Re: A simple 60 meter CW transmitter


I think the output of a typical DDS is reasonably clean.  If the object is to make a receiver that tops Bob Sherwood’s list, some extra pains might be required. 
No, phase noise can’t be filtered.  I think with a DDS it’s related to jitter on the reference and with a PLL/VCO, it’s due to “hunting” by the feedback loop which causes a small amount of FMing.  I’m not sure any frequency source can be completely free of phase noise.
A DDS can usually be had for less than $10.00 and Arduino Nanos can be gotten 3 for $10 postpaid these days.  It’s hard to justify a bare MCU chip at those prices.
As far as channel indication goes,  with enough pins you would have plenty to do an LED per channel, but yeah there are surely better ways.  Looking at my old project, I think I used a piezo beeper to announce the channel in Morse.  But if you want to get fancy, nice backlit white on blue 2x16 LCD displays can be had for $1.89 each, postpaid. Amazing.  You could also use three LEDs in a binary weighted scheme ...
There are several other advantages associated with having your LO / VFO under software control.  You could implement RIT – not that the other guy might be off frequency, but so you can adjust his pitch to your liking.  And if you choose to have a superhet receiver instead of direct conversion, it’s easy to do the offset for the IF frequency.  And that’s not all (sounding like a TV pitch man now ...) you could toggle between the right LO frequency to receive CW or switch to the right frequency for decoding USB in the channel.  That would make cross-mode QSOs possible and possibly make it a bit easier to make sure the channel isn’t in use before calling. Or just for monitoring.
Fun stuff ...
Nick, WA5BDU

Sent: Monday, June 15, 2015 12:38 PM
Subject: Re: [4sqrp] A simple 60 meter CW transmitter


No I don't have information to refute the phase noise situation.  My thoughts are that at QRP levels, the problem is not as severe and may be able to be ignored.  At higher power levels, it is definitely a problem on the bands.
What I do know is that all DDS units are not the same, some are worse for phase noise than others.

IMHO, if the design point is for small, QRP and portable, the DDS solution is viable, or something like the SI570.  The nice part about that approach is that the generator can be easily offset for CW receive so it can be used not only as the transmit VFO, but also changed to be the LO source for a superhet receiver.  It is a small matter of programming.  It also makes multiband capability possible in a smaller package.  How about an 80, 60 (channelized), and 40 meter SSB and CW transceiver the size of an Elecraft K1.

Once one accepts the fact that properly designed firmware can be used, the possibilities of a small display (like the 3 digit LED display on the Elecraft KX1) is not only possible, but can add a lot of function to a small package.
OTOH, that may be too ambitious for a 4SQRP kit if the programming skills are not available to make that happen.


On 6/15/2015 12:30 PM, Bill Cromwell wrcromwell@... [4sqrp] wrote:

Hi Don,

It's the close in phase noise that concerned me Some hams (and
engineers) have shown that it doesn't get filtered out. Apparently you
have to also filter out the desired signal to get the phase noise. Do
you have information that refutes this? It's easy to find misinformation
*everywhere* and maybe the best I have is wrong.

I would otherwise be happy to have something like that for my carrier
generator and heterodyne channel oscillators. My notion of seven xtals
(or available oscillator frequencies) is to allow for SSB *AND* CW in
one transmitter. That is one heterodyne frequency per channel (5) and
one each for either CW or SSB (2). Still five channels with a simple


Bill KU8H

On 06/15/2015 11:11 AM, Don Wilhelm wrote:
> Bill,
> Following a DDS with a low pass filter will take care of most of the
> crud and the low pass filter on the PA will take care of the rest of it.
> I have looked at the spectrum of the Elecraft KX1 and the output is
> clean even though the source is a DDS.
> With the right firmware, I think the KX1 could easily be put on 60
> meters, but that firmware is Elecraft proprietary.
> 73,
> Don W3FPR
> On 6/15/2015 9:52 AM, Bill Cromwell wrcromwell@... [4sqrp] wrote:
>> Hi Nick,
>> If we want "simple" LEDs would be overly complex and expensive. A
>> chicken-head knob can point at dymo label or sharpie numbers one through
>> five on the front panel..assuming we actually use a front panel.
>> I have been avoiding the DDS type stuff to also avoid the transmitted
>> crud that goes along with them.
>> 73,
>> Bill KU8H

Join to automatically receive all group messages.