In the real time code I use for numerical time propagation, usually we use laser pulses with a Gaussian envelop and a central frequency. To obtain a pulse with a broad spectral range (few eV) very short laser pulses need to be selected (sub fs).
I'm looking for a laser pulse with a broad spectral range (again few eV) which is also longer than 10 fs. One option would be to use chirped pulses. Are there alternatives, i.e. non chirped pulses?
As the transform limited pulse duration is set by the bandwidth, and your requirement is of a bandwidth that has a transform limit below 10 fs, then there is no other choice other than chirping the pulses to obtain longer pulses.
As you specified bandwidth in energy units, central frequency does not matter for the transform limit pulse duration.
As explained in the above answer, without chirping the pulse, the bandwidth $\Delta \omega$ and the time duration of the pulse $\Delta t$ are forced to have fixed product. $\Delta \omega \Delta t = \alpha $. However the value of such $ \alpha $ depends on the pulse shape. See for example here: Book Ultrafast Optics, Chapter 2, Section 8
Reported values of $\alpha$ do not change much, however this probably depends also on how $\Delta \omega$ and $\Delta t$ are defined. However, for my original problem, i.e. numerical simulations, moving from a Gaussian (in time) envelop to a rectangular (in time) envelop has a huge impact since the frequency window of the Gaussian envelop goes down exponentially, while the one of the rectangular envelop goes down as $1/\omega$