Building the requirement of your experiment

The number of participants is important to have clean averaged ERP data

The SNR is proportional with sqrt(Ntrials) 10 trials instead of one trial means 3x better SNR 100 trials means 10x better SNR than a single trial