I produce P3HT PCBM solar cells on ITO-coated glass substrates and evaporate six electrodes on one substrate. The problem is that I get different power conversion efficiencies on different electrodes. The difference between efficiencies is sometimes larger than 1%. What could be the problem?