I built my first combostar/picastar back in 2012. I've made several modifications over the past three years including adding a STM32 based embedded board to do FreeDV. Many of the parts were difficult to get and even more so now. I learned so much from its construction and the wealth of information from the homebrew group that I'd really like to see an update or maybe something entirely new. The DSP used is very old in the design, the TFT controller parts are dated some aren't made anymore. Here is the short list of reasons to update or come up with a new design for a homebrew HF rig.
- There are better parts for the GUI out. Integrated graphics STM32F4 would be a great upgrade. Maybe multiple screens and small OLEDs have come way down in price for adding addtional controls at low cost and ease of use.
- The DSP could be replaced with a STM32 series ARM keeping it all on the same platform.
- There are already improvements to the H Mode Mixer and maybe a direct conversion would possible now with some of the low cost TI frontends and low power/cost FPGA.
- Codec2 is widely availible and there is no reason not to include not only this digital mode but a hardware platform that may run future digital modes.
- Modularization of the GUI, RF sections etc. I really like the idea of mounting the radio somewhere and the GUI/MIC elsewhere. I would also like to see VHF/UHF modules that utilize all the work done on digital modes/GUI keeping the idea of a remote/even distributed GUI Head.
The long wait begins... The first SBHT Prototypes have went to fab and parts have been ordered. There are a lot of untested sections in this design so build up of the first rev boards will be slow.
Just a sneek peak at what I've been slowing working on for the past few years. After building a number of test sub assemblies, filter testing, device selection flops, and lots of hours I've started routing the first prototype of a VHF Transceiver capable of TDMA operation at decent wattage.
I wanted something small and almost disposable for APRS tracking. The DRA818V fits the bill but needs low pass filtering. Some photos of the prototype boards and output floor while TX outside intended band. The SMD filter works great, inexpensive, and makes for a nice package.
If you are in the market for a soldering station you can't really go wrong with MetCal. The are dependable and easy to repair should something happen. Instant heating, well placed heat, and an easy to use hand piece are just a few of the reasons to use an inductive RF soldering station. New stations have come down in price and units from Thermaltronics are compatible.
Big issues are if you haven't seen one or not generally "in the know", finding used units that are usable can be difficult. First, to have a working, station you will need 3 main components. Below is a list of recommendations about how to get everything together. I'm sticking with the high end industrial models and not the smaller units.
Power Supply / RF Unit
RFG-30 - These are the oldest MetCal in this series. Good units. No auto-shutoff. Has one LED Light
PS2e-1 - These are the next version with Auto-Shutoff. Please note that these were produced in 220V and 110V. You want a -1 or -01 unit for 110V. Has 2 LED Lights.
MX-500 - These are one release behind current. They offer two outputs but you can only use one at a time.
Any MX-500 Hand piece will work. MX-RM3E
Newer Metcal MX-H1-AV
Thermaltronics - SHP-1
You will want either the MetCal Stand or the Thermaltronics. They have magnets built it to keep tip cooler when not in use. SHH-1
This tool comes in handy for organizing kicad footprints and modules. Unfortually it doesn't build on osx normally. I have created a homebrew repo and after ignoring some build errors it seems to compile and work great on OSX.
The main site for the tool is at http://www.compuphase.com/electronics/kicadlibrarian_en.htm.
My homebrew repo is located at https://github.com/shaneburrell/homebrew-kicadlibrarian
Prototype boards for testing the RF module along with provisions for a low pass filter have been sent off for manufacture. I'm expecting to see these back in a few weeks. In the meantime below are links to the Gerbers for the test board as well as Kicad Modules for making your own boards that use the DRA818V module.
I've been looking at migrating some of the work I've done on my ComboStar build to VHF/UHF including GSMK and Codec2. With 3D printed case, lithium battery, STM32F405, DRA818V/U, and a nice OLED screen I think its possible to build a DIY 1 Watt radio that support analog, FSK (APRS), and Codec2 GSMK. I already have several working STM32 projects with successfully generating various modulations including GMSK, POCSAG and AFSK. My first step is to play with the DRV818V module and see how difficult it is to implement. From the datasheet I've included a low pass filter in this prototype which I should be able to prototype with my existing STM project board.
It has taken almost 2 years but I finally did a local test on 30 Meters of the modified combo-star radio with a STM32F4 based digital codec board running the last port I could find of Codec2. There is still a lot of work to be done packaging, winding the rest of the filters, and completion of the high power sections. Must of the time was spent on independent projects and assembly of the combo-star was quite a undertaking. Although a hacked combo-star which I'm sure most of them are in the end given the home brew nature, I've got something now I can work with that is all integrated. Eventually I'll even have it in a ready to go box.