Good:
Small and light
Cheap
5 bands 80-60-40-30-20 (other options likely soon)
SSB and CW (actually AM and FM too)
Lots of filter options
Very efficient PA (70% or better)
Built in SWR bridge
Wide range of supply voltages 5-14 Volts
Not so good:
Weird feedback on RX
Poor AGC (slow attack)
Slightly strained and odd RX audio (plus noisy AF amplifier stage)
Relatively poor ultimate rejection of filters (esp noticable on CW)
Plastic case might melt in a hot car (!)
Distorted SSB TX
Display invisible in sunlight
Similar list to yourself, Rx seems to be quite poor really.
I have used it at home, and on several activations for SSB and CW. It works, and I make contacts, but it is not quite up to providing me with a lightweight radio for multi day treks.
I am starting to think about getting a mountain topper, but they are quite expensive fora CW only radio.
I built one earlier this year. TX works ok for CW with 4-5w out across the bands, second and thrid harmonics are well suppressed. But the receiver I’m unimpressed with. I spent a fair while going through with my 'scope and found some problems, including the ~78kHz sample clock causing ringing on the analogue reference voltage due to a bad (resonant!) choice of filtering components in the design. I raised this with Manuel DL2MAN who didn’t take my suggestion of a component change very well - he seems to want to keep to his Creative Commons No Derivatives license - far enough its his design.
I don’t think the receiver has been designed carefully and the more I explored its quite clear that the poor spurious responses and image cancellation are attributable to both the circuit design and PCB layout.
I realise its not an easy task to produce a design like this in the middle of world wide component shortages, but I wouldn’t personally recommend the (tr)uSDX at this moment for SOTA.
Interesting. I received a critique of several parts of the design from a well known electronics expert. He did not want it circulated or to be identified. He noted a number of problems in the way the design has been implemented that have resulted in inconsistent performance. That being said, it’s a very clever implementation and it’s far from easy to develop something as ambitious as this.
Oh yes, we should acknowledge how much Manuel and Guido have achieved, especially with the current component shortages.
I just feel (also from observing discussions on their forum) that the hardware design is a bit ‘try it and see’ rather than reasoned through, which is a shame. As an example the analogue reference voltage issue I identified, the ripple is reduced by changing one inductor to avoid resonance with the sample clock. For some reason (perhaps to simplfy the BOM?) they have deviated from the value given in the processor reference circuit. Due to the CC-No Derivative license I am not allowed to publish circuit changes, but its always useful to read the component data sheets
I did enjoy building it. The 12v power socket has become intermittent on mine (shame its not a standard size), sometime I’ll replace it and have another go.
In addition to the points listed (esp. RX audio and filtering) what bothers me most is the effects of the TX/RX switching using a relay:
From an audio volume of “9”, the cracking when the relay drops becomes unacceptable, at least for my old hearing … so not well suited for working with weakly audible signals and audio volumes >“9”.
What is inconvenient for the user is that when changing bands, the selected tuning step is not retained (preferably per band and mode).
Owen Duffy has also made some considerations about the statements made by some regarding the SWR values displayed. https://owenduffy.net/blog/?p=25460
I was hoping that people would post to suggest that my experience was atypical. It’s disappointing to learn that it’s probably not.
Some of the problems might be fixable by others but the designers appear to have stopped this with a restrictive licence that stops anyone publishing anything. Recently someone redrew the circuit diagram to aid clarity and they were made to delete it and all links to it. Creative Commons — Attribution-NoDerivatives 4.0 International — CC BY-ND 4.0
I agree with Jonathan that is hard to recommend it in its current form, even at the low price.
I’ve had a pre-built unit for about a month now. I didn’t buy it for SOTA, but to have an inexpensive radio I could toss in a pack to take along on a run or on a trip where “radio” wasn’t the primary purpose. The radio definitely has flaws as described above, but it does work. I’ve done a bit of POTA hunting as well as using it for a SOTA activation (summit already activated by me earlier in the year so there was minimal risk).
To keep the overall station small, I use a single earbud and the internal mic. Battery is one of the internal batteries for my KX2 with a rubber band to hold it all together as a single unit. My antenna is a miniaturized 40/20 EFHW using K6ARK’s mini 49:1 transformer kit and a miniaturized loading coil (.35" diameter x 4" long with 170 turns of 28awg wire for the required 34uH). The antenna measures in at just under 40’.
It won’t replace my KX2 for “serious work”, but it’s fine for the use cases I had in mind when I bought it.
Today I activated HB/SZ-021 on SSB using the (tr)uSDX and a MP-1 antenna. It might not be the most representative summits for tests because there is a huge broadcasting tower on top which is known to make radio operation difficult - mostly on the 2m band, though.
There are some positive points to mention:
The SWR meter works to tune the sliding coil of the antenna.
When the SWR went awry for some reason (unclear to me) TX stopped to protect the device and it showed an error that I should reduce SWR.
I made 14 successful SSB QSOs using the device on a less than optimal antenna.
But mostly I experienced some of the downsides people mentioned here:
Bad modulation on TX - at least that’s what I was told and to a audio recording to prove it - could really be better.
My attempt to calibrate the frequency apparently failed. My first contact re-spotted me on the frequency they actually heard me. (To be fair, that’s mostly a user error…)
Noise level was to high. I could hear some stations but could not make out what they were saying over the high noise floor.
All in all it works about as well as I expected it to, considering the low price. But I’m definitely not going to sell my KX2 to replace it with this rig.
Ive yet to take out mine out in the field. But from what ive seen and people on qsos told me its not too bad for a device of this type.
From what ive seen my device is working quite alright. Some people reported not so good efficiency fron the “approved” seller. But that seem to be the batches after i recieved mine,im usually around 80-90%.
That said,i i didnt and do not expect too much for a device that costs 1/12th of a IC-705 so if im able to make okay activations thats great with me.
I can chuck it in a small bag into my backpack and “yomp” onto a summit and set up.
Ive bought the DL1DN battery compartment that screws onto the device and takes 3 14500.
All in all,im excited to take it out.
Edit: as for the frequency calibration,ive calibrate mine using the reverse becon website.
My complete Tr(uSDX) station weighs in at 0.8lb while my KX2 station is 2lb. While it has its flaws, the Tr(uSDX) will be great as a “take along” for running, biking, etc where I might not want to risk my KX2 (which is currently unobtainium if I needed to replace it).
It would be very interesting to look at the in band distortion on SSB transmit. I suspect that it will be absolutely awful with all sorts of sidebands. Anyone had one on a spectrum analyser?
Some people seem to using theirs with linear amplifiers which seems fraught with problems…
The way that the truSDX generates and amplifies its SSB is highly non-linear and creates all sorts of low level unwanted products. That’s why it sounds so bad if you listen to it locally.
I had a look at the schematics and it appears to be using the ATmega328 (the processor in the Arduino Uno and Nano) to run the SDR for both receive and transmit. For transmit the PA is class E which is highly non-linear. I’m surprised it works at all and not at all surprised that there are issues on both transmit and receive.
I tried using an Arduino a couple of years ago as an SDR for receive and came to the conclusion that it didn’t have enough CPU speed.
