TRS-80 Model I info (RE: MSX, TRS-80, Colour Genie, etc.)

From: Allison J Parent <allisonp_at_world.std.com>
Date: Wed Jun 25 22:14:23 1997

> The Model I originally shipped without a numeric keypad. To the right

Correct.

> of the main keyboard was a rectangular keypad-size plaque reading "Radio
> Shack TRS-80 Micro Computer System". The numeric keypad was added to
> later models, and was available as a retrofit kit for around $50. With
> the numeric keypad installed, the nameplate was moved to a horizontal
> plaque above the keyboard.

Also correct.

>
> The TRS-80 Model I lineage includes:
>
> Model I, 4K, Level I BASIC
> - This is a 3-piece system with the computer in the keyboard. It
> includes the system keyboard/cpu, monitor, tape drive (actually a
> rebadged regular Radio Shack portable cassette deck with no
> modifications), and power supply brick. Level I BASIC is similar to
> Tiny BASIC. I still have my Level I BASIC reference manual.

The Basic was an 4k microsoft basic with floating point and simple arrays
but no alphanumeric operators or transcendental functions.
  
Tiny basic was an integer language of less than 4k.


> Model I, 16K, Level II BASIC
> - The 16K and Level II upgrades went together. 16K is the maximum Model
> I memory in the system unit (8x 4116 DRAMs). Level II BASIC is similar

wrong. Either could be installed alone. Generally LII with 4k was pretty
cramped. FYI: the LII romset was only 12k.

> to Microsoft BASIC/80 with functions added for things like the TRS-80's
> 128x48 memory-mapped monochrome graphics.

It was MS12k basic with TRS extensions (graphics).

>Level II also added a
> keyboard debounce routine--Level I machines were very difficult for
> typists.

The key bounce was a bug in the original 4k software, it didn't wait long
enough. There was a cassette that when loaded fixed it. Me I'd clean the
key contacts with a swab and some contact cleaner and get the same result.

> - Expansion Interface
> Matching silver color, acts as a monitor stand, connects to system unit
> via ribbon cable. Contains dual floppy controller (WD chip), sockets
> for an additional 32K (2 banks of 4116 DRAMs) for a system maximum of
> 48K, and a parallel connection. 16K ROM BASIC occupied the remainder of
> the address space. The expansion interface also contains a card bay for
> an RS-232 interface.

The bottom 16k was 12k of rom (LII basic) 1k of ram for video and keyboard
mapped in to memory space. Some of the 4k space for the video and keyboard
was wasted due to partial decode. The upside was since the keyboard was
scanned by the cpu so alternate shift and character sets were easy to do.
The down side is no matter how you tried, keyboard type ahead was
impossible, the keyboard could not interrupt the CPU.
 
> - RS-232 interface board
> For expansion interface.

The surface connector used was very cranky.
 
> - Floppy drives
> Single-sided single-density, approx. 90KB free space.

The design used the 1771 internal data seperator which was not very tolerent
of drive spped errors or data jitter.

> - Lower case upgrade
> Provides lower case capability.

Way late in the game the "field mod" had been around over a year before
tandy did it.
 
> - Numeric keypad retrofit
> As discussed above

Popular item!
 
> Known TRS-80 Model I problems:
>
> - Unreliable cassette interface. Radio Shack later released a
> modification that improved this somewhat. The best option is a
> third-party unit called the Data Dubber by Microperipheral Corporation
> (I worked there!) that went in between the system unit and cassette and
> squared the wave.

There were two mods one largely marginal, the later one was very effective.
I had a mod I did that worked very well and was far simpler.

> - Wonky, unbuffered connection to Expansion Interface. This went
> through various modifications, and some cables you'll see have big
> buffer boxes in the middle. Later Expansion Interfaces had built-in
> buffering. Some bought third-party expansion interface clones from Lobo
> and others. Be very careful if you get an Expansion Interface without a
> cable. It might need the buffered cable, and it would be a pain to
> manufacture.

The first version with the unbuffered or buffered cable was a junk design.
the later one with local ras/cas timing was far better.

> - Bad data separator chip. The stock data separator was unreliable.
> Most people replaced theirs with a third-party improvement such as
> Percom's.

The stock circuit depended on the 1771 chips internal seperator, Even WD
the chipmaker said don't do it!

> - Unreliable connection for the Expansion Interface-mounted RS-232
> board. This board slipped over vertical post connections and never made
> good contact. Most folks used third-party alternatives that worked off
> the cassette port.

Being there at the begining was half the fun.

Allison
Received on Wed Jun 25 1997 - 22:14:23 BST