75S-2B Receiver
Bottom view. Chassis is 10 X 17 X 2 inches. Aluminum plate is mounted under three section variable to strengthen up things. Vertical shields provide rigidity, circuit isolation and mounting surfaces for switch sections. Can in center of chassis houses VFO coil.
Inside rear chassis. BFO module is on the left. Arduino shield covers Arduino Uno module under it. Plugged into the shield is a Si5351 synthesizer module from QRP Labs that provides the BFO signal to the product detector. Directly under the QRP Labs module is a homebrew amplifier built up on a proto board and described at the left. In the center is the frequency counter/LCD driver shield, with another Arduino Uno under it. Compartment on right contains filters for the ribbon cable that carries grounds from the bandswitch to the counter inputs.
Rear view. Extra holes in vertical chassis are artifacts of the initial configuration, when I thought three Arduino controlled DDS sources would work.
Close up of preselector dial and main tuning dial. Preselector uses a 5:1 vernier drive. Dial was made from an aluminum disc painted white. It was overlaid with a waterslide decal, designed using MSVisio, that shows the settings for the classic ham bands. Main tuning knob was carefully drilled and tapped to install a custom spinner grip. The tuning mechanism is very smooth with no backlash. An Arduino module drives the LCD, which displays frequency, band and mode.
BFO amplifier module using LMH6703 surface mount current mode op amp. Backside of board is covered with copper foil tape. These chips have a 1 GHz bandwidth, and without careful layout will oscillate if you just blow on them. I am not fond of all the micro-soldering required to make one of these boards. This amplifier was built when I was going to have a first oscillator going up to 33 MHz. It's overkill for the 455 kHz BFO, but since I already had it I used it anyway.
Top view of chassis. Two section variable on the left tunes the preselector. Vernier drive is mounted directly to the front panel. Vertically mounted three section variable capacitor in the middle is from a scrapped WWII ARC-5 receiver. One section tunes an analog vacuum tube VFO, the other two sections tune the 3.5 to 4.0 MHz bandpass filter. Collins filter for SSB is just to the right of it. Further to the right is a Murata ceramic filter for AM. Slug tuned coils in cans were scavenged from a piece of junk RF test equipment. Upright chassis at rear houses the Arduino modules for the frequency counter and BFO. USB ports on Arduinos are visible through holes adjacent to each power cable. This allows sketches (programs) to be loaded without removing modules from the radio.
Receiver front panel is a 4U size rack (7 by 19 inches). Radio is housed in a custom cabinet made from .062" perforated aluminum sheet. Front panel, including the preselector dial, is finished using the same processes as my other recent projects. Speaker is built-in because there was plenty of room for it, and it eliminates the need for a separate box. All water-slide decals were made using the techniques described in my homebrewing tips document.
At the buttons to the right you can download the latest schematic and a detailed technical description. The tech description has pictures that can be expanded for a closer look at the hardware. Also available is a spreadsheet tool that was used to design the preselector. The Arduino sketches (i.e., programs) are available as .txt files that can be copied and pasted into the Arduino application (IDE).
Below is a block diagram of the receiver in the final configuration. A total of 10 miniature vacuum tubes are used, so it is mostly "hollow state". Its architecture up to the output of V4 is a clone of the Drake 2-B. After this point it is just like a Collins 75S-3.
Close up of crystal heterodyne oscillator board. Cheap microprocessor crystals are used to provide a 11.0, 18.0, 25.0 or 32.0 MHz injection signal into the first mixer to cover 40 through 10 meters, where the radio operates in dual conversion mode. The crystals were less than one dollar each! For 80 meters the oscillator is turned off, and the first mixer becomes a straight through amplifier, making the radio a single conversion type. This is the same scheme used in the Drake 2-B receiver.
This radio started out to be a clone of a Collins 75S-3 receiver. I had acquired a 2.1 kHz Collins SSB filter and wanted to build something with it. My original plan was to substitute Arduino controlled AD9850 DDS modules for the crystal controlled heterodyne oscillator, Permeability Tuned Oscillator (PTO, aka VFO) and BFO used in the original receiver. That didn't quite work out due to spurious signals emanating from the first two DDS sources, so the radio ended up in the configuration described here. You can read about the trials and tribulations with the solid state signal sources in the technical description at the link at the bottom of this page. It's called the 75S-2B because in the final configuration the front end is similar to a Drake 2-B, while the rest of it looks like a Collins 75S-3.
KG7TR.com