LKAB produce several different highly refined iron ore products, fines and pellets, for the manufacturing of iron and industrial minerals applications. The different industrial minerals products have individual specifications and therefore different area of usage. At LKAB in Malmberget a problem come to attention when magnetite slurry is classified in a cyclone cluster, which is used at the production of two special products. The problem is that the quality of the overflow product is irregular. The coarse particle fractions tend to be too large. This causes that a product cannot hold its specifications, and may receive a lower product value or is rejected. This master thesis investigates if it is possible with a more even product quality as well as a lower coarse grain part in the overflow. This is investigated with different samplings procedures, in full production during the summer, and analyses that is based on literature and simulations with MODSIM. The different investigations consist of an investigation on a delimited circuit, with product balances in MatBal, and evaluation of different geometrical and process changes on the cyclone cluster. This master thesis intends to give an increased understanding of the hydrocyclone problem, with the expectation of finding a possible solution and/or proposal on further investigations. Some capacity tests were done as well as sampling for later analysis. The analysis methods chosen for the evaluation of the investigations were XRF, wet laser diffraction and moisture analysis. This master thesis have mapped and given a better understanding on how the circuit and cyclone cluster works, as well what possibilities there are for decreased coarse grains in the battery overflow. From the circuit investigation could smaller variations be observed as well as it was in fair agreement with previous sampling. The product balance compared well with previous ones, except for that the recirculation from the primary cyclone was determined to be much lower. The pressure drop investigation on the cyclone cluster contradicted partly the literature and the simulations, since the overflow gave a coarser product with increased pressure drop, and shows that an increased pressure drop should not be the solution to decrease the coarse grain part in the overflow. The investigations of decreased underflow and overflow diameter followed the simulations and the literature well, but should not be the solution to decrease the coarse grain part in the battery overflow. These changes could be the solution if the lower limits ( / <p>Validerat; 20151001 (global_studentproject_submitter)</p>