I have had a lot of problems of my own with poorly diffracting (or not at all) membrane protein crystals. After a previous discussion here, I summarised the suggestions I got in this ccp4 user wiki;
http://strucbio.biologie.uni-konstanz.de/ccp4wiki/index.php/Improving_crystal_quality There are a few ideas there that have already been put forward. My personal opinion is that (assuming your crystals are protein) detergent is the deciding factor. I second the suggestion for shorter chain detergents, but also suggest that you carefully consider your detergent concentration. While you want it to be above cmc, I have found excessive concentration to be bad for crystallisation. If you are going back to the drawing board, I can highly recommend the MemGold and MemStart/Sys screens developed by the lab of So Iwata at the Imperial College, and sold by Molecular Dimensions (in the UK). It has given me a lot of success in getting initial hits for various membrane proteins. MemGold has been designed to specifically target alpha-helical membrane proteins, as described in this paper; http://onlinelibrary.wiley.com/doi/10.1110/ps.073263108/full Hope you find something useful here. Good luck. Damon.
