(Actually, I'm not sure this is the real answer; I can revoke it as soon as someone shows me it's incorrect).
As suggested by @uhoh, we took a spectra of an halogen lamp, and as we also got an old incandescent lamp, we took one of it too:
As you can see, both lamps show a "bump" on the yellow part of the spectrum (more the tungten -old- lamp than the halogen). That seems to indicate that the diffraction grating is responsible for the "bump".
But, just to double-check, we also took two more spectra of the sun, in different conditions:
This made me realized that I didn't fully stated the conditions for taking the original spectra: as we weren't comfortable pointing the camera directly to the sun, in the first spectra we directed it to a cloud (it was a cloudy day). On another day, we took two more spectra: one pointing to the blue sky (it was a clear day), and another to a white painted screen reflecting the sun's rays. As you can see, the yellow "bump" is more prominent in the spectra taken pointing to the cloud!
So we thought that may be the water vapor has something to do with it, besides the diffraction grating.
Doing a fast internet search, we came across this paper: Operational parameters of thermal water vapor plasma torch and diagnostics of generated plasma jet, with this figure:
But the authors attribute the observed yellow bump to "noise". So it does not seem to explain our bump. :)
So, up to now, our preliminary (and certainly incomplete) conclusion to this topic question is that the yellow bump is due to our diffraction grating and the clouds we were pointing to that day.
Let me credit the other members of the team: Facundo Balerdi, Fabián Faingerch, Sofía Habu, Mariano Solís and Clara Telesca.