Timariot cavalrymen had to pay for their own costs of going to war and levy a certain number of troops depending on the size of their holding. This was determined with a regular survey the imperial government conducted until the early seventeenth century. Unlike the feudal system, the timar holders did not have a title of nobility per se and did not have the bargaining leverage that a European lord might have over the monarch. Since the timars were non-hereditary as you mentioned, they could be revoked and given to a new person. The timariot army consisted of the numerical majority of Ottoman troops during the fifteenth-sixteenth century and supplemented the already sizable kapikullu (standing) troops. The main difference between the timariot and feudal systems was the lack of hereditary title. The land was owned by the Sultan, not the timariot cavalryman. Hypothetically a holder could resist going to war, but they would lose their privileges.

The timar system was also why the Ottomans were able to field armies of tens of thousands and broke 100,000 men in some cases because the cost burden was outsourced to the timar holders. For comparison, European rulers like Charles V had to take out loans to finance mercenary armies and try to get their estates to levy men for campaigns.

There were some limitations to this system as time went on. It was workable insofar as the empire kept adding new territory it could award to new timar holders, but conquest slowed down after Suleiman and few timars were granted by the middle seventeenth century. This was also around the time that the regular land survey ended and the imperial finance system shifted toward tax farming. To make up for the reduced mobilization, the provincial governors in the seventeenth century started to recruit private mercenary forces.

----

Rhoads Murphey, Ottoman Warfare 1500-1700

Baki Tezcan, The Second Ottoman Empire: Political and Social Transformation in the Early Modern World