Click on thumbnail for full size picture.
SECTION 1 - Project Background (last update 3/21/01)
This project is inspired by all those that have pioneered the previous development with the NORCAL Sierra (i.e. Wayne Burdick W6NR) and the AD9850 Signal Generator/LO software development (found on the NJ-QRP HAM-PIC list).
I have been lucky enough to work with Mike Branca W3IRZ on his Georgia Sierra MODs for a different Sierra IF (i.e. 4.000 mHz) and a varactor diode tuned VFO operating from 5.8 - 6.0 mHz. Corresponding BP freq and inductor/cap changes were made where necessary. The details of the Georgia Sierra can be found on the noga-tech list at http://www.topica.com/lists/nogatech in the form of archived emails that the NOGA members contributed during the initial year the NOGA club members participated in this collective effort. This effort is still underway by several NOGA members and their story will unfold in time.
The "Georgia Sierra" moding effort by Mike W3IRZ was a labor of QRP Joy and in no way detracts from the Wilderness Radio effort, the original NORCAL club offerings or other inspired efforts i.e. the Elecraft K1 that has it's roots in the Sierra design philosophy.
The project code named the Appalachian is just one more MODing implementation of the NORCAL Sierra that appears in previous ARRL Handbooks with template and construction documentation on the ARRL Web site:
The NorCal Sierra: An 80-15 M CW Transceiver The NorCal
Sierra: An 80-15 M CW Transceiver Most home built QRP transceivers cover
a single band, for good reason: complexity of the circuit and physical
layout can increase dramatically when two or more bands are covered. This
holds for most approaches ...
http://www.arrl.org/tis/info/pdf/96hb1789.pdf
04/11/00, 752721 bytes
The NorCal Sierra: An 80-15 M CW Transceiver
http://www.arrl.org/notes/hbk-templates/sierra.pdf
11/10/98, 1337710 bytes
This NORCAL club effort is a classic project and to this day rates in the highest lists for QRP backpacking rigs. It's difficult to improve on such a classic but IMHO the K1 attempts to take it to the next level of incarnation.
My personal efforts are driven by the desire to use the AD9850 DDS frequency generator in a home-brew effort to accomplish the the LO and tuning portions of an all band CW QRP rig. Rather than reinvent the wheel for a basic platform I have elected to use the NORCAL Sierra style motherboard as the basics for the TX output (including keying) and the complete RX. Since it's a concepts rig I am NOT electing at this point to make a crud-proof RX front-end and find the original Sierra very adequate for casual and QRPP (<= one watt) contesting).
I am a old timer in the programming field and thought that combining this environment with a PIC controlled AD9850 would make things easier. Again on the shoulders of hams previous efforts I not only found available resources and help but a lions share of the needed code available on a HAM-PIC reflector managed by the NJ-QRP at http://www.njqrp.org/ham-pic/index.htm . I hacked my way through the existing efforts by the previous HAM-PIC contributors and added my MODed code to the list last year. You can pull the code from their archives under the HAM-PIC Member Projects section at http://www.njqrp.org/ham-pic/ae4gx_sglo/siggenlo.asm.txt .
It is a work in progress but as is offers TX offset and RX RIT so you should be able to make easy MODs to this code to allow for a IF offset and provide appropriate frequency management to the DDS depending on the LO being on the high side or low side of the desired RX range. Similiar MODs would be needed for the TX side if you elect to MOD the rig to SSB capabilities with conventional TX filters and balanced modulator methods. I'll save that effort for later.
So where am I at this point:
I have a working test bed PIC/DDS currently with 66 mHz clocking (limits me to about 22 mHz LO frequencies) (I have 100 mHz available for the Appalachian to get up to 33 mHz LO operation). I think the rule of thumb is 1/3 of the clock frequency to to safe. You can check out the AD9850 specs at the Analog Devices web site at http://www.analog.com/pdf/AD9850_e.pdf
Click on thumbnail for full size picture.
BTW Far Circuits at http://www.cl.ais.net/farcir/ carries the PIC/DDS PC board.
Next Steps:
1. Add a 10 dB linear stage to follow the present output of the AD9850 to get the output to about (0 dBm or 700 mV P-P). I'm getting about (-10dBm or 200 mV P-P) out of the basic AD9850 into a 50 ohm resistor load.
2. Build up Sierra Final and Driver sections with basic components in the original article with substitution where necessary due to parts availability. The driver needs (0 dBm input) The input to the final needs to be at the (+10dBm input) level for that stage to generate the ~1 watt output to the antenna (+25-30dBm).
*********
Note
*********
Since the DDS LO will be able to generate the actual
output TX freq (desired) then the three original Sierra sections can be
eliminated : VFO/RIT, VFO/pre-mix xtal mixes, TX mixer and carrier osc.
This makes about 1/3 of the Sierra motherboard space vacant. I plan to
overlay (with standoffs) the PIC/DDS board over where this eliminated stages.
But you ask what about the BIG band module socket. Well the good news since you can eliminate the two sets of TX BP filtering and the original pre-mix xtal stages the band module plugin can be reduced to less than half the original and just needs to handle the TX final LP filter and the input RX mixer and TR. components.
You have the choice of IF frequency and it could be the original 4.915mHz or the GA Sierra 4.00 mHz or anything you can get cheaply (I'm looking in the range 8 mHz to 12 mHz). I need to do a spur analysis on the available common xtals before selecting the IF if I don't use the 4.0mHz of the known GA Sierra method. Note: I have recently decided on 4.915 mHz to get better CW BW for the filter.
I'll be adding and changing this section as time goes on so check it out periodically.
Sam Billingsley AE4GX Atlanta, GA North Georgia QRP Club