player.
 
Plasma said:
I still don't understand the fascination with this pile of garbage. For $200 you can get a "real" XT that is properly engineered and doesn't require crap hanging out of the back for expansion...
Click to expand...

I mean, for what it is it’s… kinda cool, but if you want to run an external monitor and all that jazz, yeah, there are much better options.

FWIW, the schematic demonstrates clearly that the ”ISA bus” is electrically completely non-compliant and the only expectation you can have for external cards is if they break you’ll get to keep both pieces. This thing uses a single set of 74HC, not 74HCT, bus drivers, to run all the onboard peripherals, memory, and chipset components *and* the CF slot directly off the same lines that run to the bus connector. Which means that A: the fan out is already pretty high (which isn’t a catastrophically big deal if everything onboard is CMOS, but it does still have electrical implications) and B: expansion cards with 74LS logic on them might have problems driving the data bus low enough. General rule: HC can drive LS reliably, but the reverse is not true. It’ll *probably* work with low fanout but this thing is *not* set up well to drive a shed load of slots.

Anyway, yeah, if a desktop XT is what you want you really should be looking at either an old PC or something like a NuXT. This machine is literally a toy in terms of design, which, yeah, is fine if that is what you want, but set expectations accordingly.
 
Have you heard of "coffin apartments" in Japan? And this hotel in DC I was at recently, the room was under 200sqft. Worlds' trying to make that "normal." People just don't have space for a big box computer anymore, real estate is gettin' expensive.

One good thing I could say about this book8088 is that it's silent. But I just tried keen4 on it, that didn't run either.
edit: actually keen4 might be EGA only. Mach3 runs well on this book8088 thing (1987 title, CGA).
 
Last edited:
Someone asked if DESQView would run on this thing...

DESQView 2.25 is running, I opened two windows (one 256KB and one 128KB). But... I couldn't type in either window! It was weird, only about every other key would work (like letter a, c, g) and of the ones that did work, they would stutter (like "aaa"). When I exited and closed DESQView (press ALT, select Exit DesqView from the menu), the same keyboard worked fine in regular MS-DOS 6.22 that the system boots up with. So don't think it was a keyboard connection issue.

DESQView 3.5 wouldn't run. It just sat there, I gave it a good minute waiting, but the cursor just sat there and I had to reboot.
 
voidstar78 said:
Have you heard of "coffin apartments" in Japan? And this hotel in DC I was at recently, the room was under 200sqft. Worlds' trying to make that "normal." People just don't have space for a big box computer anymore, real estate is gettin' expensive.
Click to expand...

A NuXT MicroATX board will give you a more powerful *and* more reliable computer with four open (properly engineered) slots in a space no bigger than this thing plus the ISA backplane board hanging off it, and can be properly mounted in a box you can tuck behind/under your monitor. That argument doesn’t hold water at all if you want slots.

FWIW, there are *separate* builds of Keen4 for both EGA and CGA. The CGA build is… uninspiring, on XTs but technically works.

voidstar78 said:
DESQView 2.25 is running, I opened two windows (one 256KB and one 128KB). But... I couldn't type in either window! It was weird, only about every other key would work (like letter a, c, g) and of the ones that did work, they would stutter (like "aaa"). When I exited and closed DESQView (press ALT, select Exit DesqView from the menu), the same keyboard worked fine in regular MS-DOS 6.22 that the system boots up with. So don't think it was a keyboard connection issue.
Click to expand...

Desqview doesn’t work right with similar symptoms on my Tandy 1000s that coincidentally share some quirks with this thing, like no DMA and weird keyboard controllers so… color me not surprised.
 
sergey said:
I guess, you meant 16C450 UART. BIOS doesn't care about the UART type. It will support 8250, 16450, 16550, 16750, whatever... as long as it is 8250 compatible. The BIOS code doesn't use any advanced features of the newer UARTs (like FIFO).
Click to expand...
Apparently the UART is designated 82C450 but I found this:
Dave Platt said:
One entry in the netbsd-help archives states that "82450 and 16450 are
for all practical purposes, identical."
Click to expand...
Nothing has arrived yet except a warning from the Post Office that I will have duties and a handling fee to pay... :-(
 
voidstar78 said:
Hmmm, where's the XT-IDE? Cause, I pulled the chip with the red arrow shown below - and the whole system still seems to work just fine. booted off CF card, audio all worked, all same software, etc.

View attachment 1263203

I tried pulled those larger CPLDs - no matter which one I pull, or all of them, the system doesn't give the normal IBM PC BEEEEP BEEP BEEP at startup, and on the LCD (TFT?) it just says "VGA" as if it looking for a VGA input signal. I didn't pull the CGA chips on the side though yet.

I yanked out this SMD near the CF like you said, trying to disable the XT-IDE... and the darn thing just still booting up with XT-IDE :D
Click to expand...
Maybe I wasn't clear. The smaller 44-pin PLCC chip labeled CS82C37A is a DMA controller. It has nothing to do with XT-IDE. Actually, looking at the schematic, the compact flash card is connected to the ISA bus directly, without any buffers. The U29 "XTIO" CPLD decodes the I/O addresses and generates the chip select signals. To disable the on-board CF, you can simply pull out the CF card, and nothing will respond to the requests to the 300h-31Fh I/O ports, so it won't conflict with your NIC. (But this raises the question of how you're going to boot your system)

Again, the easier thing to do, is to configure your ISA NIC in another system to use I/O ports other than 300h-31Fh, e.g., 320h.

Based on the schematics, it appears that the best way to disable on-board CGA would be connecting a 10k pull-up resistor between U17 pin 19 and 5V (e.g. pin 20 of the same U17 IC), and then cutting the trace between U17 pin 19 and U26/CPLD pin 24, and putting a switch or a jumper there.
This way when the switch or the jumper are not connecting OE (U17 pin 19) to the DATA_GATE (U26 pin 24), the pull up resistor will keep U17 disabled, effectively disconnecting the CGA circuit from the data bus.

Simply pulling out CGA CPLDs might and likely will not work correctly, as it will leave that pin 19 of U17 floating randomly throwing garbage to the data bus, which will make the whole system unstable.

The pirated 8088 BIOS used in Book8088 does not check for CGA/MDA presence, it will continue booting without any beeps (other than the boot melody) even without display adapter present.
Feel free to submit a bug / feature request if you feel this is wrong :)
 
Got one of these hopefully on the way to me later in September......
Now what I'd be interested in seeing is the source code for the CPLDs.... with that all sorts of things would be possible.

One thing I wondered about was replacing the Upper 128K RAM with a 512K, you'd of course need to also run A17, and A18 to it. With the right CS logic for it you could have UMB blocks, in the areas not filled with CGA & ROM.

Cheers.

Phill.
 
Mine arrived, for Customs it was marked as "airconditioning parts" so no duties to pay! :)
So far I'm really pleased with it, I used to do a lot of novelties & utilities for PC magazine cover-disks and I've been reminiscing my way through my 30-year-old source.
One thing though, back in the day I was in the habit of using Scroll-lock as an enable/disable switch for various TSR functions, but the Book 8088 doesnt have a Scroll-lock key :)
Impressed with the battery - it lasts ages!
 
Eudimorphodon said:
I would have assumed that for maximum cycle compatibility the BIOS would have set up the DMA controller and timer to do these cycles anyway, but maybe not? This might explain why apparently this computer fails the Check-It system board check that looks at DMA 0?
Click to expand...

DREQ0 is pulled to ground and the PIT OUT1 line is unconnected, so no DRAM refresh. Unless that is you hack it back in.
 
GloriousCow said:
DREQ0 is pulled to ground and the PIT OUT1 line is unconnected, so no DRAM refresh. Unless that is you hack it back in.
Click to expand...
I suspect they not only used my BIOS, but also Sergey's XT or Xi 8088 designs as the base for their system. So you can blame that incompatibility on me :)
These designs use SRAMs, so I decided not to implement DRAM refresh circuit. It slightly improves the performance and also saves a flip-flop. I wasn't concerned that much about making a cycle correct IBM PC clone.
 
sergey said:
I suspect they not only used my BIOS, but also Sergey's XT or Xi 8088 designs as the base for their system. So you can blame that incompatibility on me :)
These designs use SRAMs, so I decided not to implement DRAM refresh circuit. It slightly improves the performance and also saves a flip-flop. I wasn't concerned that much about making a cycle correct IBM PC clone.
Click to expand...

I hope your board is better designed. :) There's some real head-scratchers in this thing.
 
You must log in or register to reply here.
Back
Top