Since you mention B v. I scanning and signals on Cup 3 it sounds to me you are actually using an isotope ratio mass spectrometer (IRMS). In these multi-collector instruments the three cups are used for collecting masses 28, 29 and 30 (or 44, 45 and 46) since the isotope abundance measurements of 15N (or 18O or 13C) are based on determining abundance ratios, e.g. 29/28 and 30/28 or 45/44 and 46/44. Given the difference in abundance between m, m+1 and m+2 the three cups are fitted with different resistors resulting in different amplification of measured ion currents so all signals are of a similar level.  Typical relative amplification factors for cup 1, 2 and 3 are 1, 100 and 300 (333 actually), respectively. For this reason, during a background scan (changing B) m/z 28 (either 14N2 or 12C16O) will yield a huge signal on cup3 given that 14N, 12C and 16O are the most abundant stable isotopes of nitrogen, carbon and oxygen. Under normal operating  conditions m/z 28 will fall into cup 1 (while m/z 29 and 30 will fall into cups 2 and 3, respectively).

if I understood correctly, when you analyze the CO2 from the reaction of carbonates you out in making a mass scan that you also have a high signal on the mass 28. You think it is nitrogen. But I think not! This signal observed on the mass 28 is the result of the partial decomposition of CO 2 in the CO source (i.e. that the CO has a mass 28). In reality the form of plots masses 28, 29 and 30, one can know the N2 gas or CO gas.