A scan and a vector PDF follow the same route
A PDF scan stores the appearance of a paper drawing, while another PDF may contain shapes and text created by design software. Either can present a useful page for this workflow.
The page is rendered to an image before conversion. The DXF is an interpretation of that appearance, not direct extraction of the PDF vector objects or a recovered original CAD document.
One page is one starting drawing
Choose the PDF page that fits the task. If several drawings share a sheet, consider preparing a more focused page before uploading rather than expecting the conversion to know which view you intended.
There is no batch conversion or built-in crop tool in this flow. Review the selected PDF preview, then submit one page to create a DXF you can inspect and develop in CAD.
Document dimensions do not establish project scale
The physical size of a PDF page is not the same thing as the dimensions of the subject printed on it. A plan may have been resized or placed at a particular drawing scale.
Use a known length to size the DXF, then check other important distances against reliable information. A PDF dimension label is a reference to verify, not an automatically applied CAD constraint.
Readable source pages make review easier
A PDF with faint lines, heavy stamps, or very small details can be difficult to interpret. Choose the clearest available source and inspect the page at a useful zoom before continuing.
Page rendering is bounded for browser performance, so extremely large sheets may need preparation outside this tool. Keep the original PDF to compare with the DXF and rebuild ambiguous details using better references.