Coronavirus - Towards a cheap and easy to manufacture ventilator?

It definitely seems like there’s a lot of sensors and controls on these machines. Making a functional pump is one thing, but how do you communicate to medical workers the status of the system and patient?

It can be a pretty involved process if intubation is required. And I don’t know if COVID-19 complications could be alleviated without intubation.

This is a nice technical video explaining how non-invasive ventilation works. It looks like CPAP is probably not useful here, but maybe BIPAP could be. This makes me wonder if an air compressor could be used as the pressure source to a bunch of small dual pressure regulators with a valve to switch between low pressure and high pressure. This eliminates the need for a bunch of separate motors.

Learning more. It looks like all ventilation is positive pressure, so I’m still thinking an air compressor could be used as a pressure source. One key is that the air needs to be humidified and warmed before entering the body when invasive (intubation) ventilation is used.

I’ve found multiple sources that recommend against BIPAP, so it seems like intubation is the only recommended way to ventilate for COVID-19:

Someone told me the site was not working. I just did a server update and confirmed the new user flow works. You should be able to sign up and post now!

I’ve sketched out a simple mostly mechanical design that I believe could be be easily constructed assuming the hospital has access to the intubation supplies and just needs a mechanical cycling air pump. I’m still doing more research so this might be useless or wrong, but if you need a dual pressure cycling air system, I think it could be a simple manifold with mechanical pressure adjustment for high and low pressure settings, and a timing or sensor based circuit to alternate between high and low pressure. I’m going to work on a quick YouTube video to explain what I’ve found so far. At this stage I feel confident that a simple system could be made easy to produce, but I need input from medical professionals on whether more ventilators are really a bottleneck and if so if what I’ve come up with would have any value.

See the diagram below. I will come up with a video describing it starting with a simpler block diagram in an hour or two.


Here’s a couple more diagrams.

First, a simpler block diagram.

And then a diagram showing how many of these units could be hooked to an air compressor to treat many patients all with different settings per patient.

Finally, though most of it is beyond the scope of this project, this one page fact sheet is very interesting:

I’ve uploaded a youtube video where I describe my thinking so far. I am trying to get in touch with medical practicioners who might know more about ventilators. Mostly I want to know:

  • Is pure high/low positive pressure sufficient for this situation or is negative pressure needed?
  • is a ventilator essentially a switchable high/low pressure device that cycles either with a breath sensor or with a timer?
  • Should we expect that hospitals would have the other supplies needed if we could create the alternating high/low air pressure?

Here is the video:

Naomi Wu weighs in on twitter about DIY ventilators. Makes the case that Chinese manufacturing is best equipped to ship working units quickly. Hard to disagree with. Still thinking on this.

The vibe on the internet seems to be that we could use a good design for emergency manufactured ventilators. But no obvious group has emerged to coordinate design.

Major manufacturers in the US have talked about ventilator manufacturing. I haven’t heard if they have begun yet and I don’t know if they will be able to produce enough in time.

I’ve seen comments from doctors that if you have something that will save lives they will use it even if it’s illegal. That’s what I figured, and it was good to see at least one person confirm.

Perhaps most importantly in this update, UK has published minimum requirements for a rapidly fabricated ventilator.

Hi Taylor,

I watched your video on youtube and have since reached out to a friend who works as an anesthesiologist to ask him a few questions about ventilators.

Most ventilators used these days are high tech with automatic lung capacity pressure gauging, and a multitude of sensors. Repairs must be made by technicians from the manufacturing company.

Roughly twenty years ago hospitals used what are called Volume Dependent ventilators which used a rotary dial to adjust the volume of air used for each patient and were operational from the volume source alone. they contain moisture filters that are discarded when saturated. The power of the air under pressure exceeded the need of human lungs. Adult lungs need anywhere from 15-25psi, and this air typically is piped throughout hospitals (not air tanks) and MUST be purified and filtered.

I have more info but these are the most important parts for a desperately needed functioning ventilator.
I look forward to discussing more with you.
Here’s another video increasing a single operational ventilator’s capacity:

Positive Thoughts,

1 Like

Wow great information Xandra! Thats a useful video. It makes me think if the 3D printing community could have high impact just printing three way and four way adapters. Or maybe the hospitals have enough of those fittings she used in the video? One thing to note - I think you mean 15-25 cm of water, not 15-25 PSI? 25cm of water is about 0.3psi.

Regarding emergency ventilator designs, I’ve found a few groups online that are working on them. I should find appropriate links to add to this sub. But I do get the feeling that in the USA some big name manufacturers may be able to provide the ventilators we need. Not sure.

I’ve found it a bit slow to get traction on open source collaboration here. But according to you and the UK specification, volume control or pressure and volume control seem mandatory. I’ll have to think on how that can be done. Then perhaps I can work with some of the other groups to come up with something. I don’t know if I can help get something functioning for the current pandemic, but I am really starting to think we need working open source designs for this kind of technology for emergencies like this.

Just dropping some links related to 3D printing parts for this crisis.

Reddit has a megathread on 3D printing for the relief efforts.

There’s a printable adapter to adapt scuba masks to breathing masks, though it’s unclear if this is working well or just a prototype (see comments for details):

If you can print in nylon (so it can be sterilized), files for ventilator valves are on grabcad:

There’s also 3D printed face mask holders:

That’s all I have for now.

It looks like there’s also some good communication on this Slack channel:

Hi again Taylor,

Is it possible to 3D print droplet masks? Can textiles be 3D printed, or what other disposable materials could be produced by a 3D printer to help the mask shortage.


P.S. great threads on the valves. thanks for sharing

Hi Xandra,

Sorry for the delay. Yes it looks like masks can be 3D printed. Last night I found this:

Last night I printed the face part of it, with the filter holder printing now.

On my $300 resin printer it takes 9.25 hours to print this part, plus another 45 minutes for the filter holder. So thats 10 hours per mask, or 2.4 masks per day maximum if I never stop it from running.

It is also not recommended to use normal 3D printer resins as they contain chemicals that irritate the skin and can cause allergic reactions. There is a supplier in Barcelona offering a discount of some biocompatible resin for anyone that wants to print masks:

They offer a 20% discount on top of the currently discounted rate if you email them and ask for the code.

I found a directory of open hardware projects including masks and ventilators here:

There are also a lot of people printing masks and ventilator parts here:

I can’t believe six days passed since you asked your question. I will try to respond more promptly if you have any thoughts.

What’s very interesting is that MIT released details on their open ventilator design:

They did the work I couldn’t - among other things they worked with medical professionals to determine what materials were on hand and what a minimally functional ventilator would look like.

I would like to dig in to this and see if there’s anything I can do to help that project.

There are some folks saying the 3D printed masks do not help and are a bad idea.

See also the comments here:

Real Engineering has done a great video on these efforts for low cost ventilators.

Designing a suitable unit is difficult, and most of the DIY designs on the internet would be of little use for COVID patient treatment. Accurate sensors and microprocessors are needed. It would be good to know how the cheap design from India mentioned in the video actually functions, and whether it is suitable for COVID treatment,

Tesla has designed what appears to be a functional ventilator based largely on existing Tesla parts which they can source in volume. I suppose it won’t be open source, but I am hopeful that they can ramp up production quickly at least for the US and maybe Europe.

There are some promising modifications to commercial CPAP machines which can enable BiPAP operation and potentially even suitable ventilator functionality.