Infinite-Mantra: COMM Project

[Infinite-Mantra]

Howdy howdy, y'all! I saw there were some questions about the project I had in mind, which was mentioned in the Intro thread. I figured it'd be good to just centralize that discussion here!

So, in a gist, I want to set up a computer which handles communication via AM radio signals. I was hoping to set up a system which receives the radio signals, converts the signals to binary, and then converts that binary to text. The system would present these received text messages onto a screen in a kind of "message board" layout. And whenever the user sends off their own messages, the text would be converted appropriately: text>binary>radio signal transmission.
That's a lot of data and computation to manage, so I looked into it and it seems that you can run two 65c02 microprocessors together; I figured I'd do that, with one microprocessor handling the radio reception/transmission and conversion, and with the other microprocessor handling the message board system and UI.

I expect this is an ambitious project, you know? But I think, with the proper modules and know-how and time invested, I can get this working. It's a tall order though, so I think it might be best to work out a lot of my first steps with folk here - folk who know these systems/projects better!

[GARTHWILSON]

A UART (or ACIA) could be part of converting the binary to and from the radio signals, with no need for a second processor.  The UART's serial side could go to a simple modem that does something like the Kansas City standard, if indeed you want AM.  Otherwise, perhaps it'd just turn the carrier on and off, rather than modulating it.  The binary<-->text conversion is kind of a matter of the display, not something separate the processor does.  For example if you want to display 'A', you output $41 to the display.  For input, if it's from a keyboard, the greater job is keyboard scanning and debouncing, and converting a keycode into a an ASCII code is a lesser job.  Do go through the 6502 primer.

[BigDumbDinosaur]

I agree with Garth: this is doable with a single 65C02 and the appropriate support hardware.  The WDC 65C02 can be clocked fast enough to give it ample time to tend to everything you are planning.  I foresee some interrupt programming in your future. 

[Infinite-Mantra]

Oooh, thank y'all so much! It seems a lot of the steps I had in mind would've made this so much harder than it needed to be. I'll look into the UART then! And yeah, Kansas City Standard is exactly what I'd want! I did some reading on that system and it seems to be just what the doctor ordered.
Okay: so I'll take a look into keyboard scanning, and debouncing seems like a super-useful boon to add in.

Y'all have been so helpful and so welcoming, and I really appreciate it! It can be overwhelming, you know?

[BigDumbDinosaur]

GARTHWILSON wrote:

The UART's serial side could go to a simple modem that does something like the Kansas City standard, if indeed you want AM.  Otherwise, perhaps it'd just turn the carrier on and off, rather than modulating it...

The latter idea, which is a form of continuous-wave (CW) transmission, is one way to approach it.  Carrier off is a cleared bit, carrier on is a set bit.  The data flow thus generated would need some error checking, but that’s not at all difficult.  There are some technical problems with using basic CW, but it could be done.

If Kansas City modulation is used, you now have to pay attention to the effects of carrier phase shift, possible reflections (aka “ghosting”) and oscillator drift, especially at the receiver end, assuming it is using a superhet stage between the RF and audio sections.  The Kansas City standard was not really designed to be sent over a wireless medium.  With CW none of that is an issue and a very simple receiver design is possible, as all it has to do is detect whether carrier is present or absent.

Another technique is to superimpose FSCK on top of CW, which can be done by feeding the audio output of a standard modem into the transmitter, and using a similar modem at the receiver end to reverse the process.  Not knowing the frequency band in which this is going to be transmitted and received over the air, I can’t hazard a guess on how fast it could reliably run.  However, 1200 bpS seems practical using error-correcting modems—it would run faster than most people can read.

When I was in the Navy, Tele-Types handled much communication ship-to-shore, e.g., ship’s orders, and the transmitter and receiver involved operated in CW single-sideband mode, with FSCK riding on it.  It proved to be very reliable, even when thousands of miles out at sea.

[Infinite-Mantra]

BigDumbDinosaur wrote:

When I was in the Navy, Tele-Types handled much communication ship-to-shore, e.g., ship’s orders, and the transmitter and receiver involved operated in CW single-sideband mode, with FSCK riding on it.  It proved to be very reliable, even when thousands of miles out at sea.

Hmmm... Okay, it seems I've got a lot of homework to do then! A lot of those terms are entirely new to me, but I do know that distance and reliability are my two big goals with this. So maybe FSCK and CW are something to look into - because that sounds perfect for me!
First, though: the 6502 Primer. Thanks again, Garth!

I hear AM radio signals are typically better for long-distance communication than FM radio, hence that choice. I know it probably sounds silly to rely on radio rather than other, newer means of telecommunication, but... This sort of oldschool tech approach has a really fun challenge baked into it.
Not to ramble, but it's just so exciting getting to chat with the folks here; already, you've offered some invaluable insight!

[GARTHWILSON]

Infinite-Mantra wrote:

I hear AM radio signals are typically better for long-distance communication than FM radio, hence that choice.

You're probably thinking of FM broadcast which is in the band of 88-108MHz, which is VHF, versus AM broadcast which tops out at 1.6MHz, or shortwave broadcast which is HF, all below 30MHz.  VHF is generally line-of-sight, which is probably where your idea of shorter range comes from.  At a given frequency however, FM will generally be more efficient than AM, and single side band (SSB) is more efficient than FM, and CW telegraphy can get through where nothing else can.  I won't address the newer digital methods; but I'm a radio amateur (WB6NUY) but I have not been "radioactive" in decades.  I just keep my license current so I can jump back in at any time.  When I was active though, I was usually on the 80-, 40-, 20-, and 15-meter bands (close to 3.5MHz, 7MHz, 14MHz, and 21MHz) on CW telegraphy, which is where rather than modulating the carrier, all you do is turn it on and off; and I was operating very low power ("QRP"), at 2W, and talked all over the world on that, with a random-length wire strung between the house eves and a tree, using a home-made transmatch to match the wild impedance to the 50Ω that the radio wanted to see.  At the extreme, when I was in Hawaii I contacted someone in the Middle East, which is nearly the opposite side of the earth.  The first night I had that Heathkit HW-8 radio on the air, I worked all the southeastern states from southern California, also Oregon, Washington, and Japan.  When an incoming signal is super weak, you can pull it out of the noise by using a super narrow, sharp band-pass filter.  Depending on how sharp, you may have to slow the data rate way down; but that's basically how it's done when we're listening to signals sent by spacecraft out at the outer edges of the solar system, where the signal might have been transmitted with only a few watts and has to travel hundreds of millions of miles.

To experiment, with others who aren't licensed hams, I've thought of doing basically what we're talking about here on CB radio where there's no enforcement, very little traffic, and nobody cares; just put the modem on an old AM CB radio, and send data at 300 bps.  I made myself a modem with penny parts decades ago to put data on a common cassette recorder.  It worked great, and there's no reason it wouldn't work on radio too.

[BigDumbDinosaur]

Infinite-Mantra wrote:

I hear AM radio signals are typically better for long-distance communication than FM radio, hence that choice.

As Garth notes, the modulation technique is less important than the wavelength, but FM definitely will be the better choice if possible.

With long-wave AM, you need a high-powered transmitter driving a tall antenna to go any significant distance, which is why your typical commercial AM broadcast station runs in the 10,000-50,000 watt range and has a 1000-foot mast antenna out in the country, or a shorter one atop a skyscraper.  More power will get you more range up to a point—as demonstrated by the old Mexican border-blasters, which ran well over 100 KW and could be picked up in Chicago at night time, but operational costs and problems with arcing near the antenna start to become difficult to tame.  Plus the neighbors might not appreciate the crosstalk that will inevitably attack the lower TV channels (due to harmonic distortion). 

On the other hand, short-wave (aka HF) AM or FM, mainly in the 20-15-10 meter bands (15-30 MHz range), can go half-way around the world with modest power by taking advantage of ionospheric reflection, which is reasonably dependable at those frequencies (using a 500 W, SSB transmitter, I once contacted someone via voice near Nome, Alaska in the daytime on the 10-meter band while out at sea near the Azores).  However, you need a properly-loaded antenna to get the maximum possible power transfer from the transmitter to the airwaves.  The “ideal” is either a 35 foot (or longer) vertical whip antenna with a tunable loading unit (expensive but remarkably efficient), a steerable half-dipole antenna matched to the frequency on which you are transmitting (somewhat less expensive) or a wire at one-half wavelength.  However, as Garth notes, your antenna can be quite makeshift and still perform.  As he says, the key is to achieve a proper impedance match.

Quote:

I know it probably sounds silly to rely on radio rather than other, newer means of telecommunication, but... This sort of oldschool tech approach has a really fun challenge baked into it.

Actually, radio links continue to be use in that fashion to communicate with someone who is well away from civilization, but has a local electricity supply to run radio gear.  Also, radio links are what pass digital telemetry between aircraft in flight and FBOs that monitor said aircraft’s performance—a common application in the commercial airline business.

[Martin_H]

Interesting project. As Garth and BDD have said, this is well within the 6502's capabilities. Wireless computer networking was invented in the early 1970s using computers of that era.

Computer networking has something called the OSI model which is a framework to understand and implement network communications between different systems. While over engineered for your project, the OSI model dividing the communication process into distinct layers is a useful abstraction. Specifically, that the radio link layer is a separate challenge from the digital application protocol between the two computers. To me the radio link layer seems like the biggest challenge, but I'm not a ham radio enthusiast. Hams have something called RTTY which could be the basis of a radio link layer.

To me the two biggest questions for that layer are:

* What distance are you're thinking about for the radio link?

* What data rate are you thinking about for the radio link?

[Infinite-Mantra]

Martin_H wrote:

To me the two biggest questions for that layer are:

* What distance are you're thinking about for the radio link?

* What data rate are you thinking about for the radio link?

Howdy howdy, Martin! I'm glad you like the project idea, too! I was worried this was silly, or maybe too ambitious.

So, I know that I want my range to be as vast as I can possibly get it. Ideally, I'd like it to be as long-range as possible, and I hear that that's a trade-off with data rate; I'm willing to go for a lower data rate for this project, to a point.
That point is, I don't want to actually risk data loss, if that makes sense. It doesn't matter how long a message takes to arrive, or if we have to implement a text message length limit. So long as the message arrives, and it's not messed up via data loss, I'm happy.

[GlennSmith]

Hi, you might want to look at the LoRa (RongRange) specification. There are ready-made modules designed to be used with Arduinos et al.
I have an Elector magazine from last summer where they perform some comparisons between several data communication systems & protocols - including LoRa. I'll see if I can find it, I can give you some of the details that way.

[Infinite-Mantra]

Thanks so much, Glenn! Y'all have been such a delight with this!

I still need to read through the starter info Garth provided; once that's done, I feel like it'd be good to start actually nailing down the specifics of what I want (blueprints and hardware). As I understand it, you gotta know the specific hardware you're working with before you can start incorporating it into your 65c02 code. So before I start actually coding, I need to nail down what I need.

[No True Scotsman]

Hi, Mantra!

I've coincidentally been researching various serial and wireless options for my project today. Here are a couple of suggestions.

The ready-made RYLR998 LoRa modules can directly interface with a UART, which suggests to me that you can use any serial protocol you like over them.

For the server side, you could go with an FT232H USB-to-UART or similar module connected to a second LoRa module. You may be able to directly connect the two modules without an Arduino in between, but I haven't researched that idea.

Some years ago, I wrote a PHP library that could connect any device to any other device using some combination of serial ports and TCP sockets. I had a C-64, a CoCo2, and a Model 100 set up as serial terminals on various ARM SBCs (e.g., Raspberry Pi) via FTDI cables, which latter communicated with each other via WiFi, and with Arduino projects via USB cables. In addition to endpoints the software implemented a custom IoT protocol whereby I could (for example) type a command on the CoCo2 to execute an operation on an I2C module connected to an Arduino somewhere else in the house.

I'm planning an extension board for my 65C02 project that can switch a UART between a number of hardware interfaces (RS-485, MIDI, LoRa, WiFi, etc.) under program control.

I also stumbled upon another intriguing package today. Donald Moran's Pico 6502 Bridge interfaces a Raspberry Pi Pico W with the 65C22 VIA to basically put a WiFi adapter directly on the 6502 address bus. It contains the 6502 code and Pico firmware to run it, which also implements a custom UART-to-TCP protocol.

It's a project after my own heart, I tell you.

[GlennSmith]

Just to complete what I posted above, I realized that the article I have is in French.... But here is a link to the same elektor article in english : Navigating Wireless Protocols. It costs a few €/$/£ (I dont know where you're based), but I found the article very useful.
(EDIT: the price of the SX1278 modules is SO cheap, they look like good experimenting candidates. Just check that the frequency they use (often 433MHz) is compatible with the regulations where you are.)

[Infinite-Mantra]

Howdy y'all, sorry I've been away; work has been a real doozy!
I think I want to start actually nailing down the things I want for this project. Get a little shopping list together. So I'm going to look back through this all and see which setup works for me. I want range, and I want to make sure that my transmissions come through without issue.