Revisiting the Friedberg—Lee—Sirlin soliton model