non expected discrepancy between CMB curves

[ivan rodriguez-montoya]

Hi everybody

Default parameters in CAMB include 'massive_neutrinos = 0' and consequently 'omnuh2 = 0' from which CMB spectrum fits pretty well WMAP data. you can check my params.ini

I just changed 'massive_neutrinos = 3', but mantain 'omnuh2 = 0', which should mean effectively massless neutrinos. I was expecting to obtain a very similar CMB spectrum, but am surprised to find that the change is quite significative. You can see it neu.eps. It has nothing to do with 'massive_nu_approx' parameter

Why do they differ that much?

Maybe naively, this could seem to be an inconsistency. I would appreciate your comments. Thanks

[Antony Lewis]

What did you set for massless_neutrinos ?