there are many, and the usage depends a lot on what do you want to do with it, and how big is your dataset. Most on-line servers have a cap either in the number of sequences and/or in the lenght of the alignment. So in which stage are you?
I am going to assume that you have a bunch of protein sequences of this "novel protein" in different species. The first thing you have to do is align them... muscle works well http://www.ebi.ac.uk/Tools/msa/muscle/. You have to check that your alignent is good, both visualy and systematically. and trimm the alignment, so you dont introduce noise. I would reccomend BMGE for that, but is only command line, no online version. Also you can try it manually, although as a personal opinion, i think that this affects reproducibility. After that, you can use RAxML (you have to save the alignment in phylip format) http://embnet.vital-it.ch/raxml-bb/.
Again... this is very general and depends on the type and quantity of data.