The Pandemic Ventilator




Introduction: The Pandemic Ventilator

Assembly instructions for a DIY Ventilator prototype.
This could be useful in an Avian Flu pandemic.
Constructed with commonly available components.

Many of us modify, hack, re-purpose, and DIY to save money, build something unique, create art, or show the world that there is a better way to use some device. And sometimes, just because it's cool. This is something different. It is a ventilator, and ventilators are meant to save lives. This project is called the Pandemic Ventilator, because it is meant to be used as a ventilator of last resort during a possible avian (bird) flu pandemic.

Many health authorities are preparing for the possibility of a flu pandemic in the next few years.
If a pandemic occurs that is related to the type of virus that is currently spreading in birds, they fear that it may be as bad or possibly worse than the 1918 Spanish flu pandemic. It is expected that the number of people that require treatment with ventilators may be much greater than the current number of ventilators in existence. If a pandemic were to strike, the hospitals could not just go out and buy all the ventilators they need, because there would not be enough parts or manufacturing capability. Many governments already have plans for triage and rationing programs that will determine who gets access to the limited number of ventilators and who will be left to die. When I first heard about this, I thought, "This is not good enough, if someone I know or love needs a ventilator, I would get one, I would build one myself if I had to". Thus the idea was born.

The earliest ventilators of the 1950s were primitive devices with even more primitive control and sensor systems, but they worked, and they saved many lives. Some of the early ones were built in workshops. This ventilator has a very primitive and basic design, but then it does benefit from a modern electronic control system. This is a basic ventilator design using materials that would still be readily available (or re-purposed) if a pandemic were to occur. It uses wood, tape, plastic bags, threaded pipe, solenoid valves, security system magnetic switches, and a PLC (Programmable Logic Controller). The prototype shown does not yet incorporate all of the proposed design features and the control program still needs some work to make it more stable and failsafe but it does function, as you can see from the video.

The Pandemic Ventilator Project is an open source hardware project. If you build your own development unit, please share your ideas, and experience at .

The information in this instructable is presented as is for development and investigative purposes only. The prototypes presented are not fully functional devices and have had no safety testing done. A ventilator is a potentially hazardous device and should only be operated by a trained and certified respiratory therapist. Any usage guidelines will be published for emergency use only, and only when a fully functional and validated unit has been completed. Anyone using this information to build or use a device agrees to waive any and all liability.

The Pandemic Ventilator Project

Teacher Notes

Teachers! Did you use this instructable in your classroom?
Add a Teacher Note to share how you incorporated it into your lesson.

Step 1: What It Looks Like

The Pandemic Ventilator Project
It basically consists of the bellows unit, which is made of wood, valves and piping, a PLC controller, some wires and switches and a power supply unit.
The whole unit is mounted on a piece of 1/2 inch thick plywood that is18 inches by 21 inches.

Step 2: Valves and Pipes

The Pandemic Ventilator Project
To get started you will need the valves. You may be able to find some used ones somewhere. New valves are expensive, over $100 each. The inlet valve can be 1/4 inch normally open valve, the other two should be a minimum 1/2 inch diameter, one normally open (NO) and the other normally closed (NC). They must be of a direct acting solenoid type.

Direct acting solenoid types are required to operate with air. Pilot operated types will only operate with liquids. Pressure ratings do not matter for the 1/2 inch valves, but the 1/4 inch valve should have at least a 50 psi rating. If you can only get either NO or NC 1/2 inch valves you can still make it work by adjusting the PLC program. The inlet 1/4inch valve must be a NC valve or you will end up blowing up a lot of bags in the testing phase.

The valves are connected with pipe and mounted in such a way that the T to the Bellows lines up with the center of the bellows unit.

I used threaded pipe fittings. with teflon tape on the fittings.

I also used:

Two 1/2 inch NPT Ts, One 1/4 inch airline fitting (to connect to an air supply)
Four 1/4 inch pipe nipples (short pipe sections threaded at both ends)
Three 1/4 inch to 1/2 inch adapters
Two 1/2 inch pipe nipples
One 1/2 inch plug (this plugs the line that will go to the manometer planned for later)

Step 3: Making the Bellows

The Pandemic Ventilator Project
The bellows unit has a hinged section of quarter inch plywood that is 10 1/2 inches by 12 1/2 inches.

There is a one and a 1 1/2 and a 1/2 inch by 9 inch sensor pole attached beside the bellows that is used to position the magnetic switches. These are magnetically activated reed switches that are usually used for security systems. The sensor pole should be angled at the bottom so that the top leans back about 1 1/2 inches from the bottom.

The bellows hinge is constructed of 4 pieces of 1 1/2 by 7 inch 5/8 inch plywood pieces and a 1 1/2 by 1 1/2 inch by 17 inch piece of wood, two 3 inch hinges and a 2 inch by 12 1/2 reinforcement.

The bellows is made by screwing down the bottom 2 plywood pieces to the backing board.

Arrange the other 2 plywood pieces directly over the first two, cover with the 17 inch piece of wood and and clamp together.

Drill holes centered 2 and 3/4 inches from each end to accept 1/4 inch carriage bolts inserted from the bottom.
Un-clamp and remove the top 17 inch piece of wood.

Insert carriage bolts from the bottom to line up the middle layer plywood pieces and clamp again.

Screw hinges to middle layer plywood. Ensure the screws do not protrude through the plywood to the bottom layer.

Screw the other side of the hinges to the bellows lid between the lid and the reinforcing strip.

The bag is clamped between the two plywood sections during operation using the nuts and washers on the carriage bolts.

The magnet is attached to the end of the bellows near the sensor pole, and the sensors are attached to the sensor pole. I just used duct tape to attach them so that they can be easily adjusted during the testing phase.

Step 4: Adding the Bag

The Pandemic Ventilator Project
To make the bag for the bellows, I used a large size Ziplock freezer bag.

Cut off the ziplock part.

Inert 1/2 inch plastic tubing into the center and use Tuck tape to seal and reinforce the edges.

The tubing should stick out of the bag far enough to be able to be slipped over the end of the 1/4 inch nipple section of piping.

The taped seam of the bellows bag should be in on the bottom plywood section.

Install the hinged cover and then the top 17 inch section.

Clamp together with the 4inch long 1/4 inch carriage bolts, two nuts and 2 washers.

Step 5: PLC, Program, and Wiring

The Pandemic Ventilator Project
The PLC unit is a Direct Logic 06 DO-06DR from Automation Direct. Automation Direct
Their units are low cost, flexible enough and they have a lot of free software to program with.

You could use other PLC units and write your own control program. If you do. please share your work at The Pandemic Ventilator Project so that others can use your program and insights as well.

Besides the PLC you will also need a 24V power supply, and an on off switch to start the system.


Also below is the ladder logic program that runs the unit. (My son does the programming) As I said it is still very rudimentary. It will require some improvement to make it more stable. The program will also have to be extended to accept the inputs from the future manometer, and also incorporate safety systems and alarms.

Basically what it does is:

It opens valve 1 and closes valve 2 until the bellows is full which is indicated when the top magnetic switch closes.

It then closes valve 1, opens valve 2 and closes valve 3 so that the bellows can deflate and pump the air to the patient.

When the bellows drops to the lower limit, the lower magnetic switch closes and then valve 2 closes and valve 1 opens again to refill the bellows.

A timer lets the patients lungs deflate with valve 3 open. When the timer expires, valve 2 opens and valve 3 closes to start the next respiration cycle.

In the future I hope to get the manometer to be able to control the respiration cycles as well as the timer.


Everything is wired back to the PLC connectors and power supplies..

My unit has all NC valves so the program should be slightly different if you use a NO valve for valve 3 as I recommended. Also I used all 120V valves. If you use 24V valves, then the output CO will be connected to 24V instead of 120V and the output returns would be connected to the 24V ground. Here is the wiring chart.

X0 Top bellows mag switch
X1 Bottom bellows mag switch
X2 on off switch
C0 24V
All returns to Ground

Y0 Unused
Y1 Inhale valve (V2)
Y2 Exhale valve (V3)
Y3 Bellows fill valve (V1)
C0 120VAC line
All returns to line Neutral

Step 6: Ventilator Running

The Pandemic Ventilator Project

Here is a video of the Pandemic Ventilator running.

I am using an earlier bellows design for my patient simulator. If you wish to test run your own unit you can build a second bellows unit for a lung simulator. No not try it on yourself or anyone else. The patient airline that runs to the simulator is 1/2 inch ID plastic tubing that has been Y ed together at the patient end. You need a regulated compressed air source to run it. It is best if you can get the pressure down to about 10 psi. The higher the inlet pressure, the greater the chance that the bellows will inflate too fast for the upper level switch to detect. If this happens, you blow up your bellows bag. When I initially test ran my unit, I used switches to simulate the bellows mag switches to confirm that the program would function as intended.

I put some wrenches for weights on the bellows and simulator. The weight on the bellows provides the pressure to inflate the patients lungs. The weight on the patient simulator bellows simulated the lung expelling the air. The bellows weight has to be greater than the simulator weight to make it work.

Be the First to Share


    • Tiny Speed Challenge

      Tiny Speed Challenge
    • Woodworking Contest

      Woodworking Contest
    • Trash to Treasure Contest

      Trash to Treasure Contest

    67 Discussions


    Tip 1 day ago

    Dear Everybody.
    This is a great thing to do, and in the current COVID-19 situation, it can be acceptable with a DIY version.
    HOWEVER: I would, as a pretty experienced constructor of equipment do the following changes:
    1) I would remove all the valves.
    2) I would use a 12 volt motor from a wiper from a car. These NEVER EVER fail. They continue working, no matter what.
    3) I would use a big car battery - or truck battery - 200 Ah - to ensure that power is available even during "power breaks".
    4) I would put a car charger to charge - and a power supply supplying the car charger. This to ensure that battery would indeed be maintained well, and not suddenly and invisibly, lose charge capacity - rendering a patient dead when power goes (and everybody thinks that the ventilator of course continues.
    5) I would equip the system (for use in 3rd world countries) - with a set of pedals and a bike-seat, for manual use, in case everything else fails.
    6) I would make the "in-ventilation" be controlled by a permanent weight, adjusted such that it provides proper inflation - I would guess some 0.05 kg/cm² would be OK as air pressure
    7) I would use the wiper motor to lift the permanent weight, and would let the weight itself provide the air-pressure for the patient to breathe in via an air-container where the weight is put on top (for air pressure). The permanent weight should be adjusted such that the patient feel comfortable.
    8) I would by all means avoid any computer if I in any way could. They fail... (ie. a mobile phone coming too near, hardware malfunction etc.
    9) I would regulate the breathe in and breathe out periods with a very primitive electronic circuit, ie. a NE555 timer - and would make the breathe in drive for the motor with variable power (PWM control of the wiper motor) - The regulation would be separate in terms of 3 factors:
    9.A) The time between letting go of pressure and applying pressure again.
    9.B) The time between applying pressure again (letting the weight fall) and letting go.
    9.C) The speed of the motor.
    - The above 3 should have some 3 knobs to adjust timing and speed.
    10) A DC-based wiper motor uses some 1-2 Ampère when there is no resistance. Thereby it ought being able to run for days without charging.
    11) I would - at appropriate places - apply even a Solar Cell with some 8 times the wiper motor charging capacity (to provide power the 4 out of 5 hours where there is no sunshine) + charging unit - such that it can even be setup anywhere as long as there is daylight.
    12) I would use much larger tubes.
    13) The valves I would use such that the air container with the weight on, fetches fresh air from outside, and when the weight is applied, it blows into the patient's lungs.
    14) I would ensure that the air container is extraordinarily solid and durable, as it will have to operate some 10,000 breathing operations per day.
    15) I would make the wiper motor slow-start, with a Capacitor - thereby ensure durability and pleasance for patient.
    16) I would 3D-print the piece which would fit with a mouth piece.
    17) I would 3D-print the mouth piece too.
    18) The 3 regulators of let-go time, breathe-in time and speed of motor, should be available for the patient.
    19) The mouth piece should be fit with standard gauge pipes.
    20) We should indeed get the dimensionality of this.


    Reply 1 day ago

    I agree on some but what I disagrees on is that there is a need of more advanced controller than a 555 timer but not a computer as you point put. An ardruino should do, the PC is important for monitoring of the patient maybe even remote.
    Removing the pressured air is difficult and I would rader use that wiper motor as a generator for air pressure. The rest I agree on. :)


    Reply 8 hours ago

    I have seen many an Arduino halt or lock, or start executing from funny places in the software. Software is inherently unstable, in the Arduino case also due to its sensitivity to electric pulses/sparks/electromagnetic disturbance. I would therefore use an NE555 or similar timing circuit, such as to avoid sudden malfunctions.

    An Arduino could be made with an external (NE555?) timing device as a watch-dog timer - but again - the more equipment, the more parts can fail.

    My thought is to remove the valves, again due to that these too fail. Therefore I would use a bellow, with a fixed weight on top (adjustable along a stick) (see image).

    I would therefore raise the top wood, which is hinged at the right hand side - by use of the motor - and then the weight on top thereof to push the air into the lungs. The two red valves would be gravity-held valves, which are made of simple rubber and hinged in one side. (Metal hinge plus rubber to tighten).

    All in all, I would make a primitive - absolutely primitive - solution - such as to ensure that as few components as possible can fail.

    We are having problems finding out how to make a bellow. (The black thing): We are using one of these ventilation/air tubes - but would like to know if there is any bellow type or style or technology which just works?

    Bælgberegning ved 0.05 bar.png

    7 days ago

    What if we use a large peristaltic air pump? it need no valves just for exhaust, easy to clean and sterilize with gas or plasma systems,even disposable tubing, for low pressure air the tubing thicknes could be very small, step motor controlled, accurate flow no need for flow control,just some security, please comments??? is a low torque pump, so it can be even 3D printed.


    Reply 1 day ago

    Dear Ibonavino,
    I would not put anything which can lock or get spoiled. I would really keep it simple. Stepper motor - how about if it gets stuck?
    The tubes ought being big, so that the air-speed is low. (large volume, low air speed)
    On air pump, I propose a permanent weight, which is being lifted and released to pump air into the patient.


    Reply 11 hours ago

    I use to repair medical devices for labs, peristaltic pumps are very popular, never seen one stuck, mainly because are low torque jobs for the motor, most critic fails are due to cheap motor drivers not handling correctly reverse currents, any system could fail, the great advantage of this is, in case of emergency or critic failure, could easily be operated manually by spining the motor, a few low power leds could show the recomended speed. Anyone knows the air volume/speed ranges on ventilators?


    Reply 1 day ago

    Looking at some images of peristaltic pumps for the first time, I like your general direction of using the rotational nature of a motor itself to directly control timing. Whether it squeezes a tube (as appears the premise of a peristaltic pump, not 100% sure), or pushes on a bellows with a geared-down cam, this has the potential I would imagine, to have the fewest operating parts. When I conceive of the design of a solution for this problem it is ONLY in the context of the most urgent, "my loved one or me will die if I can't automate air into them" context. The big manufacturers might/should fill the COVID-19 ventilator gap in a few weeks (from March 27, 2020) but the whole point of any such DIY is if/when they don't and one's choice is between no ventilator or the worst ventilator in the world. I will attempt to be making the worst ventilator in the world ASAP and will likely use the direct-cam approach on a bellows. Great work by all.


    18 hours ago

    $200-400 form working device, needs nurse to install, but not £18,000 per unit?

    someone show how to build one in his garage. Raspberry pi device $50 will display via GM do to television.tell the press!


    2 days ago

    I have so many problems with this.
    This was written when there wasn't a pandemic in progress. At that time, if a ventilator is on your list for preparation then get a real one and get qualified to use it.
    We are now in the grips of a pandemic and a shortage of ventilators could be a reality but it would be ill advised to use a device like this on a person unless they are turning blue and there is no chance of real emergency care.
    As for the plan itself it apparently relies on some kind of compressor. Compressors for supplying breathable are are special made so chances are you don't have one.
    I was surprised at the requirement of a plc. The controller isn't that complex and the chances of having a few relays available seems more likely. Ford is working on ventilators made from blowers rather than a compressor. This looks like a safer and more dependable approach.
    The bottom line is that you don't want to kill somebody while trying to save them.


    Reply 2 days ago

    I asked a few days ago, if he could provide a COMPLETE shopping list of items that were used for this, but haven't heard back. I like the idea of using a PI or Arduino as the controller which I think is more realistic... Regardless if you use a blower or a compressor... there's also a Regulator to control the flow of pressure. As a pilot, I use these on my oxygen system every day to regulate how much oxygen goes through my mask at various altitudes. Granted, what you're saying is accurate.... some form of "pressurized" air would be needed at the inlet of Valve 1. Though this could be a constantly supplied air source that is pressurized via a turbo inlet (like in a car) or a blower or somtehing, and then all the regulation could be handled by an in-line regulator attached to the plastic tubing.

    All this said, between this page and the link to the pandemic ventilator, I still haven't seen a complete "step by step guide" to create what this is intended to be used for.... "a LAST RESORT, Pandemic Ventilator...." However, given that there are so few ventilators available even with FEMA's stockpile, and companies trying to re-configure themselves to create them... We probably need something like this. Though I'm not suggesting people learn to intubate. It could be something a sick person brings with them to the hospital.


    2 days ago on Step 5

    Be advised, plc relay outputs cannot handle high current loads like direct acting valves require. Simple solution is to drive a relay with your output which triggers the solenoid. In industry it is referred to as"isolated outputs".

    Without the relay, the contacts will not last long.


    2 days ago on Introduction

    Have you been in contact with any cities about building this unit?


    4 days ago

    What does the code look like that you used for the PLC controller?


    Reply 3 days ago

    The PLC isn't coded through written code, but rather through the ladder logic diagram. The diagram is in one of the pictures in the article.


    5 days ago on Step 6

    Absolutely brilliant and something as simple as this could save so many lives. Well done, fantastic instructable.


    11 days ago

    Raspberry pi - $10 stand-a-lone credit card computer device. Order on eBay, curry’s, pcworld etc. Has usb interfaces, hdmi cable as standard, so tv acts as monitor.

    has memory stick, so can run Linux. Just need programmed - probably 1-2 days work for monitoring app, main question usb-to-what connection into ventilator?

    we need pi page, so we get developer (coder), project manager ( order of build,test) , some testers, then create proof of concept. Doctor for hygiene issues ( wet wipe between patients or more...? Do we sterilise the breathing tube ( say in a microwave with water steam steriliser - or using baby bottle steriliser (order on Amazon)

    millions of these devices exist in world.

    please add pi project page and have two stages

    1) ventilator build
    2) pi controller project

    we can do this independent of governments. Each town, can use library (all will close) as workshop to build production line for pi installations ( libraries have good internet) .

    lets get this working!

    amazon has1000s of these devices in stock in every county!

    I have pi device already!

    help us ( the world to help you)

    Awooga ! We can do this together!


    Reply 5 days ago

    100 % Right


    Reply 5 days ago

    If you don't get an MD for your project, would an RN with 15 years experience in ICU's and similar high aquity health care settings be helpful? I am very familiar with the operation and use of Mechanical Ventillators.


    Reply 7 days ago

    check my comment proposing to use peristaltic pump


    12 years ago on Introduction

    >>"Most authorities believe that a flu pandemic will strike the world within the next few years. " No offense meant but I do not believe Instructables is here to spread falsehoods or to scare people. "Most" implies the majority and the majority would include the US, UK and other World Health Organizations. Since none have officially come out and said they "believe that a flu pandemic will strike the world within the next few years" then by definition you are spreading unproven and unfounded falsehoods. In my eyes, thats not cool. Not here anyway. You can go somewhere else to spread that.. say like the ATS forums. FYI: The amount of people who died last year from falls from a 3 foot ladder far outweigh the number dead from any "bird" flu. I like your instructable, it's creative, but its premise is .. seriously flawed.