• Please review our updated Terms and Rules here

My 1st-hand experience on a 2nd-hand 8088

D

Dca2

Member
Joined
Oct 5, 2011
Messages
40
Location
Canada
Well, it's garage sale season again in my area. The beauty of this season is that I can never plan how I'd be surprised next. As a 386 guy, my wish-list was full of 386 stuff but nothing earlier than that. Up until recently, I occasionally came across this guy (as pictured). It was a 8088 based XT clone, Acer-710. Although I was not really expecting to extend my retro hobby that far backwards, I decided to take the given chance and explore it for fun.

A710_1.jpg

What I found under the cover:
- CPU 8088-1 10Mhz (or 4.77Mhz in non-turo mode)
- 768kB 120ns memory in massive array of DIP chips
- Integrated CGA/MDA/MGA video
- Two 5.25" 360k floppy drives
- Minicribe 8438 31M hard disk
- Adaptec ACB-2072 RLL controller card
- One parallel port and one serial port

The manufacturer spec is available here:
http://classictech.files.wordpress.com/2010/03/1987-acer-presskit-acer-technologies-targets-comdex-to-make-pc-industry-debut-6-1-87.pdf

and the mobo documentation here:
http://stason.org/TULARC/pc/motherboards/A/ACER-INC-8088-ACER-710.html#.UZ42S4JdX10

A710_6.jpg

It might not be an ideal candidate for any ultimate retro project due to its less-customizable integrated mobo. But all I wanted to do was to make it working, so that I could revive my old school BASIC/Fortran/C codes on a true 8088 PC. The machine came without any peripherals -- no monitor, no keyboard and no mouse, which yielded a bit of challenges in my home lab with merely AT/ATX stuff. The first thing I wanted to avoid was to hunt for a space-taking 9-pin monitor, although I did miss the CGA look & feel for a long time. Instead, I quickly dropped in a generic Trident 8900C VGA card and fired it up with a 17" VGA CRT. The good news was that the machine beeped and passed the memory test, the bad news was that the HD didn't spin up. I was also a little bit disappointed about the silence during the memory test (in either fast or slow test mode). I loved to hear the tick sound as a kind of feedback from counting those very expensive silicon bytes at that era.

A710_2.jpg

A710_3.jpg

After several power on/off, the HD finally spun up loudly (compares to those early IDE HDs). I then discovered that I have to power it on everyday or otherwise it would randomly choose a day as its holiday. Another interesting thing was that the RLL controller card refused to work with all my 386DX motherboards but worked well at the first kiss with my PIII motherboard. It was fancy (for a newbie) to see such a cross-century marriage between RLL HD and PIII 1GHz Coppermine. There could be something to do with the BIOS I thought. But anyways, this allowed me to easily dump MS-DOS 3.30 and many other software to the HD (without via the 5.25" 360 disk). I've also managed to install Windows 3.0 in VGA mode with an effective help from this thread:

http://www.vintage-computer.com/vcforum/showthread.php?35593-Windows-3-0-VGA-color-driver-for-8088-XT

The true trouble started with the keyboard. The oldest keyboard I could find didn't work, although it physically had a correct 5-pin DIN head. I've then recalled a "common sense" that AT keyboards are not XT compatible. The issue was thoroughly discussed in this thread:

http://www.vintage-computer.com/vcforum/showthread.php?15907-AT-to-XT-Keyboard-Converter

So, before I decided to follow the guide and build the convertor with a PIC chip, I asked myself why not just open the keyboard and speculate in? I was happy to see that the controller IC in it was a well documented IC. After I desoldered the jumper wire from the pin #1, the keyboard could be *sometimes* recognized by the XT box, which was a good sign. I then soldered a 1k resister from pin #1 to +5v, which reliably locked the keyboard in XT mode. I was lucky enough because the IC in this keyboard had both XT and AT logic in its ROM and was switchable by pin #1.

A710_4.jpg
 
The next trouble should not be existing at all if I could resist the temptation of adding a sound card to it. I knew that SB 1.5/2.0/Pro were perfect for XT, but I didn't have one in hand and I also felt responsible for finding good home for my spare SB16s. So, I picked up a ViBRA16X (CT4170) card because I saw that all data/address pins were absent from the 16-bit portion of its ISA interface (which implies a desired capability of passing 16-bit stereo sound over a 8-bit ISA bus). It's a PnP card though, but PnP should not be an issue because there was a tool called CTCM.exe, right? … wrong! Every time I tried with CTCM.exe, the computer hung forever. The issue was thoroughly discussed in this thread:

http://www.vintage-computer.com/vcforum/showthread.php?7987-SB-16-on-XT

So, CTCM.exe and a few other Creative Labs utilities require 80186+ CPU to run! It almost sounded like the end of world. But wait, theoretically, SB PnP card should be ISA PnP standard compliant in general. The ISA PnP Spec is well documented here:

http://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&ved=0CDEQFjAB&url=http%3A%2F%2Fwww.osdever.net%2Fdocuments%2FPNP-ISA-v1.0a.pdf&ei=_tqbUfGJFaGDywGEj4C4DA&usg=AFQjCNGFUTbBQlH23YgxrqJCP5yvFjhjNA&sig2=-0DaLjDnCQfQNvbALeVy9w&bvm=bv.46751780,d.aWM

Fortunately, I happened to have ISA PnP general programming experience and still had my old Turbo C source code in my legacy repository. It took me a few hours to customize the code for my CT4170 and eventually brought it to work on my Acer-710 without eating up any memory (non-TSR). I've merely tested it with Wolfenstein (re-engineered for 8086 by Mike Chambers) and LENNI, the MIDI and WAVE worked as expected.

A710_5.jpg

If someone else here is also bothered by the same issue, I uploaded the utility here for test.
View attachment CT4170.zip
It's better to invoke it in AUTOEXEC.BAT and position it AFTER the 'BLASTER' environment var. In my case, the IRQ 5 was already taken by the RLL card, I had to set my 'BLASTER' string as following:
set BLASTER=A220 I2 D1 H1 P330 T2
The utility should also work on AT systems with 8/16-bit ISA bus but may not be necessary since the Creative's original CTCM.EXE is available on AT.

There is one trouble unsolved and still makes me sick today. The machine is not able to remember date/time. Of course, I replaced the dead button battery with a brand new one (no leak was found). The strange thing is that the mobo can't remember date/time even after a soft-reboot, which is not a common behavior on an AT mobo with a dead battery.

A710_7.jpg
 
Last edited:
Most real time clock solutions on XT machines require special software to load and save the date. This could possibly be why you're having the clock problem.

Nice looking machine :)
 
SpidersWeb said:
Most real time clock solutions on XT machines require special software to load and save the date. This could possibly be why you're having the clock problem.

Nice looking machine :)
Click to expand...

Oops, if this is the nature of XT, perhaps, this is also why DOS asks for manual input of date/time if config.sys and auto exec.bat are absenct.
 
Those MiniScribe 3.5" hard drives all suffer from stiction. To get it to go again, just rotate the little thing that has warning labels next to it saying "Do not rotate interrupter". That will unstick the heads from the platters. Now you can apply power and it should spin up fine.
 
I think SpidersWeb has it right. It almost certainly needs software to set the clock chip, and then to read the chip and set the DOS clock on startup. You might check out this page.

http://www.minuszerodegrees.net/rtc.htm

I didn't see Acer listed there, but if you notice near the bottom of the page is the Strickland real time clock programs. Included in the zip file is a doc explaining how they work. I have found these useful when I didn't have the original manufacturers software.

I too am curious about the keyboard chip. Can't quite make it out in the photo.
 
Last edited:
Haha! I thought it only happens to my miniScribe :) yea- rotating the interruptor always helps.

vwestlife said:
Those MiniScribe 3.5" hard drives all suffer from stiction. To get it to go again, just rotate the little thing that has warning labels next to it saying "Do not rotate interrupter". That will unstick the heads from the platters. Now you can apply power and it should spin up fine.
Click to expand...
 
Dca2 said:
Oops, if this is the nature of XT, perhaps, this is also why DOS asks for manual input of date/time if config.sys and auto exec.bat are absenct.
Click to expand...

That's pretty normal for MS DOS on any machine :) But hopefully someone around has the Acer disk or knows more than I.
 
vwestlife said:
Those MiniScribe 3.5" hard drives all suffer from stiction. To get it to go again, just rotate the little thing that has warning labels next to it saying "Do not rotate interrupter". That will unstick the heads from the platters. Now you can apply power and it should spin up fine.
Click to expand...

This tip works immediately with my first try! According to the previous experience, I may need to repeatedly play this trick every (x-) days.
 
fatwizard said:
You might check out this page.

http://www.minuszerodegrees.net/rtc.htm
Click to expand...

I quickly downloaded and tried the Strickland real time clock programs. I put the CLKDRV.sys in config.sys and rebooted. It said no RTC detected. But I will do more experiment tonight with other programs provided in the package or on the same web page. I believe this is the right direction to go. The document of the motherboard also shows that there is a switch determining whether the on-board RTC is enabled.
 
SpidersWeb said:
That's pretty normal for MS DOS on any machine :) But hopefully someone around has the Acer disk or knows more than I.
Click to expand...

Yeah, while this "feature" seems useless on AT+ PCs, that makes me curious about its original design rational. :D

fatwlzard's post opens a new hope, I will experiment and report back soon.
 
fatwizard said:
I too am curious about the keyboard chip. Can't quite make it out in the photo.
Click to expand...


My knowledge about keyboard circuit is very limited. My assumption (instead of understudying) is that only those early AT keyboards actually have a hidden ROM "partition" with XT protocol built-in inside the controller chip. Unfortunately, there is no universal way to detect or activate the hidden XT "partition" across different chips or with different PCBs. I think that the photo in my post just demonstrates an individual case :blush:
 
Last edited:
Good news, the RTC problem has been solved with the following utility:
TIMER.COM version 2.x

After failed with Strickland, I've tried with PC Timer 2.x. First, I used TIMER /I to initialize the date/time, and then put TIMER /S in autoexec.bat. The system clock works after both soft-reboot and cold-reboot (as expected).

Thanks to SpidersWeb for pulling me out of the trap of fighting with battery, and thanks to fatwizard for pointing me to the correct resource (including the PC Timer utility).
 
Last edited:
A while back i pulled an old Cherry AT keyboard apart to give it a good clean and noticed on the circuit board it had a jumper and just by the side was marked "SW" so after giving it all a good clean i cut the jumper and wired in a small toggle switch, It worked fine on my XT and AT, I seem to remember reading somewhere that at the time some manufactures used 1 rom file which contained the XT and AT code and wired the circuit boards accordingly.

Dca2 said:
My knowledge about keyboard circuit is very limited. My assumption (instead of understudying) is that only those early AT keyboards actually have a hidden ROM "partition" with XT protocol built-in inside the controller chip. Unfortunately, there is no universal way to detect or activate the hidden XT "partition" across different chips or with different PCBs. I think that the photo in my post just demonstrates an individual case :blush:
Click to expand...
 
My mistake was that I just cut the jumper and test. The result was unstable -- less than half of the time the keyboard could work in XT mode. Perhaps, the reason was that the pin was left suspended and thus act as a high impedance antenna responding to random noises from the air. A 1k resistor resolved the problem by persistently pulling the pin up to +5V. A toggle switch is certainly a better solution for ease of switch between AT and XT mode.


Malc said:
A while back i pulled an old Cherry AT keyboard apart to give it a good clean and noticed on the circuit board it had a jumper and just by the side was marked "SW" so after giving it all a good clean i cut the jumper and wired in a small toggle switch, It worked fine on my XT and AT.
Click to expand...
 
Hi!
Sorry for my english, please.
I have ACER 710. It works, but when it turned-off, I listen some sound in power supply.
Inside power supply I see, that somedoby try to fix it before. It's terribly.
Some wire on fuse, obscure resistor on one leg and... What iz TNZ 101? I dont know, what is TNZ :( May be I can use some analogs?
635bfe837ceb.jpgc8c7d7e62b03.jpg
Thank's!
 
You must log in or register to reply here.
Back
Top