Ron K0EIA has just “finished” his uBitx build and made several contacts on CW and SSB with good reports. He also listened quite a lot to the contest, shortwave, and BCAM.
He says, “Receiver sensitivity is very good. My commercial radio is an old Kenwood TS-830, and I live in a quiet location up the side of a mountain. I have tuned very weak stations on the KW, and A/B tested with the uBitx. I have yet to find a signal that I can’t hear on the uBitx. It’s a good little radio.”
He finds it difficult to overstate his enjoyment of this build project. He said “finished” above because there are still a lot of experiments to try with my now fully working radio. Ron gives kudos to Farhan, HF Signals, Raj, and all the helpful people in the community.
- v5 board
- stock display
- KD8CEC v1.22.
Ron first assembled the rig with stock software on an open aluminium chassis to make sure he understood the parts. He also has a few pieces of test equipment which really helped: a frequency counter, an oscilloscope, and a software controlled dongle (Analog Discovery 2) that can do FFT (a poor man’s spectrum analyzer).
By far the most time was spent by Ron in reading through the IO Groups list postings and various blogs to understand details of the design or program features. In turns out he had the encoder wired up backwards. He also had several plug jacks wired wrong. etc. There were the normal assembler errors – each another research exploration.
Calibration was challenging for a few reasons. The calibration procedure for the stock software was fine, but it’s different from the procedure for the CEC software, so he had to relearn the procedure.
Ron had to figure out the CW Shift programming, to understand the design for sending a CW carrier. He finally adopted Ian’s mathematical iteration approach using the uBitx manager 1.11 tool, with a spreadsheet and the frequency counter (the third calibration procedure learned). Ron was able to get the uBitx to within just a few Hz of the frequency counter reading. Ron says, “Of course my counter isn’t calibrated, and the uBitx probably isn’t temperature stable so this was mainly a learning exercise. I will be checking the calibration regularly though!”
Setting the BFO
The BFO setting was another significant challenge. Ron’s arrived from HF Signals with a setting of around 11052. He could not hear anything because the BFO was putting the signal way outside the bandpass of the crystal filter. He suspects that this is a common occurence, based on how many people report not hearing anything when they turn on their rig for the first time. Ron ended up with the BFO close to 11055.5 KHz, some 3.5Khz away.
Many on this list recommend using FLDIGI or other audio frequency displays to refine the bandpass. Copying this approach, he put his AD2 on the speaker terminals and looked at 0Hz-5000Hz. Perfect! Ron could see the bandpass of his crystal filter, and adjusted this to get the bottom skirt down around 350Hz. This is still a work in progress since the same filter is suppressing the carrier on transmit, so there is a balance to be struck.
On transmit, Ron was able to put the AD2 on the RF output (using a 47dB attenuator tap on his dummy load! i.e. Don’t blow up your test equipment!.
Ron can look at the difference in the fundamental power on CW, vs keying the mic in SSB with no audio input. It looks like his carrier is suppressed about 30-35dB. This seems ok but he can still hear the carrier whine if he tunes near it using a nearby receiver. He is still thinking about the tradeoff between carrier suppression and good low end audio. He thinks he is likely to fiddle with the BFO some more sometime soon.
Ron spent a lot of time figuring out which version was the latest version, for his particular board and his particular display, and getting all the steps squared away. Once it is figured out it’s just a moment to load, but it took quite a while to get everything organised.
Spurs and harmonics
The big question everyone has is what about spurs and harmonics. Using his AD2 as a spectrum analyser, Ron’s radio looks pretty clean, but he can see the third harmonic. Using his non-calibrated bench equipment, it looks like his build has the third harmonic down about 38dB. He played around quite a bit with different bands, adjusting the gain on RV1 down to output of 5, 4, 3 watts. He also tried using CW and putting a 1000Hz at 50mV signal into the mic circuit. He says, “This testing is not conclusive because it’s at the upper limit of my skill and tools … but I definitely see the third harmonic and it’s not below 43dB on my amateur bench. So I’m going to do more testing …”
For the moment, Ron is not losing any sleep over a stray 500mV of RF signal. He is most likely simply going to add an external filter to take out the remainder of the harmonics.
Ron’s advice to a new person who is thinking about building this radio, “it’s pretty obvious, but this is not a plug and play building experience. It’s not the linear kit build of the Heathkit days, with check boxes and step by step instructions. The uBitx is a kit of parts, you have to figure out how to put it together … and there are a lot of options and choices you have to make.”
Ron did not keep track of how much time he spent on my build, but it wasn’t soldering iron time. By far the time was spent reading through this group, other builders’ blog posts, and some basic electrical component theory (how does the click-switch thing on a variable pot actually work? How does a rotary encoder actually work?).
When he finally figured out how it is supposed to work, soldering up (or loading software) represented a small amount of time. He has built a few things (smaller than this) with step by step linear instructions, but he took on the uBitx build as a step-up challenge.
Ron’s test equipment really made a difference. He believes that some really skilled builders get it built using nothing but a soldering iron. However, in his view, that is an extra level of difficulty! His advice is not to try to do this kit as your first beginner kit, and don’t handicap yourself with having no test equipment at all.
All this being said, he loved doing this build. It was a great learning experience and he wanted to thank group members for being such a helpful and friendly community.