The paper is devoted to quadratic Poisson structures compatible with the canonical linear Poisson structures on trivial 1-dimensional central extensions of semisimple Lie algebras. In particular, we develop the general theory of such structures and study related families of functions in involution. We also show that there exists a 10-parametric family of quadratic Poisson structures on gl(3) * compatible with the canonical linear Poisson structure and containing the 3-parametric family of quadratic bivectors recently introduced by Vladimir Sokolov. The involutive family of polynomial functions related to the corresponding Poisson pencils contains the hamiltonian of the polynomial form of the elliptic Calogero-Moser system.1 We will use the term "polyvector" instead of "polyvector field" for short, with the exception of vector fields.By linear, quadratic etc. polyvectors on a vector space we mean polyvectors with homogeneous linear, quadratic etc. coefficients with respect to some linear system of coordinates.We will say that a Poisson pencil is linear-quadratic if its generators are linear and quadratic bivectors on some vector space respectively. The cases of "linear-constant" and "linear-linear" Poisson pencils are already very important, see for instance [1], [20] and references therein. The linear-quadratic Poisson pencils are the main objects of this paper.There are basically two classes of linear-quadratic Poisson pencils known in the literature. The first one can be described as follows ( [4], [9], [10]). Let r ∈ g ∧ g be a skew-symmetric solution to the classical Yang-Baxter equation on a Lie algebra g, i.e. an "algebraic Poisson bivector" on g. Assume that a linear representation of g in a vector space V is given and ξ 1 , . . . , ξ n are the fundamental vector fields of this representation corresponding to a basis e 1 , . . . , e n of g. Then, if r = r ij e i ∧ e j , the bivector π 2 := r ij ξ i ∧ ξ j is a quadratic Poisson bivector on V . If, moreover, another bivector π 1 on V is given, invariant under the action of g, the bivectors π 1 and π 2 are compatible. In particular, if V = g * with the canonical Lie-Poisson bivector π 1 , and the representation is the coadjoint one, we get a linear-quadratic Poisson pair (π 1 , π 2 ), where π 2 = r ij π 1 (x i ) ∧ π 1 (x j ) (here x i are the corresponding coordinates on g * ).The second class concerns a specific situation when a quadratic Poisson bivector π 2 is given on a vector space V and there exists a point a ∈ V , a = 0, such that π 2 (a) = 0 (not for every quadratic Poisson bivector such points exist). Then the linearization of π 2 at a is a linear Poisson bivector which is automatically compatible with π 2 .One can find numerous examples of such a situation in the literature. One class of examples is related to Poisson-Lie groups. Given such a group (G, π), its Poisson tensor π necessarily vanishes at the neutral element e ∈ G and there is a correctly defined linearization π 1 of π defined on the linear space T e G (it coincides with the r-matrix bra...