To be honest, I would be quite unhappy if a plot of DiracDelta
would actually show anything! What you can do is making something visible by showing its convolution with some "sharp" function, e.g.:
f[t_] := a Sin[t]
g[\[FormalK]_] = FourierTransform[f[\[FormalT]], \[FormalT], \[FormalK]];
plotDelta[f_, k_] := Total@*ReIm@Convolve[f[\[FormalX]], HeavisideLambda[50 \[FormalX]], \[FormalX], k]
Plot[Evaluate[plotDelta[g, k] /. a -> 1], {k, -2, 2}, PlotRange -> All]
which gives:
