The 17-meter transceiver project is underway. I got the idea for this project when I was disassembling a CB rig a couple of weeks ago. There was an 11050-kHz crystal inside, and I wondered what I could do with it. I figured out I could mix it with a 7-MHz signal to produce output on the 17-meter band. My first attempt was to build two VXOs, one using the CB crystal and the other using a 7040-kHz crystal. They would be tuned with a dual-section variable capacitor at first, and varicaps in the finished version. After some experimenting I determined I wouldn't be able to get enough frequency shift from the CB crystal, so I decided to build a 7-MHz varactor tuned VFO instead. View a block diagram of the transceiver (subject to change as work continues).
This transceiver will have a direct-conversion receiver. The transmitter will probably be switchable between 1 and 5 watts out. I'll probably use a pair of SA612 mixer ICs, one to premix the VFO and crystal oscillator and another as the detector.
Building the VFO was simple enough; what I needed to determine was the voltage range the varactor (MV209) needed to swing from 7018-7050 kHz, enough to cover the band when mixed with the crystal. That was the purpose of the breadboards shown in the next two photos (click the photos for larger versions if you really want to see 'ugly' construction!).
These breadboards tend to take on a life of their own. The 11050-kHz oscillator is at bottom left. You can see the varactor poking its head out at top center. With the exception of the toroid, all parts came from salvaged hamfest "junk." That's one of the beauties of building from scratch: It's cheap!
I used a couple of trimmer caps to set the frequency limits and measured the varicap voltage. When I reconstruct the almost-finished version (homebrew stuff is never finished) I'll be able to set it up pretty quickly. You can see one of the trimmers towering above the other parts.
The VCO reassembled. The 1N4148 diode, lower left, is in series with the tuning potentiometer, and positioned right next to the varactor. It will provide temperature compensation for the varactor.
The best laid plans ... I had to do some final tweaking to get the VCO on frequency. Fortunately, it was only a matter of removing a few turns from the coil, as you can see. The coil now sits on a plastic standoff pressed into a hole on the board. Once everything is on frequency and ready to travel, I'll hot-glue the coil in place. Q-Dope® is for glue sniffers.
The pretty VCO shows off its pretty sinewave. Tomorrow* I'll finish fixing up the coil, reinstall the crystal oscillator, and give them both buffer stages. The oscillators will eventually be separated by a partition.
* It's a good idea to know when to quit for the day.SUNDAY UPDATE
All of the work today revolved around the VCO. I only need 32 kHz of span to cover the 17-meter band. I made some measurements and determined the voltage range I needed to apply to the varactor. It's just about a volt. I hung a couple of diodes in series across the output of an adjustable regulator, with a series resistor. I was able to pick off the voltage I needed by connecting the tuning pot across the diodes, which gave me 1.2 volts at about 4.5-5.7 volts above ground. Perfect! The regulated voltage is also the right value for running both oscillators and the SA612 mixers.
The VCO buffer uses two 2N706 transistors salvaged from a Systron Donner module, and an output transformer salvaged from a CB rig. I measured the primary and secondary inductances, determined the ratio was about right (4:1) and hung a cap across the primary. When I connected the scope I was able to peak it right up.
The other accomplishment for today was to name the rig, in accordance with QRP tradition. As it uses two oscillators, at 7 and 11 MHz, the right name was obvious. "7-Eleven" is a registered trademark, so I made it "7-11." I hope they don't mind.
The oscillators are in place and on frequency. I decided to try using the crystal oscillator without a buffer, but there's (barely) room for one if needed. Click the image to see a larger version.
From left to right, top row: Regulated power supply, 11050-kHz crystal oscillator. Bottom row: VCO and buffer.
Now on to the premixer, after which it will be a very QRPp 17-meter transmitter!
Hard to believe that 10 days passed since the last update. The SA612 mixers were giving me grief, and I decided to rework the board so it would fit in a Radio Shack project box. Right now I'm using an SBL-1 mixer to premix the VCO and crystal oscillator. I may venture to try an SA612 on the receive side. My goal is to have the prototype finished next week, because I'm meeting a friend at the Gainesville convention and he'll bug me if I don't bring it along in some state of completion. What are friends for?
The oscillators are now nicely mixed. L-R, the crystal oscillator, room for a post-mixer filter, SBL-1 mixer, VCO.
I finally finished the 7-11, and one of these days I'm going to digitize and post the schematics. But here's a photo of the inside.
I ended up using SA612 mixers to combine paralleled 7.040- and a single 11.059-MHz crystal, and for the receiver detector. The toggle switch removes the inductor and lets the 7.040s move up a few kHz. The spot switch adds the transmit-offset capacitor in receive mode, so I can zero beat the other station, then release the button and know I'm on the right frequency, an important consideration with direct-conversion receivers.
The transmitter is along the back wall, the VXOs are in front, left. The receiver and audio filter/amp are on the right. Output is 1 W, using a 2N1711 scrounged from some surplus test gear. The two T50-6 cores at right angles and separated by a shield are the LO bandpass filter, to ensure only the 18-MHz product reaches the receiver and transmitter. The receiver has its own bandpass filter. I can hear the leakage on my URM-25 signal generator. Aside from the PA, the circuit uses 2N2222As, an op-amp audio filter, and an LM386 audio amp.