You can use an on-the-fly logit program built into Lighthouse to calculate utilities for each item in your MaxDiff experiment. Then you can use the function MaxDiffRankAttValue(ExerciseName, Rank#) to build a constructed list of ranks 1, 2 and 3 (you can find more information about this function in the Lighthouse help menu). It's a little tricky to use so make sure to test it to make sure it's functioning properly.
The on-the-fly logit routine would likes more than the usual 3 observations at the respondent level (we usually recommend at least 4-5 views per item) but it looks like you may have to get by with just 3 - not ideal, but something I've had to do myself from time to time.
Funny you should ask - I was doing a similar thing just as your question came in.