There is a variety of possibilities including, but not limited to:

• It's a focus issue. This can be verified by testing a different species with otherwise similar properties, i.e. also something ball-milled in a solvent.
• If you don't get good Raman peaks even with formally good scatterers, check if your optics and fibers are OK in the first place.
• Have you checked for fluorescence in the background? This may easily overpower the Raman signal and may indicate the necessity to use a different laser wavelength.
• If you get good peaks with unprocessed graphite, but only your particular material doesn't give rise to these, it may simply be that there is something wrong about the process and the resulting material is no longer graphitic enough in order to give rise to the wanted peaks.

Hello,

Check this article where variations in Raman spectra were discussed:

Article The Electrochemical and Structural Changes of Phosphorus-Dop...

Raman spectra is quite sensitive to the crystallite size of the sample. Ball milling, especially in ethanol, will surely reduce this significantly, it can separate the graphite layers and can even break them apart. This structural disruption broadens the Raman peaks, sometimes to the point where they’re barely visible or completely lost. So what you're seeing makes sense.

you may wanna do a partial ball milling in small time to see if you can spot the tiny variations.

What do you mean by good peaks?

You can try to do Raman of the Graphite sample at first and then proceed with the dispersed sample and compare the results. For the ball milled samples, give a heat treatment of 50-55 degree celsius prior to any characterization.