Arduino based AM TX

[K1WIZ]

I'm trying to design an AM TX based around the arduino.  The purpose for the arduino is carrier generation.  I know the arduino outputs a square wave.  I chose the arduino because at this frequency, it's pretty stable and adjustable in code if need be (and they are super cheap! $6).  I've already written the code and I'm getting a 1 MHz signal on the output of the board.

My thinking was this:

1) arduino generates a 1MHz square wave
2) square wave excites a 1 MHz resonant tank - becomes sine at output?
3) buffer & filter tank output
4) pass buffered/filtered tank signal as input to a final running as class C (tube or solid state but solid state preferred)
5) modulate class C final directly (unless the option to modulate smaller signal at final input?)
6) output filter after final to 50 ohm load

the goal is to keep the external component count low, and use readily available components.  Prefer a power FET as the final, unless there is a good reason to use an ordinary transistor/tube.

[w8khk]

That is an interesting approach.  Are you looking at QRP or moderate QRO?  I am curious what you will interface for controlling the frequency?  Maybe a shaft encoder?  Frequency displays are very inexpensive for the Arduino platform too.

Although I am just finishing up an all-tube 40 meter exciter for my legal limit tube rig, I have been studying the Arduino possibilities, and the route I am considering is interfacing the AD8950 DDS to the Arduino.  There are lots of examples of this on the web, and parts count is small, and VERY inexpensive.

What I would like to do is use the DDS output as the receive local oscillator for a single conversion superhet, then on transmit reduce the DDS frequency output by 455KHz, for the transmit VFO.  This would allow easy netting; the transmitter would automatically be on the net frequency just by zero-beating the received signal.  Would work great with an old HRO or RME.

I'm very interested in following your progress!

[K1WIZ]

Rick,

Thank you for your reply.  I am trying to build a QRP Broadcash band TX for playing some OTR and Art Bell's internet radio show (Dark Matter Digital Network).  Keith Rowland (Art's Webmaster) has already given me written permission to broadcash the show in my town.  In my town, 1MHz is crystal clear (strangely even at night), and I figured it'd also sound good on my RCA Strato-world tube receiver.  (been dying to put it to good use).

I'm not looking to change the freq, rather I wrote the code to divide the clock frequency (16 MHz on the Nano) and output the wave on pin 9.  I plan to document the project on my personal tech blog: http://wizworks.net   to share the idea with others as well who are into LPAM. 

When I write up the project, I plan to also publish the code so others can take it, and maybe even improve it. 

I'm just trying to get a little help on the RF engineering so I can be sure the square out of the Arduino is properly converted to sine and filtered, as well as ideas for modulation.  I figured modulating the final in class C might be the easiest, but maybe some guys here have a better idea to be considered.

Grateful for any help I can get.   

Regards,

John

[w8khk]

John, that sounds good.  I think it's great that you have the opportunity to broadcast locally with your own transmitter.  I often thought that would be fun when I was younger.

Obviously no need to change the freq for your application.  (At first read, I thought you just used one MHz as an example.)   I am thinking that a non-symmetrical square wave, with a narrower on-time duty cycle would drive a class C final just fine.  I believe all the flywheel action needed  to recreate the carrier sine wave should happen in the plate tank, maybe a Pi-L network would be best. 

So many ways to modulate it...  Standard plate modulation, Class E PDM, etc.  If you use a tube, you can get some fabulous fidelity with screen modulation.  So many options to try.  At that low  power level, solid state all the way should work fine, the attention to detail will be in the frequency stability and the linear modulation.

Have fun with your project.  I will be sure to read the blog as it transpires!

73, Rick

[R. Fry SWL]

... I am trying to build a QRP Broadcash band TX for playing some OTR and Art Bell's internet radio show (Dark Matter Digital Network). ...

Background: unlicensed use of the AM broadcast band in the US is permitted by FCC §15.219, but the d-c input to the final r-f stage is limited to 100 mW, and the length of the radiating conductors of the antenna system (including the conductor to r-f ground) is limited to 3 meters.

You might want to consider using a carrier frequency more at the upper end of the AM broadcast band, where the radiation efficiency of that antenna system is best.

The useful coverage radius of these systems to indoor, consumer-level AM receivers typically is 1/4 to 1/2 mile.

Below is a plot of field intensity at 1 mile, for the conditions shown:

[K1WIZ]

Thanks for the tip on the part 15 use of the MW band.  I chose 1MHz because the top side of the band is littered with stations and I didn't want to interfere with a DX listener who may be living nearby in my area.   

[W7SOE]

It is difficult to get higher frequencys from toggling a microcontroller port pin, perhaps using a timer with dedicated output....  Sometimes the best they can do is 100's of kHz and even then the waveform is degraded.  How about a dds controlled by an Arduino?

Rich