The following is explicitly a synopsis of the CPR only to provide a background and give basic insight necessary for understanding the answer to the question. Some points are slightly modified or amended in later works.
On the transcendental method in the CPR
Kant goes through our three faculties of knowledge (Sensuousness, Understanding, Reason) in the Transcendental Aesthetics, Analytics and Dialectics. In every faculty he is searching for transcendental a priori, which means nothing more and nothing less than the necessary conditions of the possibility of (this aspect of) objects of experience (see Prolegomena, 4:473f. fn.). Insofar it is possible, the a priori insights are established as real beyond reasonable doubt in the Transcendental Deduction (for more on how a priori knowledge is established in Kant, see this answer of mine).
On the different a priori for the three faculties of knowledge
For the faculty of Sensuousness, the faculty through which we acquire content, he says that the content needs to be ordered in the Space and Time in order for us to be presented with different (space), persistent (time) objects of experience at all (as opposed to a single, multimodal manifold). Thus, space and time are the transcendental forms of Sensuousness.
For the faculty of Understanding, the faculty that subsumes content under concepts, he says that there are fundamental conceptual categories that allow us to conceptualise objects and their relationships in the first place. These categories are, therefore, transcendental concepts.
For the faculty Reason, the faculty of inference from what is to consequences and conditions, he says that there are ultimate conditions of the experience of objects itself. Those are, according to him Soul, Freedom, God: The soul as ultimate condition of the unity of subjective experience (the self), Freedom as ultimate condition of the unity of objective experience (the world that includes all causal conditions of objects from its very beginning), and God as ultimate condition of the unity of being and thinking itself (harmony of subjective and objective experience). Since they are projections of our thinking that are thought to be the ultimate conditions, he calls them transcendental ideas.
On the commonalities, differences, and relations between the transcendental forms, concepts, and ideas
Transcendental a priori are all of them in the sense that they are necessary conditions of our relationality to objects of experience (question from his letter to Herz from 1773). They are not thought nor known "before" any experience in a sense of time, but in a sense of inference: Without them, our experience (or thought) would not be possible. It is a common, yet mistaken interpretation to say that we are born with an inventory of a priori knowledge. Otherwise, it would not have taken some thousand years to work them out, right?
Only space, time, and the categories do have corresponding objects of experience (i.e. every sensual intuition) and are therefore knowledge a priori proper. The transcendental ideas are presented as necessary end points of the search for end points in inference chains but as they are outside of experience proper and not about the forming of objects of experience, they are only problematic and not knowledge. The highest level of "confirmation" for them we get is that they are able to be thought without contradiction (definition A254|B310). Note that he changes that in his second Critique at least pragmatically, ie. establishes them as real in the sense of a real aspect of human action: they are realised in action but still not known in the same sense as objects of experience as we got no way to perceive them.
Regarding the relation between space and time on the one hand and the categories on the other hand there is another aspect already established in his dissertation and repeated in the Transcendental Deduction: Time and space as forms of intuition are used for the conclusion that sensual intuitions are appearances (in the sense of a way to represent the objects that are sensed) since we do know that they are aspects of our way of representing things but we cannot know inhowfar this representational mode has correspondence in the represented objects without being represented in a conscious mind.
What is new in CPR is that Categories are the most abstract concepts of relations between objects of sensual experience (which can only be given via intuitions!) and thus can only be applied to intuitions. Therefore, the form of space and time, as forms of intuition are "logically prior" to the categories as well. How could we differentiate unity and multiplicity without differences in temporal and spatial loci, which are necessary to have a definite object of representation in the first place, after all? In the same sense, the transcendental ideas are at the end of logical priority: only given we have established different definite empirical objects and the relations between them, we can start to bother about the grand picture and whether their relations point us to some common, ultimate condition.
I suggest reading Eckart Förster's The 25 Years of Philosophy for a quite good presentation of the overarching argument.