Has anyone checked for voltage leakage? The Feeltech units have had well documented ungrounded connections.
Leakage from where to where?
The unit takes external 5V DC supply, so it is fully isolated from mains.
Only by as much as the 1nF Y class capacitor (typically used in all 5 to 10 W rated EMC compliant class II smpsu based wallwarts) keeps it isolated from half live mains voltage.
If you run it off a power bank from its USB output charging port, that'll give you the isolation you crave, otherwise forget it when using a cheap universal mains voltage wallwart with its half mains live 'touch voltage (1.6M ohm impedance at 50Hz connection to half of whatever your local supply voltage happens to be.
The only virtue to using an external plug in 5vdc power source is that you can at least elect to use an earth protected 5v mains psu alternative to the now ubiquitous cheap commodity wallwart. The only issue with such earthed supplies is that the common ground return may also be connected to the protective earth pin which introduces unwanted mains earth loop noise and interference to any DUTs.
A neat way to achieve the superb fix to this issue on the FY6600 (and its successors) which killed off this unwanted touch voltage without committing the sin of introducing the mains earth loop issue by the Feeltech idiots' spiteful fix to all the complaints that had been made by their FY6600 owners over this ESD hazard to DUTs is to obtain a suitable stout plastic project box, fit it with an IEC C14 mains socket, crack open a cheap wallwart to transplant its guts into said box and wire a 4k7 half watt or so resistor between the zero volt terminal on the DC output and the protective earth pin of the C14 socket.
You can also embellish this side project with a mains indicator lamp (LED and a resistor across the 5v rail) and fuse according to taste but whatever the case, you will now be the proud possessor of a supply with less than 100mVAC leakage on its output with respect to ground on 240vac supplies without that troublesome low impedance ground loop to allow the free flow of unwanted interference (including unwanted random DC offsets) in either direction - that 4k7 resistor will attenuate such interference by some 40 to 60 dB or so.
For those of a nervous disposition when it comes to cracking cheap 5v 1A wallwarts open to extract the smpsu board and wire it into a project case as described, there is a more disposable way to use the project box by adding a C14 socket as before along with a mains outlet socket and a DC jack socket (or three to widen your options) to plug in whatever 5v wallwart comes to hand for the job (swift swap out if the wallwart dies unexpectedly). You can either use another DC jack or else a suitably plug ended captive DC lead to match the DC jack of your cheap FG - your choice. The final task after wiring the C14 to the mains outlet socket is to connect that 4k7 resistor between the C14's PE tag and the zero volt connection of the DC output on the DC jack socket(s).
The only downside to this solution is that the smpsu board remains confined by its original unventilated plastic housing (unventilated to minimise the risk of accidental liquid ingress). Plastic is famous for its poor thermal conductivity so that smpsu pcb has to tolerate very high temperature operation, high efficiency notwithstanding.
Extracting the wallwart's cct board as per the first solution does offer the benefit of much cooler operation simply by virtue of the much larger surface area of the plastic project box used. An even cooler solution is to use a metal project box but you need to take more care in mounting the PCB and properly connecting the metal case to the C14's PE tag whist making sure the only connection between this earth and the zero volt DC output connection is via the 4k7 resistor.
Anyway, there you have it. A few suggestions on how to eliminate the unwanted half live mains voltage ESD hazard to your DUTs when using such cheap "Toy FGs"
JBG