Re: Test equipment recommendations
There's already a lot of good recommendations here, but thought I'd add my 2 cents.
I am in the fortunate position of having access to decent test equipment via my work when needed. However, I rarely actually take advantage of this, as for most hobby stuff, I can make due with the hobby grade stuff I have at home. One of the most important skills as a tinkerer is to learn how your tools work, so you can use them to their full potential, but perhaps more importantly, so that you know your tools' limitations and either work around them, or know when to buy better tools.
The Internet is an extremely valuable piece of "test equipment". If you look around here, or any of the other numerous radio forums, you'll find tons of examples of people describing basic symptoms, which others have already encountered, and can suggest fixes for, without ever using a physical diagnostic tool. However, the Internet by itself is also very limited in what it can test, so if you have an uncommon or non-obvious problem, it can't help much without additional test tools.
A cheap ($10-20USD) multimeter is by far the best testing per dollar investment. The one I have at home is known to be somewhat inaccurate, but has been good enough for diagnosing my hobby stuff for many years now. The continuity buzzer is probably the feature I use most anyway, to find shorts or opens where they shouldn't be. You can spend hundreds on a good multimeter, but do you REALLY need values accurate to 5 digits, or can you get by with 1-2 digits and a little error? You might need the accuracy if you're doing matching for a mixer, or precision resistor measurements, but most of the time I suspect not. The main thing to be careful with on the cheap end is high voltage or current, since many cheap models skimp on safety features like fuses. If you're working QRP or microcontroller electronics, this shouldn't be a big problem, but if you're working on high power rigs, then you'll want to make sure your multimeter has the proper fuses and isolation.
A SWR/power meter is useful for every-day operation. I have one built into my antenna matcher, and use it so much that I hardly think of it as test equipment, even though it obviously is. While an antenna analyzer can help you with SWR (discussed below), the forward power meter part lets you measure your actual output. Turns out SWR doesn't matter if your transmitter isn't actually outputting anything!
For general RF detection, an LED and a diode make for a super cheap indicator. Take a look at https://www.youtube.com/watch?v=V5SMF9p-4Q0 for example.
A dummy load is also piece test equipment I often take for granted. It's great for verifying transmitter stuff without actually broadcasting (appreciably), can be used in conjunction with a power meter to verify full power output without worrying about SWR, and if you have another radio, the leaked power on my dummies is often large enough to be detected by an adjacent receiver for further validation.
For radio use, it's a toss up for me if the next best investment is an oscilloscope, or a network/antenna analyzer. If you're planning on doing lots of antenna design or tweaking, then the analyzer would serve you better. If you're planning on spending more time working on rigs or other electronics, then the oscilloscope probably is the better choice. Of course, if you can afford both, then even better.
For oscilloscopes, as mentioned by others, the key is to have enough bandwidth for the signals you actually care about. The DSO138 you mentioned isn't necessarily a bad choice, as long as all of your signals of interest are in the audio frequencies, or are slow-ish digital (like Arduino stuff usually is). It's low price is attractive, but it's actually significantly more expensive in Hz per dollar than the "real" entry level oscilloscopes out there (200kHz/$10 for the cheapest DSO138 kit I see = 20kHz/$1, vs e.g. Hantek 2C42's 40MHz/$104 = 385kHz/$1), and is not at all competitive in performance. I have a XMEGA-based 200kHz (2MSPS) oscilloscope that I bought a number of years back, and it definitely has helped me solve problems that my multimeter just can't. Having even a 200kHz time domain can be incredibly useful, so if you have a tight budget, something like the DSO138 could be worth while. However, for analyzing RF, you'll need an oscilloscope that has a much larger analog bandwidth. The uBiTX IF goes up to 75MHz (officially, more if you ask the Si55351 to :P), so if you plan to work on the uBiTX's RF chain at all frequencies, you should consider buying a scope capable of ~75MHz analog bandwidth. As Robert mentioned, DSOs are sometimes advertised with their sampling frequency, rather than analog bandwidth. Shannon/Nyquist tells us we need at least 2x sampling for a given frequency, so you'd need an absolute minimum of 150MSPS to get 75MHz analog bandwidth measurements, but realistically you should aim for higher samples per second so that you can actually see the shape of the wave. Analog oscilloscopes don't have discrete samples, so as Ashhar pointed out, you can use them at higher frequencies - they'll just measure lower amplitudes than actually are present past their designated bandwidth. I would agree with Bill, and advise against purchasing anything on eBay that's not listed as being in working condition with pictures of sample traces, unless you're ready to spend time diagnosing and fixing it yourself, which will be particularly difficult if you don't already have a working oscilloscope. Randy's suggestion of looking at hamfests is a good one.
For the antenna analyzer, the NanoVNA you mentioned is a great tool for the money. It's capable of all the stuff the more expensive units are (e.g. MFJ's SWR-584C, or RigExpert offerings), but the NanoVNA has a bit steeper of a learning curve than those more expensive options too. K7TRF wrote up a nice short comparison recently over at https://forums.qrz.com/index.php?threads/antenna-analyzer-so-many-choices.689265/#post-5335185 . One thing to watch out for with the NanoVNA specifically is all of the clones. I'd strongly recommend buying directly from one of the NanoVNA's developer's links (such as hugen79 or flyoob) to ensure you get one that doesn't take any shortcuts, and directly benefits the parties responsible for the tool development.
Once you have most of the equipment above, an SDR would probably be my next recommendation. The normal RTL dongles won't work at HF (though you can find some RTL dongles with HF upconverters built in), but there are a good number of options in the ~$100USD range. There are tons of things you can do with SDR, but as a test tool specifically for the uBiTX, I've found it most useful for calibration of the oscillator, and verification of signal output in various modes. They can also work as a windowed spectrum analyzer, useful for finding unwanted harmonics or spurs, or measure relative power output.