You can check several other options. 

some ion channels fall apart very fast right after SEC. Max time is 2-3 hrs or even less. depending on your experiment, purify in small amounts and use them right away. I have some mutants of ion channel that crashes very soon and I used to set up my crystals in a hour.

Main problem is the detergent. which family does your ion channel belongs to. Since you having considerable amount of aggregation in SEC, you protein is not happy in the detergent which shows later. did you screen for other detergents. DM, UDM or OG. try a detergent screening and figure out the optimal one.

I am not sure if cleaving of his-tag is the reason for the protein to crash. I used to work with MBP tag which stabilizes the protein and cleavage of the tag tends to aggregate the protein but his-tag is just few residues and not sure if it is really stabilizing the protein.

Minor things to check is the temperature. Do you perform your experiments  and store the protein at 4 C? Try Gel filtration buffer at pH 7.4. 

Thank you for your input. I am currently working on a magnesium sensitive ion channel. Previous characterization of this protein has been optimized to work in DDM. I have double checked the buffer pH to be at 7.2. His-tag is not being cut in this case. I would also like to add two points to this. Firstly, the precipitation is accelerating in presence of the cognate ion and secondly the protein seems to phase out to a certain extent after that the remaining protein is stable.

You may be over-concentrating the protein beyond the ability of the detergent to keep it in solution. Either lower the protein concentration or use a higher detergent concentration.

GOOD AFTERNOON DEAR. if you have to make protein stable in the maximum concentration may you use the washing solution in the high concentrations.