Old, older and .... what next?

[kellys_eye]

My first sideswiper was a twin paddle 'thing' I purchased when shoreside visiting Southampton and I found an electronics shop that sold me a 'blue box' with red plastic paddles and internals built around the Curtis 8044 chip.  I loved it!  As a merchant marine operator I was perfectly at home with a straight key but suffered, on occasion, from a ganglion on my sending wrist that severely restricted wrist movement (giggidy) so getting hold of a bug key sort of saved my life.

I was surprised at how easy it was to pick up the technique too - only a day playing with it - and it showed my how limited my hand keying really was as I could exceed 45 or even 50wpm all day, very day with the Curtis thing but would break out into a sweat after 10 minutes at 45wpm on a straight key!

Regrettably that key was consigned to a junk box some years after leaving the merchant navy and 'carefully disposed of' by my (ex) wife in clear out..... grrrr....

I've search in vain for a Curtis 8044 chip even though there are many PIC devices that can replicate the functions just as easily.  Typical nostalgia trip for me I suppose.

Anyway, last year I purchased a Heathkit bug key , the HD-1410 (E-bay, sold 'as seen' and not working) that I've nearly finished renovating and it will, I hope, give me that same kick that the Curtis did i.e. does the job, competently, inexpensively and with 'fun' as a big part of it.

Prior to the Heathkit purchase I bought a naked squeeze key (Hi-mound 703) and a small board to knock up a keyer but couldn't 'gel' with it and ended up flogging it back on eBay!

 

[Ham4CW]

I was after an 8044 keyer chip a few years ago. I managed to find one on a keyer board removed from an Icom HF rig (on ebay). By the time I had found one the project that I intended to use it for 'gone off the boil' so to speak, so the chip was never used in the end. Still have it in the cupboard somewhere... :-*

I have an old Heathkit HD-10 to restore one of the days. I purchased it as 'spares or repairs' from a seller in the States. It's a 110V unit as it stands, but I have a 240-110V transformer to drop the UK mains voltage to something more suitable to run the item.

Found that upon powering the keyer up it produces a continuous tone and does not respond to the inbuilt paddles at all. Needs some TLC!

I think I might have an Heathkit HD-1410 in my 'collection', also have the Hi-Mound 703 key. Too many keys,, not enough hours in the day. ;D

73, Mark...

 

[Mickey Morse]

I found an archive of a usenet thread where someone necro'd a many year old post to offer an 8044 for sale, back in 2016:

https://www.radiobanter.com/homebrew/80486-curtis-8044-keyer-chip-2.html

There is also a link to the 8044 application note, which I found interesting. 

These days, it might be possible to build a drop-in (pin compatible) substitute using an smt processor and some wiring on a DIP-format pcb adapter, along these lines:

https://store.open-electronics.org/uChip_Arduino_Zero_compatible_headers_soldered

 

[Ham4CW]

From an electronics point of view, although it would be quite feasible to program such a module to emulate the 'actions' of an 8044, you need to keep in mind that unlike the original Curtis IC, many of the modern PIC and Atmel devices were nevr intended to be used in an environment where high RF levels are present, so in some cases you can get problems with erratic keying and false triggering.

From experience I can say it varies a LOT, with some micro's working without a hitch, but others will completely lock up and cease to work until the offending RF source is removed. It can even vary from band to band, with some frequencies being more problematic than others. Having said that, with some additional decoupling added most circuits can be made to work.

73, Mark...

 

[Mickey Morse]

Oh thanks, that is an interesting point about RF sensitivity.  I'll try to get further advice about this if I try to make a real keyer to key a transmitter with.  Step 1 was going to be to just be for code practice without a radio.  Now my thought is to shield the microprocessor circuit very well (with copper tape and that sort of thing), and use an optoisolator to communicate with the transmitter.  I will also have to find out (not now) what kind of switching is required to operate the transmitter itself.  I think some old transmitters actually send the whole transmitter power through the key.

 

[Ham4CW]

In the past I've found that at VHF/UHF frequencies, much of the circuitry can be large (long) enough to act as an antenna, so the circuit becomes overloaded by direct pick-up of the RF radiating from the transmitter.

As you come down the frequency scale then the antenna type used with any particular transmitter can have quite varying effects. Over the years I've found that a keyer circuit (or other peripheral devices such as speech processors), are not so much bothered by radiated RF from the transmitter, but more so by earth/ground currents.

So say a balanced antenna, like a dipole, with a current transformer at the feed-point and some choking on the feeder itself, will work quite happily even at quite high RF levels. But the same transceiver, used with an unbalanced antenna (say a random length of wire), and an ineffective counterpoise or ground, will tend to have high earth/ground voltages present, and these in turn get back to the connecting wires for the keyer, and even on to the Morse key connecting lead itself. So all of a sudden you can find high voltages, many tens of volts or more, presented to the input and output connections on the keyer, resulting in overload, false triggering, or even destruction of the keyer electronics!

A scheme like you suggested, with say a battery powered keyer circuit, using opto isolation or even a good old relay or reed switch, should be fine in most cases.

The vast majority of the modern rigs only require low level switching, but as you quite rightly say, some of the very old transceivers or transmitters can have high voltages (I've seen some with 118V on the keying input), and or high power levels present. So long as its not AC, then many of the modern IGBT devices can be used to interface a low voltage keyer to a high voltage/high power source for switching.

Most IGBT's can easily handle many tens of volts, and in a lot of cases hundreds of volts at quite high current levels. There are also devices called "Logic Level Switched FETs", and they are designed to switched by a 5V logic signal. Again, many of them can handle high power levels without requiring a heatsink. One I used at work a few years ago for a project could switch up to 60A without a heatsink, and up to 160A if cooling was provided!

You could also 'borrow' ideas from commercial products for transceiver switching. One that comes to mind is the MFJ-704 amplifier to transceiver interface. To quote:-

The input is designed to be compatible with any transmitter or transceiver, and the output is compatible with AC control lines or DC positive or negative amplifier control lines having up to 200 Volts peak open circuit voltage and 300 mA of operating current.

The schematic can be found here: https://forums.qrz.com/index.php?attachments/arb-704-jpeg.82154/

I have a couple of old keyers in the attic that can directly switch high voltages, in both cases they simply use a reed relay. Being electro-mechanical devices they will  eventually wear out, but having said that the ones here are 50yrs old or more and still work fine! bd

I haven't checked, but there may well be off the shelf MCUs that have some kind of high voltage/high power switching peripherals already inbuilt.

73, Mark...

 

[Mickey Morse]

Thanks, hmm.  My original thought (that touched off my re-interest in CW) was that the Raspberry Pi Pico would be a good platform for a keyer and code practice oscillator.  Now I see that there are these other issues as well.  Does a relay actually switch fast enough to keep up with high speed CW?  If yes, that is the simplest approach.

I did a little reading last night and saw that in the spark era, the key in lower powered transmitters (but that may have meant 100s of watts or even low kilowatts) directly switched the AC current in and out of the transformer primary, with the high voltage secondary energizing the spark gap.  Maybe that explains why those Marconi-era keys are so big.  Higher powered stations (some were 100+ KW) apparently used relays, so that has me wondering what kinds of relays could have been fast enough at that power level.

I've also discovered that everything I envisioned with the Pico has already been done with the Morserino, though the Morserino is pretty expensive and I don't know how it is in terms of RF shielding or transmitter switching.

I guess if I do anything with the Pico in this direction, I'll start with simple code practice, and worry about actual transmitters only if I make some progress with the simpler parts.  Thanks for all this info.

 

[Ham4CW]

Does a relay actually switch fast enough to keep up with high speed CW?  If yes, that is the simplest approach.

Yup! Certainly most of the small ones can switch at surprisingly high speeds, more than fast enough for an isolated keyer.
The larger relays tend to require a few hundred mA on the coil, but more of a problem is the time taken for the core to loose the magnetic field after the current has been removed (hysteresis). This effect makes the relay sluggish in operation. The smaller ones are generally more than enough though, many can switch surprisingly high voltages at respectable currents.

Reed relays are fast, but they are not the easiest of things to find these days, and when you do find them they are not really all that cheap for what they do.

Large relays can be very noisy too, the smaller ones are not so bad, but you can still hear an audible click in most cases (though not loud enough to be a bother).

For practice  I bought a "Mini Memory Iambic Keyer" from Electro-Resales. They just sell the item, it's someone else's design.

The instruction sheet is here: https://nebula.wsimg.com/30f1953d31dfb03830096167c987bf3f?AccessKeyId=13A6205A294059260963&disposition=0&alloworigin=1

Much more info, including a schematic and the code for the micro used can be found here: http://www.strozzi.it/users/carlo/hamradio/iz4kbs-keyer/

It uses a small piezo buzzer for audio output, and a single FET for keying a radio. The whole lot is run from a button cell (CR2032 in my unit), which seems to last ages.

There are LOTS of simple keyer circuits around the web, very often using little more than a single micro and a few support devices. Well worth considering rather than re-inventing the wheel.

The biggest problem with using stand-alone micros tends to be programming them. Without a suitable programmer it can be pretty well impossible. And by the time you have purchased a programmer, or made one, the whole project starts to seem less attractive.

I'm lucky enough to have a few (programmers) laying around that I use for work, so the tools are already in place if I need to put some code in to a micro, whether a PIC or AVR device.

That's the good thing about Arduinos and the like, you don't need anything more than a free USB port to program them.

I have a couple of Morserinos here, one built and one still in kit form. I've never tried keying a transmitter with one. They do have an internal VHF transmitter module (only very QRP I think), so that two unit can communicate directly with each other. They can communicate over wifi too, but overall it's way more than you would need for your project.

You're quite correct about the old spark transmitters using quite chunky keys. I've seen a few examples that heat-sinks on the contacts to keep them cool! Also arcing was a big problem too. Mind you, all spark transmitters were outlawed back in 1923 for Amateurs so thankfully it's no longer an issue.

Low level keying is always best if possible, since keying high powers can lead to all kinds of 'nasties' being radiated, including key clicks etc. You can also get wide side-bands generated too, which can spread over tens of kilohertz either side of your transmission frequency. You might get lots of contacts but for all the wrong reasons! ;D

 

[Mickey Morse]

The pi pico is an amazing little board that just screams for project ideas, of which a keyer was one.  They are 4 USD in the US which is crazy cheap.  For some reason even though RPi is a UK company, the cost more in the UK (maybe from VAT?):

https://shop.pimoroni.com/products/raspberry-pi-pico

Anyway they have tons of both ram and program memory, and you just plug one into a USB port and you have an interactive Python prompt, if they are set up like the Adafruit boards.  They don't have true DACs but they have PWM output pins, so can produce sidetone that way, maybe requiring an external audio amp or transistor though.  I don't have one yet but I have an Adafruit Trinket M0 which is probably something of a predecessor:

https://www.adafruit.com/product/3500

You plug it into your laptop USB port and it implements two USB devices at the same time: a serial port (so you can interact with the Python interpreter) and a USB flash drive (so you can drag and drop Python programs and other files to the "disk").  It removes a huge amount of pain from programming microcontrollers.  Of course you can also program them in C if you prefer that, and similarly with the Pico.  Anyway, my hope with the Pico was that you could have a keyer and code practice system with just a single 4 dollar board, plus some kind of paddle (I love the hacksaw blade one) and I guess an audio amp.

A little bit more upscale is the BBC Micro:Bit v2:

https://shop.pimoroni.com/products/new-micro-bit-v2

The older v1 version could do similar things but had a lot fewer cpu resources, so make sure to get v2 if you want to play with one of these.  The Pico seems more interesting to me though, because of its lower price and more standalone programming system.

I'll certainly look at existing keyer designs if I try to build a real keyer, but I think at this point I'm not terribly serious about it.