Design of proteins de novo allows access to diverse, unexplored areas of sequence space. Given the enormity of sequence space, a directed approach is expected to yield a higher fraction of folded and functional proteins than a stochastic sampling of random sequences. One such approach is the design of combinatorial libraries by binary patterning of hydrophobic and hydrophilic amino acids. This approach utilizes hydrophobic collapse to bias protein folding towards specific secondary and tertiary structures. We have previously investigated the possibilities of binary patterning by building a library of de novo four-helix bundles. The structure of the most stable protein from the library was solved previously and proved to be consistent with design. However, solving one structure does not fully assess the overall success of the binary patterning strategy, nor does it account for differences in the stabilities of individual proteins. To more fully probe the quality of the binary patterned library, we solved the structure of a second protein, S836, by NMR. Protein S836 proved to be a four-helix bundle consistent with the design. The high similarity of structural features between the two solved structures reinforces previously published experimental evidence that proteins in the library are stable, monomeric, four-helix bundles. Despite their structural similarity and high sequence identity, the two proteins have hydrophobic cores that are packed at different degrees of tightness. This may account for the difference in their stabilities. The relationship between interior packing and protein stability was probed by Modelfree dynamical analysis of both proteins. This analysis showed high frequency of chemical exchange coinciding with less well-packed hydrophobics. We conclude that although the binary patterning may be sufficient to drive protein folding, the overall stability of the fold is moderated by the identity and packing of specific residues in the hydrophobic core.