Setting a diagonal covariance matrix for random effects

Hi,
I think the syntax 1 || participant, which avoids estimating the correlations (i.e. forces a diagonal covariance matrix) is what you need.

Does that resolve your issue?