The vast majority of rapid-response emergency ventilators are variations on the “bag-squeezer” design originally developed at MIT in 2010 (http://news.mit.edu/2010/itw-ventilator-0715). In these designs a mechanical apparatus compresses a BVM (bag-valve-mask – also known as an Ambu-bag after the brand-name Ambu) bag forcing air into the patient’s lungs. The bag squeezers have the advantage that they do not need compressed air or O2, but have a number of disadvantages. First, they are heavy, more difficult to assemble , have more moving parts, and as a result are less reliable than a PSV ventilator. The PSV has a single moving part – the plunger of the solenoid valve. A bag squeezer has motors, a gear train, and arms. The more complex linkage of a bag squeezer also makes it more difficult to control. There is more mass and backlash between the PWM control from the microprocessor and airflow limiting the control and making it difficult to deliver very precise flow and pressure waveforms. The electronics are also more complex since rather than providing a few hundred milli Amps to drive a solenoid, the controller must provide 10s of Amps to drive motors. This larger current draw also makes it more difficult to provide battery backup. Some argue that bag squeezers have the advantage that a caregiver can take over manually by removing the bag from the squeezer. However, manual ventilation is possible for all of the designs by disconnecting the ventilator from the ET tube and using a separate BVM resuscitator.
Several designs also use non-proportional valves where the flow is manually adjusted via a fixed flow meter or valve and then this fixed flow is gated by an on/off valve. The University of Florida design does this using a Rain-Bird sprinkler valve (https://simulation.health.ufl.edu/technology-development/open-source-ventilator-project/). This approach has several disadvantages having to do with regulating flow. First, manually adjusting flow makes it difficult and error-prone to set parameters for a patient. Changing tidal volume, respiratory rate, or i:e ratio requires calculating a new flow and manually setting it – with the prospect for human error. This approach also makes it impossible to regulate on pressure (short of a pop-off valve). The flow is set and if breath stacking occurs and pressure builds, the same flow will continue to be delivered until the pop-off valve opens. Breaths can be ended early, but flow cannot be reduced. These systems are also open-loop, the flow may vary due to varying back pressure and supply pressure and this flow cannot be automatically adjusted.