Model 1 Overview
The TRS-80 Model 1 was Radio Shack's first personal computer. It was
developed back in the late 1970's when the only home computers available
to the general public were things like the Altair kits, and (if I remember
correctly) the first Apple machine.
The Model 1 was first made available with a 4 kilobyte tiny BASIC in
ROM, and either 4 or 16 kilobytes of RAM. The CPU board of the computer
was contained in the keyboard unit, and it used a separate black and white
monitor. The monitor was really a RCA black and white TV without the TV
guts in it. A friend of mine had one, and I was almost able to talk her
out of it. (Thanks to Robert Bottomley who corrected me when I remembered
it as a Sony!)
The Level 1 BASIC was very limited. There were 26 number variables
available, A through Z, two string variables (A$ and B$) of up to sixteen(!)
whole characters each, and one array, A(x) that allowed as many elements as you
had RAM for. With a 4K machine, that was right around 875. I managed to find the
box that my Level I BASIC manual was in, and finally looked it up.
Here's something I got in an email from Rob Ricci <
I would like to say that I'm enjoying your TRS-80 page. I have some
info that you may find interesting on Level 1 BASIC. While I didn't own
one, I used to go down to Radio Shack and play with them whenever I
could. This was on a 4K machine.
The highest line number that could be used was 9999. Same for the
RND() [random number generator] function.
Level I had only 3 error messages. WHAT? came whenever a syntax
error occured, HOW? came whenever an illegal function (ex. dividing by
0) occured, and SORRY showed its ugly face when availiable memory ran
Later, the ROM was upgraded to a 12K "Level 2" BASIC, which
was much more advanced. Level 2 BASIC was actually useful to write real
programs in, though all it supported for storage of programs and data was
cassette tape. Level 2 BASIC and 16K of RAM became the "standard"
configuration. If you're interested, there's a Level II BASIC reference
on the web. You can find it at
To get real use out of the Model 1, you needed the Expansion Interface
which would let you go above 16K of RAM to 32K or even 48K. It also had
a provision for using two cassette tape units, a printer port, the floppy
disk controller (which controlled up to four floppy disk drives)
, and a serial port. Finally, it supplied the processor
with a real-time clock interrupt that was used by the various DOS systems
(and a couple of cassette based ROM enhancements) for time keeping functions.
Microsoft had a Level III BASIC that loaded from cassette and made several
useful additions to the standard BASIC.
The main problem with the machine was the interconnection between the
keyboard and the expansion interface. It used a 40 conductor ribbon cable,
and the connectors would get dirty quickly. There was nothing more frustrating
than to have a large program typed in only to have the machine decide it
wanted to reset itself. The other problem was flakey connections at the
serial port board. All these contributed to the name "Trash-80".
The interconnect problem was bad enough to give rise to several companies
that sold gold-plated connectors that could be soldered to the machine.
Another problem that occurred with the expansion interface had to do
with timing and buffering. Some revisions of the CPU unit and some of
the expansion interface required a special cable between the two of them.
I've been lucky so far, in that all my CPUs and EIs have worked together
with just a plain old ribbon cable.
Model 1 Disk Operating Systems
There were quite a few disk operating systems available
for the TRS-80 Model 1. It all stemmed from the fact that the original
TRSDOS from Radio Shack was utter garbage. Several other companies developed
their own (surprisingly compatible) operating systems for the machine,
and in no time at all, the Model 1 was a useful machine. <grin>
There was even a version of CP/M 1.4 available for the Model 1, but since
the ROM, keyboard and video were unable to be remapped out of the low memory
area, special CP/M programs had to be used with it. That sort of defeated the
whole purpose of running CP/M, however. I've still got some old CP/M 1.4 disks
around here somewhere. If I can find them, I'll boot them up and try to find out
who they're from.
In all the different TRSDOS type operating systems, there were several common features
they all shared. Among these were: Disk drives were refered to as :0 to :3,
file names were eight characters with an optional three character extension,
the separator for the extension was a slash (/), and if a file was protected with a
password, it consisted of up to eight characters separated from the filename by the
period character. There are many other common features, but they will not be discussed here.
Below, the major operating systems for the TRS-80 Model 1 are listed, along
with some comments from myself and others.
TRSDOS was the standard disk operating system for the Model 1. When
you bought a disk based machine, that's what you got. Version 2.1 was the
first widely distributed version, and it was seriously buggy. The main
problem it had was it would "lose" files for you. Version 2.2
fixed most of the problems (but not all), and was released a little bit
after two guys in Denver released their own fixes, called NEWDOS. The final
version, 2.3, is not too bad, but it still had a few bugs.
NEWDOS and NEWDOS/80
TRSDOS (before version 2.3) was, to be blunt, horrible. Two fellows
out in Colorado started a company called Apparat, and they fixed almost
all the problems with TRSDOS 2.1 and released NEWDOS. Not only did they
get it out the door with very few bugs, they admitted them when they were
found, and sent out fixes right away. That's quite a contrast from the
attitude Tandy had... It seemed like Tandy's favorite saying at the time
was "not fully implemented". NEWDOS included a disk editor, a
program relocator, and extensions to the Radio Shack editor/assembler package,
making it a very good choice for programmers. The picture above is of
a NEWDOS 2.1 system after displaying the list of library commands.
NEWDOS/80 was their later release. It was another extremely powerful
operating system, which allowed the user to specify just about anything
they wanted to about their disks. It was so powerful, it was more of a
"programmer's only" operating system than one for users. This
is a picture of NEWDOS/80 version 1.0. On the Model III page, there is a
picture of version 2.0
NewDos/80 was probably the most configurable operating system ever devised
by man. You want a custom disk format? No problem... just learn how to
use the PDRIVE utility and you can set up just about anything you want.
The only problem I had with NewDos/80 was remembering what all the switches
did. There were TONS of them. Everywhere. On every command. Agghhhh! Lots
of people swear by it, and it's a good system. I mostly swore *at*
From Shawn Sijnstra (email@example.com),
NewDos/80 v2 was the real one worth getting. The earlier ones didn't
actually do much.
The disk format was VERY configurable although not quite configurable
enough to be compatible with certain formats of other DOSes (mainly double
sided double density being the problem). This was solved with enhancements
written by certain individuals.
NewDos clearly was a programmers DOS since in Australia there were two
separate 'commercial' enhancements for it. One was by Alan Johnston.. called
AJ's DOS. When it got to version 6.0 The Alternate Source started to sell
it for him. It got as far as 6.0A.
Then there was Warwick Sands' package that started out from some enhancements
to BASIC (full screen editor, graphics commands, more machine language
interfacing etc...) but then the DOS got done over. Many features were
added including command line calculator, lots of things that were seen
in other DOSes, a powerful built in shell, and so on. The first version
was called NewDos86 because of when it was released. The 86 caused some
confusion about whether it was an intel version of NewDos so the next release
was called NewDos90. This was released in August 1991.
In case you are wondering... I didn't use a TRS-80 much. I used the
copy. In Australia it was called the System-80. In the states it was the
PMC-80, in Europe the Video Genie and in South Africa the TRZ-80. There
was a colour version planned, designed, beta versions made but it never
Perhaps you could also mention the LNW-80, LNW-Team and Lobo-Max 80
which I think were TRS-80 compatible... started to stretch the memory here....
You just did mention them. (grin) Thanks Shawn - That helps fill this section out a bit.
There's a TRS-80 Clone info page, too, which gives a little more information about the
Model I clones.
If you would like to see it, there is a page of NewDos/80
Commands that gives a rundown of the commands available on both the Model 1 and Model 3
versions of NewDos/80.
Multidos was written by Vernon Hestor, and sold through Cosmopolitan
Electronics. Somewhere around here I've got about half of an older (relatively
older, that is) revision of the manual. If I can ever find the silly thing, I'll see what
I can add.
Vernon recently emailed me, asking what I would like to know about Multidos, so I
should be able to add a bunch of stuff to this part of the page. I am not sure where
to start asking questions, though, since I don't know much about it. I just told him
that whatever he wanted to share about it would be nice. Hopefully I'll be getting
From: Shawn Sijnstra <
Multidos had a reasonable amount of flexibility with disk formats. It
also had a very powerful set of BASIC extensions. Through a utility called
VFU (versatile file utility) it claimed to be able to copy files to and
from any native TRS80 DOS although NewDos could still cause problems. The
most interesting feature of the disk format of MultiDos was that you could
format a double sided disk as one or two volumes. In effect, you could
treat the disk as two separate disks if you wanted to.
From: Albert Hartman <
MultiDos was a TRS-DOS compatible operating system written by a man
named Vernon Hester. It came in both Model I and Model III Versions. The
same disk could be used to make Single and Double Density versions of it.
It had MANY re-written and more advanced utilities and did many things
(such as detecting diskette density and types) automatically. It was as
versatile as NewDos/80 and faster than anything around. The modules of
the OS were written in re-entrant code and In the many phone calls I had
with Vernon I was impressed with his knowledge. I still have my copies
and manuals. In fact, should I ever fix my Model I or spot an LNW-80 at
a show, I'll boot it up yet again and play. Vernon also wrote and sold
a smaller version of MultiDOS called VDOS. This was a tiny version of MultiDOS.
I think this was for licensing to Games and Program vendors.
Finally, I have this bit of info about Vernon and MultiDOS from Johnn Audritsch
I found your page very interesting as I had a ball during the early TRS80
days. A model I, Level II (16k) was my 1st personal computer. I quickly
grew with an Expansion Interface (remember the "spray the connections with
WD40 to fix things" hoax?). I got involved writing a BBS (after
disassembling the Host80 code) and sold a few copies. I also wrote drivers
to send/receive to NC machinery, a driver to run a real-time display for a
multi-monitor, wholesale jeweler and some simple games.
But the best times were related to being a member of the Dearborn TRS80
User Group. Members included Vernon Hester, Dave Welsh (Lazy Writer) and I
believe Kim Watt (sp?) once attended..
I tested versions for Vernon and Dave (I think Vernon thanked me and a ton
of others in the front of one of his manuals). I suggested the Multidos
MOVE command (Cut/Paste a file), some enhancements to VFU and some other
Vernon was amazing. He wrote the fastest Basic Interpreter (Multibasic) -
so fast that at one meeting, he was challenged by a fellow from a
commercial company to "benchmark" Multibasic against the compiler. I
believe it was a large FOR/NEXT loop using an integer. Before the
executable loaded, Multibasic was done. The vendor was a bit embarrassed
and impressed and they tried a few others tests - and Multibasic blew them
Vernon worked as a purchasing agent for Ford Motor Company and taught
himself about computers and Z80 assembler. I was a software programmer,
also working for Ford, and Vernon's code was so elegant, compact and
efficient as to put me to shame.
It's been some years now, but he was a total gentlemen, always willing to
help and so engrossed that his wife often found him asleep at the keyboard.
A company down in Florida, Micro-Systems Software, was responsible for
DOSPLUS. This was another good Model 1 operating system, with quite a few
bells and whistles. The first version I ever used was DOSPLUS 3.5, and
I was immediately impressed. Like NEWDOS, there were quite a few embellishments
for programmer types, but there were also a lot of things that made it
easier for the user, too. A program called MEDIC (or was it CODIR?) was available
to give a nice menu of all the programs on a disk, and let the user pick what to
run with the arrow keys. Enhancements to the Disk BASIC included an INPUT@
function, an array sorting utility, and several others.
Much like LDOS (discussed later), DOSPLUS supported the idea of device
independence. The screen, keyboard and printer devices could all be redirected
to each other, or to a file. This was generally a "Good Thing".
DOSPLUS supported hard disks, though I never used one with it.
Sorry, but there's no picture of this one.
Randy Cook, the author of TRSDOS 2.1, released his own TRS-80 operating
system, called VTOS 3.0. It fixed a lot of bugs in TRSDOS 2.1, but it also
had its own share of them. Not many people took a liking to it, even though
it had a ton of bells and whistles, mainly because it was "copy protected".
Sector 4 of track 0 was unformatted on system disks, and it wouldn't run
Tim Mann sent me some amazing email not too long ago. He got into the TRS-80
operating system business (he worked for Logical Systems) by buying a copy of VTOS 3.0, disassembling it, and
fixing some bugs. While he was at it, he learned (just for fun) how to disable
the copy protection. In addition to that, he found that there was code on
the disk that you could activate to turn the disk into a reproduction master!
Once you booted the modified disk, all attached floppy drives would
start scanning for new blank floppies. Whenever you inserted one, it
would be formatted and made into a copy-protected master disk that could
in turn make non-copyable system disks! Talk about defeating copy protection!
At the time, Randy was distributing VTOS 4.0 through PowerSoft. He
was later contacted by Lobo Drives (makers of the LX-80 Expansion Interface clone and
later the MX-80 computer) to get an operating system for the LX-80,
which wasn't fully compatible with Radio Shack hardware. Unfortunately for Powersoft,
he sold VTOS 4.0 to Lobo. Apparently, from what I've heard, he was a little slow getting
bug fixes to them, though, and refused to deliver any of the source code to
the operating system, so they brought in some other people (including Bill Schroeder of Galactic
Software and Roy Soltoff of Misosys) and started working on it themselves. Eventually, they
turned VTOS into LDOS, or "Lobo Disk Operating System".
LDOS, from Logical Systems, Inc., was the result of Bill and Roy realizing that
they had a good operating system for non-Lobo systems. They formed Logical Systems
(named partly because they needed something that started with the letter L) after
Lobo became less involved.
LDOS was was another of the extremely powerful TRS-80 operating systems (it seems
like they *all* were, other than TRSDOS), which used the concept of device
independence in almost everything it did. The printer could be routed to
a disk file, the keyboard and display could be linked to the comm port,
and filters could be applied to the keyboard to allow one key to generate
dozens of keystrokes. LDOS was probably the best all-around operating system
available for the TRS-80 line. In fact, it was chosen by Tandy to be the
operating system for Model 1 and Model III hard disk systems. This led to LSI's
eventual development of TRSDOS 6 for the Model 4.
Logical Systems tried (and succeeded) in making all their operating
systems compatible with each other as far as disk formats go. A Model 1
LDOS disk can be read in Model 3 LDOS or TRSDOS 6 with no problems. This
made it my favorite operating system, though the others have their strong
LDOS was the last operating system still supported for the Model 1.
The final version of it, 5.3.1, added support for dates past the year 2000.
If you're interested, here is a page full of LDOS
Internals. This has lots of good programming information for both the
Model 1 and Model 3 versions of LDOS. If you'd rather have the full LDOS 5.x
manual in Microsoft(TM) Word (probably another (TM) goes here) or ASCII format,
you should head over to
Tim Mann's TRS-80 page. He also has disk images of LDOS 5.3.1 that can be used
with the various Model 1 and III emulators that are available.
Other hardware and stuff
Double Density Floppies
In addition to the standard Western Digital FD1771 based disk controller,
which only supported single-density, there were a number of different double-density
kits made available. The third-party ones worked rather well. The Radio
Shack one also worked, but was incompatible with most of the others. (What
else would you expect!) I never had the "pleasure" of running
the double-density TRSDOS that Radio Shack came out with, mainly because
I used the wrong Double Density controller, but from what I hear, the
other operating systems were better, anyway. Pictured above is one of the
double density adapters available, the Percom Doubler II. I used the one
from Aerocomp, but don't have a picture of it scanned in yet.
Exatron Stringy Floppy
Another nice addition was the Exatron Stringy Floppy. These were small
tapes that recorded at high speed, making a non-disk TRS-80 into a much
easier to use system. I really need more information about the ESF, though,
because I've never actually seen one "live" as it were.
The Omikron Mapper, delivered starting 1979, is a circuit board with some
components and a bootrom which is plugged into the Z80 socket. The Mapper
displays a boot menu. When the option "CP/M" is chosen, the TRS-80 memory is
remapped so that RAM starts at address 0000 and keyboard/video memory is in
upper memory. This enables the TRS-80 to run standard CP/M 2.2, which was
supplied with the Mapper.
The Micro Merlin (by Micro Projects Engineering 1983-1984) add-on enabled
the TRS-80 Model I to run MS-DOS. The Micro Merlin consisted of a metal box
containing a board with 8088 processor, 256 kB memory, printer port and
serial port. The board of the Micro Merlin was connected to the TRS-80 40-pin
expension connector or to the TRS-80 standard expansion interface. The
TRS-80 is booted with a TRS-80 compatible Dos (e.g. TRSDOS). A TRSDOS
program is then started, which makes the TRS-80 a slave to the Micro Merlin.
The TRS-80 does all keyboard, screen and disk I/O for the Micro Merlin
board. Now MS-Dos can be booted. The Micro Merlin was supplied with MS-DOS
Speaking of hardware... One of the best books I ever read for the Model 1 was called The
Custom TRS-80 and Other Mysteries, written by Dennis
Bathory-Kitsz, and published by IJG, Inc. There were plans for a music
synthesizer, a flashing-light front panel, and all sorts of modifications
that could be made to the machine to make it run faster and better. There
were also a number of other books in the "..and Other Mysteries"
series, and they were all good. If you're interested, Dennis is on the
World Wide Web, himself. Check out his home page at
(Special COCO lovers note: He apparently still has some copies of "Learning
the 6809" that he'll let go for ten bucks apiece. Email him for more
Back to the Model 1, more hardware information, such as important memory
and port addresses, can be found on the Model 1
Internals Page, which might still be under construction.
The TRS-80 Home Page created and maintained by Pete Cervasio
Copyright © 1998 Pete Cervasio