I have produced a sawtooth wave using a combination or even harmonics, but have been asked to provide a better representation of the wave using a simpler method. So by terms of simpler method it rules out Fourier analysis. What are you meaning by time domain or frequency domain representation?
In my previous post, I gave the exact equation for it. Are you instead desiring the Fourier expansion of a sawtooth wave? You can look that up in many math handbooks, such as page 455 of the CRC Standard Mathematical Tables, 26th edition.
You need to phrase your question in a more detailed form to get an appropriate answer.