Title: Towards early dark energy and $n_s$=1 with Planck, ACT and SPT

Abstract: We investigate the constraints on early dark energy (EDE) by combining the most recent CMB observations available, ACT DR4, SPT-3G, and Planck2018 ($\ell_\text{TT,max}=1000$) data. This combined CMB dataset favors non-zero EDE fractions and large Hubble constants. The inclusion of BAO+Pantheon data has little effect on the results, leads to $H_0=71.6( 72.9)_{-1.5}^{+2.0}$ and $73.17(72.74)^{+0.55}_{-0.77}$ km/s/Mpc for axion-like EDE and AdS-EDE, respectively. The axion-like EDE can fit the data significantly better ($\Delta \chi^2 \lesssim -10$) than $\Lambda$CDM, which is mainly driven by the ACT data. It is found again that if the current $H_0$ measured locally is correct, complete resolution of the Hubble tension seems to be pointing towards a scale invariant Harrison-Zeldovich spectrum of primordial scalar perturbation, i.e. $n_s=1$ for $H_0\sim 73$ km/s/Mpc.