Solar simulations and observations show that the detection of long-period Earth-like planets is expected to be very difficult with radial velocity techniques in the solar case because of activity. The inhibition of the convective blueshift in active regions (which is then dominating the signal) is expected to decrease toward lower mass stars, which would provide more suitable conditions. In this paper we build synthetic time series to be able to precisely estimate the effects of activity on exoplanet detectability for stars with a wide range of spectral type (F6-K4) and activity levels (old main-sequence stars). We simulated a very large number of realistic time series of radial velocity, chromospheric emission, photometry, and astrometry. We built a coherent grid of stellar parameters that covers a wide range in the (B-V, LogRHK) space based on our current knowledge of stellar activity, to be able to produce these time series. We describe the model and assumptions in detail. We present first results on chromospheric emission. We find the average LogRHK to correspond well to the target values that are expected from the model, and observe a strong effect of inclination on the average LogRHK (over time) and its long-term amplitude. This very large set of synthetic time series offers many possibilities for future analysis, for example, for the parameter effect, correction method, and detection limits of exoplanets.