When you say the "other tool" are you meaning MJ or the launch window planner? Assuming the latter: You do have enough info to use the planner. It requires only origin and destination for inputs. Once you have the resultant info create a manuever node faithful to those numbers. It seems from your post you are ok up till here. If your manuever node isn't exactly like the planner (ejection angle, speed etc) then you will get different results. Is this where you're having problems? Assuming yes: If your results DV and time-wise are wildly different from the math than the issue is execution. If you feel you are following the math exactly and are still not getting the correct trajectory after burn, then run two of them. Let MJ plot and burn one, then plot and burn one yourself (the exact one MJ just did) and see what comes of it. Something to keep in mind is the planner is a best case scenario. There may be many times when it's results are inconvenient, especially if LS is an issue. In this case you can experiment with different departure times and angles, none of which are likely to be more efficient (dv) but may be much faster; i.e rather than ninety days to Duna for 2500k, seventy days for 3500k.