Pseudo genes, as paralogues (defined on protein sequence similarity), are not necessarily a problem, only the level of DNA sequence identity is important. Do a full DNA-DNA alignment of the coding sequences and try to design primers using the sequences that are divergent enough, by specifying those unique regions as input to your primer designer. Do an in-silico PCR with the designed primers to detect other non-specific sites.

In eprimer3, parameters

-targetregion, -excludedregion, -includedregion can be used to control this.

The -mispriminglibraryfile parameter could be used to input sequences to avoid for amplification e.g. the coding sequence of the pseudogene.

Use -explain parameter to get an explanation in case no primer pair can be found.