Normally you need to tell your keyboard which voice you want it to use by sending it the Bank Select MSB, Bank Select LSB, and Program Change values that select the voice you want.
Some DAWs let you use a patch list file or instrument definition file that contains a list of the instrument's voices, along with the Bank Select MSB, Bank Select LSB, and Program Change values for each voice, so you can select the voice you want by its name.
But even if your DAW doesn't let you use those files, or if a ready-made file isn't available for your keyboard, you can still select the voice you want by specifying its Bank Select MSB, Bank Select LSB, and Program Change values.
Now, if you've set up MIDI output on your keyboard, the keyboard will output the Bank Select, Program Change, and other voice-related parameter values (such as Channel Volume, Reverb Send Level, Chorus Send Level, Filter Cutoff, Filter Resonance, and so forth) whenever you select the voice you want to use. If you're recording your performance into Logic (or other DAW) but the selected voice isn't being picked up, this is probably because the keyboard output all of those voice-related settings before you started to record.
There are three ways to resolve that.
(1) Start recording in the DAW before you select the desired voice on the keyboard. Sometimes this doesn't output all of the voice-related settings, because if some of the settings for the new voice happen to be the same as the settings for the old voice, the keyboard might not output those settings when you select the new voice. Also, if you start recording and then start making a lot of changes to your keyboard's settings, the DAW will record all of those changes, and not just the final settings.
(2) Your keyboard may have a "Send Initial Settings" function or something similar, which will output all of the voice-related and effects-related settings for the keyboard's current setup. This is a better option than the first, because you can be more confident that you'll capture all of the necessary settings-- not just the ones that changed when you selected the voice-- and that you record just the final settings and not a long series of interim changes. So you would set up the keyboard as desired, then start recording in the DAW, then execute the function that sends all of the settings for the keyboard's current setup, then start playing.
(3) The alternative is to manually enter the desired settings into the DAW. This can be a lot more tedious than sending the settings from the keyboard, but one advantage is that you don't end up recording what appears to be a lot if "silence" at the beginning of the track, which you might be tempted to edit out of the track but which has the MIDI messages for setting up the voice, so you don't want to chop out that seemingly "silent" beginning of the track. DAWs usually let you enter all of the preliminary settings for a track, which are saved in a sort of "header" for the track so they aren't "in" the track per se, which lets you chop off the beginning of the track if you want without losing those settings. It isn't that tedious to specify the Bank Select and Program Change this way; it's all of the other miscellaneous voice-related settings which can be tedious to enter manually. Sometimes the keyboard's "preset" voices will have default values for those miscellaneous settings, and if you simply send the Bank Select and Program Change then the keyboard will automatically use all of the default miscellaneous settings associated with that particular preset voice, such that you don't need to set those miscellaneous settings in the DAW unless you want to use settings which are different from the default values.
