RF generation

Michael OBannon

I'm interested in using the Propeller for a ham radio application. Several threads on using the counters for signal generation seem to indicate that RF signals with high spectral purity can only be generated at a small number of frequencies. (Please correct me if I am wrong.) Is there any way to use the Propeller to generate high quality RF in the 3.5 to 30 MHz range by adding a few external components?

Thanks,
Michael

RinksCustoms

um, yes. sift throught this thread and you shal find something very close to what your looking for, written by Chip Gracey i think...


http://forums.parallax.com/forums/default.aspx?f=25&m=113109

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Definetly a E³ (Electronics Engineer Extrodinare!)
"I laugh in the face of imposible,... not because i know it all, ... but because I don't know well enough!"

TransistorToaster

I looked at that problem before. Look at these threads:

http://forums.parallax.com/showthread.php?p=627973

http://forums.parallax.com/showthread.php?p=583301

The Prop DDS operates as a 1 bit DAC and 32 bit tuning word DDS.

I was thinking about "'wiring" the internal Prop counter to a several bit R2R DAC. The idea would be to have an assembly routine copy the most significant bits of the counter phase register and place it on the output pins. You could compare this idea with the fact that the ad9850 DDS is 32 bits with 10 bits running up to 125MHz.

rjo_

That's an MRI frequency... so maybe we can add very large magnets to the list of Propeller peripherals in the future[noparse]:)[/noparse]

Rich

Michael OBannon

After reading all the threads again, it appears that using the Propeller as a signal generator in the 3.5 to 30MHz range will result in both harmonics and subharmonics for most frequencies. This is because the signal is not a perfect square wave and there is some phase instability in the signal. (For a few frequencies, a perfect square wave will be generated and this will not be a problem.)

The harmonics are a not a serious problem for amateur radio applications. Low pass filters are relatively easily designed. The subharmonics are more of a problem. Since I do not have a spectrum analyzer, it is hard to test possible fixes.

TransistorToaster, could you say some more about using an external R2R DAC? This sounds like a workable approach. I am just getting up to speed on DDS, so my understanding is a bit rudimentary at this point.

Don D

Let's not get confused and cause all kinds of havoc with our neighbours!

Perfect sinusoidal waves do not cause harmonics. Anything else, especially square waves with any degree of perfection will create a plethora of unintentional interfering rf energy which can easily be transmitted through walls. Not too much of a problem if you live in the country, quite another situation if you live downtown anywhere!

For most applications, I would opt for the corded and shielded video signal with a metal project box: this always gives a better image.

Please take care: we wouldn't want our favourite chip to make headline news THAT way... ; )

TransistorToaster

Michael,
>TransistorToaster, could you say some more about using an external R2R DAC? This sounds like a workable approach. I am just getting up to speed on DDS, so my understanding is a bit rudimentary at this point.

You probably know about how an R2R DAC works. A DDS contains a phase accumulator that essentially generates a sawtooth waveform http://en.wikipedia.org/wiki/Sawtooth_wave. The minimum value is 0rad while the max one is 2pi. The speed at which the counter progresses through the ramp is the phase delta =freq in the DDS accumulator: theta:=theta + freq modulo 2pi. Taking the MSB is essentially just comparing the sawtooth to the middle value. If you take a look at the length of time the MSB is 0 vs the length of time it is 1, it will in general vary thus causing jitter. A commercial DDS takes cos(phase) and places the result on a DAC. On the propeller, the math function is sign(phase).
What I am saying is that if you output the phase onto the DAC, the jitter should be much reduced. You will still have harmonics above the fundamental frequency just like the current Prop sign(phase). If I am not mistaken, the technique I am presenting here should not jitter. I'll give it a bit extra thought. Obviously, if you hook an external parallel eeprom to do a lookup Cos function of the phase(address) it receives, then you would get even better results.

TransistorToaster

>The harmonics are a not a serious problem for amateur radio applications. Low pass filters are relatively easily designed.
It depends on what you are trying to do. I would tend not to venture on a frequency that is not a r/n multiplier of the Prop's clock and use only values of n that divides 2^32 exactly.