Be prepared for some noisy-heavy signals. I haven't experimented around with enough different acquisition methods to recommend a particular device, but each one is going to have their own challenges (either $ getting them or some negative signal quality issue).
Once you have a system there should be some basic demos with regards to observing a few EEG bandpower based estimates. Beyond that I would recommend looking at SSEVP (steady state evoked visual potential) as all of the conference papers/talks/posters/etc about that topic make it seem reasonably approachable compared to other tasks, being EEG (rather than LFP) friendly (IIRC), and it makes for some cool applications.
source - I spent a few years researching EEG/LFP recordings in the context of speech and sleep (separately)
Once you have a system there should be some basic demos with regards to observing a few EEG bandpower based estimates. Beyond that I would recommend looking at SSEVP (steady state evoked visual potential) as all of the conference papers/talks/posters/etc about that topic make it seem reasonably approachable compared to other tasks, being EEG (rather than LFP) friendly (IIRC), and it makes for some cool applications.
source - I spent a few years researching EEG/LFP recordings in the context of speech and sleep (separately)