It is nice gadget, and somehow I still like it (particular due to partability and battery operation), but calling it "excellent" is IMO exaggerated.
Features are just elementary, but yes, it does display waveforms sufficiently for a number of simple use cases.
Frontend bandwidth is ideed more than sufficient. Below is a list of things I do not like so much:
- UI experience is rather inconvenient. I'd also favor a touchscreen. Operating a NanoVNA or TinySA is much more pleasant.
- Data transfer to the PC app, and consequently the screen update in the PC app, is horribly slow.
- Frontent is a pretty noisy, in all V/div ranges. No wonder, though, given that all input ranges are attenuated to 10mV/div first, and then get re-amplifed.
- A range smaller than 10mV/div were nice, too. But I'm unsure whether it would make sense, given the noisiness.
- There is almost no memory depth.
- The lack of memory also prevents a trace averaging feature in order to reduce noise.
- Input capacitance changes, depending on the input attenuator relay position (1:1 vs 50:1). This impacts adjustment of 10:1 probes. But the same seems to apply to various other low-cost scopes, too.
- There is a significant feed-through from the AWG under load (i.e. when a load is connected to the AWG) to the scope frontend. Maximum feed-through happens at ~5kHz. 1)
I did track the path, and obviously it happens via the negative power supply rail. The rail is shared between AWG and scope frontend, and the PSRR of the frontend is obviously not very good - just roughly 10dB, IIRC. - In two-channel mode, the two channels are not sampled simultaneously, but with a time offset of 1/2 ADC clock cycle (which depends on selected time base). 2)
I.e. if you feed the same signal into both channels, you oberve a phase difference, which is not present at the inputs. Also well visible in x-y mode, when an expected straight line turns into an ellipse. - The AWG cannot reach the full specified output voltage swing w/o clipping (at least under load). Root cause is that the negative supply voltage for the output opamp is too low (and the opamp is not rail-to-rail either).
- When I got mine, the AWG did produce significant spikes in the waveform. This was IMO a timing error in the FPGA (DAC's setup times from data to clock edge not met).
A FPGA update from hantek fixed it mostly, but I still see sporadic fringes in the waveform -- it still does not look as clean as from other AWGs. - The AWG does of course not have any analog attenuators, so the output amplitude is only controlled digitally (-> the lower the amplitude, the lower the relative resolution).
1) See first attached image. AWG generates a 5.5kHz signal, and drives a 50 Ohm load. Scope inputs are shorted with 50 Ohm terminators (in order not to pick-up noise).
Still the scope channels display the fed-through signal fom the AWG (with an aplitude equivalent to a 500mVpp input signal, at 500mV/div)
2) The next three images demonstrate the 1/2 ADC clock cycle time offset between the two channels.