• Please review our updated Terms and Rules here

Morse-Baudot

new_castle_j

Experienced Member
Joined
Dec 3, 2011
Messages
318
Location
Texas, USA
Does anyone else have one of these Morse-Baudot S-100 boards by Curtis Electro? It looks like basically a fancy UART with onboard ROMs to presumably handle ASCII to Morse conversion. I'm interested to learn how it operates, would it take an audio input from a HAM radio via the RCA jack and decode the dots and dashes? There's also an empty DIP socket in the upper left, perhaps a connection to some other peripheral. I don't know HAM radio, so wondering if someone else can take a more educated guess.

The big chip on the board is an AY-5-1013a

IMG_2511.jpg
 
Not sure at this stage in computing any one was good at converting received morse to dots and dashes on screen. The trouble is that certaining when this card was built a lot of morse was hand sent so there is a lot of variability in for example the lengths of the dots and dashs, which computer programs find hard to work with. Do I have that right? There are 8k of code on there? I think you need a manual
 
Looking at the back of the board, the empty socket may have had an op-amp. It looks to have unusual power pins. I suspect they are +/- 15V.
Dwight
 
Using an IMSAI with the Cromemco 8K bytesaver (I knew it would come in handy some day) I was able to dump the contents of the ROMs into a HEX file capture on my laptop terminal and from there convert it to a 7K binary. The address switches on the Morse-Baudot board were set to A0, presumably that's where the ROM code would start. It should be Intel 8080 code.

It is a curious question how this thing would work if decoding human generated Morse code, perhaps it was only meant to receive from another Morse-Baudot board, that way one could expect a fixed rate of transmission.View attachment ROMs.txt
 
I was able to dump the contents of the ROMs into a HEX file capture on my laptop terminal and from there convert it to a 7K binary. View attachment 55721

Just a question about the file as the whole Rom thing worries me in terms of data retention and I'm fairly new to programming and using Roms. In that file, towards the end, there are some broad zones where a number of the byte values have gone to (or were originally FF). Is that the sort of appearance one gets in a Rom file when there is Rom rot ? Or would it look different ? I noticed that there were no light occlusive covers over the Rom windows, so they might have been more at risk of it ?
 
Sometimes (and you find this in a lot of early code), routines need to be at fixed locations, so the intervening unused spaces simply aren't programmed.
 
Sometimes (and you find this in a lot of early code), routines need to be at fixed locations, so the intervening unused spaces simply aren't programmed.

So you are saying that the sort of appearance seen there in that file with the zones of FF's you would judge as probably original programming ?
 
According to the attached article, the ROM software on this board was developed on an ALTAIR with a Tarbell cassette system, an ASR-33 Teletype and some SOL-20 boards and utilities. Oh, and a Cromemco Bytesaver! Examining the HEX, there's some oddball sections like this one:

RAMjunk.jpg

Stuff like that is likely garbage that was left over in RAM when the code was burned into ROM. There's probably a bunch of other garbage scattered throughout the ROM image as well. After reading the manual on how ROMs were written from an Altair front panel and a Bytesaver board, it makes perfect sense. You would get your code running in RAM, then copy that RAM into the ROM, it would include all the artifacts that were left in RAM at the time.

curtisElectro.jpg
 
There were also editor/assemblers that were completely RAM resident (e.g. Processor Technology) along with the source files. I used one on my Altair. Worked pretty well for what it was. That's likely where the source code comes from--the editor was one that used a BASIC-type line numbering system.
 
So you are saying that the sort of appearance seen there in that file with the zones of FF's you would judge as probably original programming ?

Some would intentionally fill with 0's. Others would leave these areas blank. Some would even fill with whatever garbage was in the unspecified RAM at that time. I know that the programmer for Heathkit was religious as he'd put something related in the empty locations. Most programs that used a checksum byte at the end of a ROM image would pad with 0's. My point was that erased by age would float to 0FFh rather than 00.
Dwight
 
The only way to really be certain is to dump the PROMs and disassemble them to see if the code makes any sort of sense.

My money's on the 0ffh areas simply being unprogrammed. I imagine that the PROM programmer source was in the form of Intel .HEX files, which need not present contiguous data--there can be gaps and it's considered to be perfectly normal. A smart EPROM programmer would ensure that unused areas remained ff's.
 
I have seen this board listed on eBay during the summer (several months).

From the 1970s to mid-1980s there were a variety of Morse/Bardot products (discrete & IC designs),
before the usage of Digital Signal Processors (DSP) in late 1980s (Texas Instruments & Motorola chips).

Since the FCC did not permit the usage of ASCII during the 1970s, the majority of the products
decoded Continuous Wave (CW)/Morse Code and 5-bit Baudot / Radio Teletype (RTTY). This changed by 1984 (AT&T Break-up).

The more advanced amateur radio products used keyboard (computer’s or dedicated)
to encode for transmission (Audio Output & Push-To-Talk)
==
Curtis Electro Devices was founded by Jack Curtis, K6KU (sk, 2018 ).
Jack designed “purpose built” ICs at Signetics (his 8044 series) in 1970s.
This specific board was very expensive, and not often seen.
http://www.arrl.org/news/curtis-keyer-chip-developer-jack-curtis-k6ku-sk

The RCA Jack should connect to Audio Output of a Shortwave Receiver,
THEN tune the radio to a CW (Morse) or Bardot (RTTY) transmission.

greg
w9gb
==
In mid-1960s Bill Henry (Illinois BSEE) founded HAL Communications in Urbana, IL (before A. Clarke/S. Kubrick “2001”)
and had a number of dedicated RTTY and CW products throughout same period (1970-1990s).
Bill retired in 2012 ... but is still with us.
http://www.halcomm.com/about/history/
 
Last edited:
I'm interested to learn how it operates, would it take an audio input from a HAM radio via the RCA jack and decode the dots and dashes? I don't know HAM radio, so wondering if someone else can take a more educated guess.View attachment 55678
As G4UGM noted, no system for Decoding CW was fool-proof.
Even when the other station was keyboard/computer sending CW (Morse), the QRN (atmospheric static) and QRM (interference from SW broadcast or other amateur radio stations) would result in gaps or “gibberish”.
An appreciation & respect for the human ear/brain by Morse operators (military in WW2) increased in message reception.

Today, amateur radio using dedicated DSP or the DSP chips within computer sound cards are using a variety of encoding methods (Protocols). JT65 was developed by Joe Taylor, K1JT (Princeton Physics)
https://phy.princeton.edu/people/joseph-taylor
for Earth-Moon-Earth (EME) communications — bouncing radio signal off moon’s surface.
https://physics.princeton.edu/pulsar/K1JT/wsjtx.html
 
Wonderful, thank you for all the responses. Any guesses as to what the empty socket in the upper left of the board is for? The two chips to the left of the empty socket are ILD74, ILQ74 opto couplers. I'm guessing that this must be some kind of interface to transmit morse via a Ham radio. Perhaps it hooks up to a morse paddle key? The advertisement for this device also said that it can produce either TTY or CRT output, could that socket be some kind of hookup to a teletype?

IMG_4176.jpg IMG_4177.jpg IMG_3.jpg
 
I think your right. It is likely a connection to something like a teletype. It would be worth tracing the leads out.
I notice that the additional panel has two ribbon connectors on it. It also has lots of signal filters.
Dwight
 
A little bit of progress, thanks to Jim Battle's SOL-20 emulator. I can load the ROM image from the Baudot board into the emulator and execute it. The ROM code relies on the SOL personality module (SOLOS) to provide the I/O routines. Starting the program is as easy as jumping to A000 and executing.

When I execute it in the emulator, the copyright message displays, and under that "TTY" and a few seconds later the cursor starts flying back and forth and the screen begins scrolling. At any time I can type something in at the keyboard and it will pause and the letters I typed show up on the screen (only lower case, SHIFT key doesn't cause upper case letters on the screen), then it will go back to scrolling a couple seconds after I stop typing. I am guessing that this is the expected behavior, it seem like it just begins polling the Baudot board for an incoming stream of characters that would get re-directed to the screen. Perhaps typing in some characters interrupts the receive function and causes the board to transmit what was just typed. The ROM image didn't contain any ASCII strings in it other than the copyright message, no interactive menus or anything. I think the operation of this board is just that simple, plug it into your SOL, attach a Ham radio's audio out to the RCA jack, jump to A000, tune in a morse code transmission, and wait for some characters to display.

Does anybody out there have a SOL-20 with SOLOS personality module, 40K of memory, a VDM-1 display board and a Ham radio? Would be neat to see if this thing works, PM me if you're interested in trying and making a video.
 
A little more progress... maybe. I was able to port the ROM on the Morse-Baudot board over to a COM file that is executable under CP/M. I could see the ROM code calling the SOLOS output routine while it was running in emulation and I patched in a routine to call CP/M's CONOUT instead. I tested in on real hardware and it runs, the copyright message displays when executed. Tonight I hooked up the Morse board to the audio out on my laptop, there's a website that you can type in a message and it will output a midi file of the morse equivalent. I fed that audio into the Morse board's RCA jack while the program running under CP/M and...... nothing happened. It was a long shot anyway.
 
Back
Top