We print one timecode per paragraph. Traditionally, a new paragraph begins when there is a speaker change or a pause in the text of a few seconds long.
So if your clip consists of one big “wall” of text (one paragraph), then the pdf export will only have one timecode.
You can create paragraphs yourself by pressing ENTER in the transcript. You won’t see these timecodes in the user interface, but the pdf export will have a timecode per paragraph.