Censoring a truncated distribution

As silly as it may be that means you can use a StudentT with large fixed nu, to get what you want. That will behaves basically like a Normal, except the likelihood could be a bit slower.