Abstract
The locations of hot pulsating variables in the H–R diagram are found using the effective temperatures derived from spectroscopic analysis and luminosities from Gaia parallaxes. Frequency peaks extracted from TESS photometry were used to compare with model predictions. A large number of stars with pulsation frequencies similar to δ Scuti variables were found between the predicted δ Scuti and β Cephei instability regions, contrary to the models. These Maia variables cannot be explained by rapid rotation. There is a serious mismatch between the observed and predicted frequencies for stars within the known δ Scuti instability strip. In δ Scuti and Maia stars, the frequency at the maximum amplitude as a function of the effective temperature was found to have a surprisingly well-defined upper envelope. The majority of γ Doradus stars were found within the δ Scuti instability strip. This is difficult to understand unless pulsational driving is non-linear. Non-linearity may also explain the huge variety in frequency patterns and the presence of low frequencies in hot δ Scuti stars. γ Doradus stars were found all along the main sequence and into the B-star region, where they merged with SPB variables. There seemed to be no distinct instability regions in the H–R diagram. It was concluded that current models do not offer a satisfactory description of observations.