I have tried to save size as much as money on a compact $40.00 generic bench PSU. It died within a couple of months. The voltage starting to drift way high when it heated up.
I spent a little more on a Korad, which Dave does a review of. It has been doing the job for a couple of years, now, and I'm fully satisfied.
If you want to have a switched variable output, use a 3v3 regulator and then lift the ground pin voltage with diodes. For example a 3v3 regulator with a reverse biased 4v7 zener between the ground pin and the 0v rail will give you 3v3+4v7 = 8v out.
I'm going to teach you something so you never have to do this terrible hack, again, lol.
1. a 3.3V regulator can't go below 3.3V
2. an adjustable linear regulator such as an LM317, et al, usually goes down to about 1.25V.
3. even if a fixed linear regulator is all you have, adding diodes to it doesn't make it "adjustable." It means you "adjusted it." And considering the "resolution" of available diodes, you can hardly call it that, even.
4. If you want to increase the voltage output of a fixed linear regulator, all you have to do it wire it up like an LM317, using the ground pin as the "ADJ" pin. And in the formula, replace "1.25V" with w/e is the voltage of your fixed linear regulator. Just be sure to know that if the device package has a heatsink, this heatsink is now going to be at Vout minus the initial voltage rating... which is a stupid number (and same as result of your diode trick). But using potentiometer, at least this way your output is actually adjustable. Please, I want to hear no more of this terrible diode hack that I have heard of for at least 10 years.
5. If you want to make multiple fixed setting to switch between, again, this is way more tunable by switching in different resistors for R2... this is not a reason to use a 3.3V fixed regulator over an adjustable regulator. The diode hack will work just as badly on an adjustable regulator.
The only thing "special" about an adjustable linear regulator is that
1. the "fixed" regulated voltage is the minimum possible to increase the range of possible output.
2. the "ground" pin is called an "ADJ" pin. (In a linear regulator, the voltage is dropped in series... the "ground" pin is just a feedback pin. The regulator keeps the voltage between output pin and "ground" pin at the "fixed" voltage - in this case it's ~1.25V.)
3. the heatsink is in continuity with the Vout, rather than the "ground," since you're going to be raising this "ground" node above ground in most applications. And having heatsink at Vout is a lot more sensible than Vout minus 1.25V.