To the second part (the ventilator or other device pressurizing to 20 cmH2O): The same as in the experiment with the syringe applies here. Since compensations were not turned on, the additional compliance of the compressible volume makes the overall C just a bit over 1.5 mL/cmH2O. And, again, the piston excursion of the ASL 5000 by itself, as depicted in the raw data plot, follows the set C of 1.0 mL/cmH2). So, the proper way to set up experiments for volume accuracy of ventilators, specifically when dealing with the small volumes of the neonatal range, is to activate compensations in the settings of the lung model. Furthermore, it is important to remember that the raw data only depicts volume contributed by the movement of the internal piston in the device, whereas the data processing provides the volumes that your ventilator will "see" when attached to the ASL 5000 as the patient.

For clarification: The discrepancy between the set C (1.0 mL/cmH2O) and the resulting C (1.5 mL/cmH2O) stems from the fact that compensations were not activated in the settings (video at 2:45). With a cylinder URC (uncompensated residual volume) of 300mL, plus the 200 mL of "dead space" (together 500 mL) the expected parasitary C is 0.5 mL/cmH2O, which is what resulted in the experiment. The raw data, as the name implies, is not to be used as the measure of what actually has been input, it is not processed yet. The volume information in the raw data is showing only the piston excursion, which was set to 1 mL/cmH2O. Since the pressure increased to only 12.5 cmH2O, the piston moves to the 12.5 mL position. The overall volume that is "absorbed" in the system is, however, the full 20 mL of the syringe, because part of that volume goes into the compressibility of the overall volume (to the tune of the additional C 0.5).