Homepage Forums Open Source Ventilator Project Microcontrollers

Viewing 36 reply threads
  • Author
    • #2516

      Please discuss microcontrollers here such as the arduino.

    • #2669
      Thomas sciano

      Esp32’s are dual core and wifi with ble. They can connect to wifi cheaply for a network monitoring system. There is also the esp now mesh network

    • #2735
      Elie Sarraf

      I wrote a comment on the GIT forum, which I’m going to repeat here. Basically, the items described in the git pages are not available/won’t arrive until April 18 – mid May.

      Arduino is a nice piece of hardware, but I don’t think it is scalable for a big project. You should consider using a more “industrial” embedded system like a PIC micro-controller. Ideally the items can ordered from a wholesaler such as Digikey and not Amazon.

      My 2 cents

    • #2738
      Tom Sciano

      I agree that a pic would be nice. But this needs to be something readily available. And arduinos are more readily available in my opinion than getting it from mouser or digikey. Yeah in the long run you would want a custom solution.

    • #2741
      Elie Sarraf

      Links on the git site:
      Arduino board (only 12 left in stock)
      Arrives: Saturday, April 18

      LCD board:
      Arrives: May 15 – June 8

      I can go straight to arduino.cc and get the board but I don’t know how many are available. This is essence, scavenging for parts and from my perspective, and is not off-the-shelf ready.

      Compared to
      Where there is 23000 microcontrollers in stock

      as well as
      With 27k and 5.5k item availble.

      I love what you are doing and I want to help, but I want to be able to build more then 10 ventilators, and this is what I mean by scalable.

    • #2744
      Tom Sciano

      I’m just someone passing through trying to help. I didn’t make any of this. I’m just going to make it when they say it’s been approved

    • #2762
      Barry Allen

      Just a question – Could potentiometers and a 555 timer be used instead of the Arduino controller? It seems like that would be an easier, more accessible build if it could work.

    • #2765
      Pablo Rattin

      Hi folks!, The ESP32 have major hardware bugs in continuos use, it reboots after a few time of operation.
      The networking capacities of the ESP32 an others wifi integrated microcontrollers is not the focus in this development case.
      I encourage you to prototype from arduino and later move directly to a microcontrolle like PIC and CCS C.

      Arduino is a great platform, and the discusion about it performance in production environments become a real war :).
      In my experience is a reliable platform, the problem is about the people who progrming this devices.

      A lot of people come from Software design and hobbyst enthusiast.
      The embeded programing needs another competencies because the hardware is closer.
      I talk about handling interrupts, timers, and microcontroller hardware in general.

      The code can be functional but if the programer don’t take care about this concepts the device eventualy fails.

    • #2825
      Travis Johnson

      Hey all,

      Thank you for the responses. Keep in mind what we have on our site is not final.

      For instance we have recent made the decision to go with an Arduino Nano. What we are currently working on is a base version so that we can run tests and validate our specs. Once the specs are finalized we will update our site with the relevant documents, in hopes that others will switch out components and parts as they see fit as long as it meets the specs.

      Those mentioning using other solutions aside from an Arduino, we agree but do not currently have the resources to fab and test every alternative. We are relying on volunteers outside of our team to do so. If you are willing to do so please stand by as we may need your help in tackling these issues soon.

      Thank you,
      Travis Johnson
      Daily operations manager – CSSALT

    • #3026
      Adam Squagle

      The underlying atmel chips that the Arduino project is using are easily available and can be included on the final PCB very easily. Past that, the Arduino project and its IDE are, in my opinion, the most widely disseminated and generic platform to develop a project of this type on. The ESP32 and others may be more powerful or have more features but the goal is the minimum required, not the extras.

    • #3044
      Mathew Benson

      I’m new to the forum. My background is 16 years as a software engineer working for the NASA community. More specifically, my experience is primarily in safety critical embedded systems. I’ve used Arduino professionally, but never for safety critical applications.

      I recommend against anything with bluetooth or wifi. Keep it simple, with no outward facing interfaces that require testing or safety and security certification.

      Also, I’ve seen a few mentions of performance. I think safety, reliability, availability, and cost are more important than performance and clock speed. Unless there is a specific performance requirement that I’m not aware of.

      I’ve had good experience with the SparkFun ProMicro (https://www.digikey.com/catalog/en/partgroup/pro-micro-5v-16mhz-development-board/51926), but availability might be a problem. My company has developed a design functionally identical to the ProMicro, and released the kicad files open source. We just can’t, and haven’t had a need, to mass produce it.

    • #3080
      Eric Paton

      I have an Arduino Uno R3, so I only need to purchase the LCD. The one posted on Github has longer lead time, and more expensive than this one on Amazon with Prime. Will this be equivalent? The specs look the same.

    • #3146
      Dario Pennisi

      i’m the HW & FW development manager at Arduino. if we can be of any assistance please let me know.
      i agree with most comments here: a solid design separates wireless interfaces from the main processing and that’s why most of our connected boards have a main processor in addition to wireless modules with their own intelligence.
      we can assist you in product selection and help you with logistics as we’re already geared up to deliver large quantities of our boards directly if needed. in addition to the traditional Uno, Due, Mega, and Nano boards we may supply also the MKR family and the newly introduced Portenta H7 which has very high specs and is for more industrial applications.
      let me know how we can help… in the meantime you may want to join us later today in the conference we arranged with people interested in discussing covid fighting gear: https://www.arduino.cc/covid-19/conference


      Dario Pennisi

    • #3161
      Mark Oleszkowicz

      Have small micro Programmable Controllers been considered as the control system solution. they are readily available in large quantities, with all the hardware developed to control valves and actuators with great timing precision. In addition, they are tightly integrated with HMI panels that are easily programmed, very robust touch screens with easy to understand graphics. Rockwell Automation is a highly regarded source around the world and their products have extensive support by systems integrators. These are highly reliable systems, designed for highly accurate and repeatable automation platforms used in high speed manufacturing etc. https://www.rockwellautomation.com/en_NA/products/programmable-controllers/overview.page?

    • #3164
      Mark Oleszkowicz

      Your local Rockwell Automation supplier that supports UF is: Rexel, 4010 W Osborne Ave Tampa, Florida 33614-6526 +18132493200
      they are the Rockwell distributor responsible for the Univ. Of Fla. Contact them for readily available industrial controllers that will easily handle these specifications, with off the shelf and proven hardware in use in critical applications. Safety rated PLC’s are also available.

    • #3170
      Tom Sciano

      Thank you Dario!

    • #3173
      Christopher Gates

      I have been a developer of medical devices for over 40 years and worked with teams on 6 new ventilator designs over the years.

      I view both Arduinos and PICs (i.e. non-ARM cores) as toys for hobbyists to use as a learning experience.
      However, given the exponential ramp up on the virus, what we need is the fastest solution possible, not the best solution!

      I do agree that the host CPU should


      be shared with any communications activity (Wi-Fi or BLE) for a whole host of reasons, including cybersecurity.
      While the FDA allows these very crude and risky designs for emergency use, they stated that security was not to be omitted.

      The best way to achieve that is no communications with the vent, second to that would be for the vent to only broadcast BLE beacons with alerts & alarms in the beacon, to be utilized for remote ‘monitoring’ by overwhelmed staff. That and a watchdog timer and a very loud local alarm!

    • #3179
      Greg Bakker

      Here is a suggestion related to the April 2 Arduino COVID-19 webcast. I initially was not able to connect to the Zoom meeting because it said that it had reached its max of 100 participants. It looks like the conference is also being livecast on the Arduino YouTube channel at https://www.youtube.com/user/arduinoteam/videos

    • #3182


      I would like to help with the ventilator development. I have experience with register level programming with ATMEL 2560 (Used in Arduino Mega) which can easily be translated to ATMEL 328 (used by UNO), and realtime and interrupt based programming with the aforementioned microcontrollers.

      If you need help with microcontroller and electronics interfacing, time critical or realtime programming,
      feedback control, mechatronics, etc. just let me know.


    • #3308
      John Anderson

      This project in no way requires a microcontroller, and requires zero software. There is absolutely no need for LCD display.

      I’ve got a 555 timer based solution testing now, should have MTBF in many 10’s millions hours. 10 to 30 breaths per minute, ratio 1:1 to 1:4. 2 pots give 100% independent control, and no menus to fumble with for setup. Turn it on and dial in pots with infinite control for perfect ventilation on every patient. Power supply is 6V to 36V DC or AC and can use the standard sprinkler valve transformer. Valve can be DC or AC. Finalizing schematics and PCB layout for publication. Can be built on perf board or PCB, and parts are in good supply. In case anyone is interested. I tried contacting developers but never got a response. Maybe this needs a new topic?

    • #3323
      Alex Theodossiou

      Upvote @ John Anderson – keep it simple.

    • #3326
      Christopher Gates

      Love it… the KISS rule always is the best

    • #3329
      Christopher Gates

      @John Anderson, how about another 555 and a pressure switch just before the breathing circuit to ask as an alarm. Would detect controller failure and circuit disconnect.

    • #3332
      Barry Allen

      @John Anderson That’s exactly what I was thinking! I replied onto this board but nobody seemed interested. I think the Arduino controller is a little much for something this simple. Now that I’ve thought about it more, a 556 timer might work better to open the gate for the exhalation after the inhalation is complete.

      @Christopher Gates – I think this would work. Pressure gauges would need to be included so that an analog readout could be added.

      I would assume that the respiratory therapists are trained on displays similar to the original build. I would think this would be a benefit to people in countries where FDA regulations aren’t a consideration and retraining wouldn’t be an issue.

      Good job guys! Let me know if I can help on this design in any capacity. It may be 1950’s technology but when life is on the line, who really cares about the technology behind the device?

    • #3338
      Christopher Gates

      The FDA will not be a problem, they have issued an emergency use authorization (EAU) for ventilators. The only one issue that really left in place was for cybersecurity, so again the KISS rule applies, no communications, no CPU/MCU. Hard to hack a 555 timer {grin}

    • #3341
      John Anderson

      I’ll have preliminary schematics ready in a few days. A new set of parts arrives 4/7 and then I can publish a more finished set.

      Turns out the sprinkler valves work best on 24VAC and that’s what my design works with. Driving valves on DC may present a reliability issue longer term – more on that later.

      Analog gauges are not a problem – in fact 0-15 and 0-30″ H20 (convert scales to cm H20) are readily available at Aamazon and other suppliers. Tested on a CPAP machine and work fine – around $20~$30.

      If you’re gasping for air, you want a solution that works in the simplest most reliable way possible.

    • #3359
      Mark Oleszkowicz

      You need to carefully examine whether or not a “sprinkler” valve is viable for this application. At a minimum you must know the valve CV or in this case CG and what the pressure drop is under actual patient in-use applications. High pressure water is not the same as very low pressure air. http://blog.parker.com/choosing-a-high-flow-proportional-valve-for-ventilators-and-anesthesia-machines

    • #3362
      Mark Oleszkowicz

      You need to have selected the proper airflow control valve before you even begin to design the controller. That valve selection is critical.

    • #3398
      Bill Persyn

      I built one of these in my garage yesterday and used a Rockwell Automation Micro820 Industrial Automation controller. It has a nice display and a couple of buttons on top to change rates and stuff. https://literature.rockwellautomation.com/idc/groups/literature/documents/pp/2080-pp001_-en-p.pdf

    • #3422
      Mark Oleszkowicz

      Yes, using Rockwell Automation or other industrial PLC with easily integrated HMI is the ONLY way to go, because it is proven technology, exceeds the requirements, and is off the shelf. That married with the proper Parker Hannifin approved respiratory valve gives you a real shot at a working design. You need proportional control that is accurate and repeatable and a PLC with a proven instruction set that includes PID control, analog in and out etc. Real time trending is also available on their Panelview products, with touchscreen entry. Avoid the low buck hobby shop stuff. Timing to deliver is all that matters here.

    • #3428
      Dave Lizdas

      we need a battlefield tourniquet ventilator to bridge the gap until new ventilators become available. what is available everywhere including third world? Parker is not taking new customers because they are struggling to meet demand on the vent manufacturers they currently support.

    • #3449
      Mark Oleszkowicz

      Well at least you know now that the Parker valve is definitely the right valve to use and design for. Whatever controller design should be predicated on at least using that valve. Hope that initially they can give you one sample to use to validate your controls design. Production act can then maybe free-up others to build their valve to Parker specs.

    • #3452
      Mark Oleszkowicz
    • #3458

      we are NOT designing beyond covid crisis, these vents should NOT be used unless there is literally NO OTHER OPTION AVAILABLE using parts everywhere. we don’t want to take real vent parts away from vent manufactures, the point is to close a gap. I use parker valves in my work. Dr. Lampotang is mech engineer PHD, not anesthesiologist, and has designed and built ventilators. Thank you for the link! can you identify suitable replacement valve in that catalog and call manufacturer on lead time, price and availability? will it hurt vent manufacturers if the are not available?

    • #3590
      John Anderson

      Running tests on multiple 555-driven pulse circuits across 10 sprinkler valves from Rain Bird, Orbit etc. Cycling as fast as possible on valves, using common irrigation 24VAC transformer as power suply to drive valves and run the pulmo-pulse circuit.

      HEAD’S UP WARNING: We have seen a failure in one Rain Bird valve as shown on this web site. It is not the valve’s fault but pay attention: These have a tendency to stick “open” if there is any particulate matter in the air supply line. We’ve seen an effect where a very small sliver of teflon tape used to seal the joints on the incoming air regulator wound up in the solenoid pilot pintel valve seat. Keep your air supply piping perfectly clean – we now assemble the test units in a clean room area, flush all piping before test, and use an inline filter ahead of the air regulator.

      Also watch out for increased O2 concentration on that (probable EPDM blend) rubber diaphragm in the valve and in the air regulator assembly.- we’re taking a look to see if that might cause a release of particulates from the rubber surface. Keep an eye out for any other source of particulates in the system, and watch out for static buildup that occurs on PVC that can cause debris to stick to the surface, and then get released later.

    • #3605
      Gustavo Pohls

      Hi Guys,

      I am a Mechanical Engineer and a Amateur Astronomer who has built his own low budget automation projects in Mexico. I am watching very closely all the discussions in this website and I have decided that I should build the 1.2 version of the DIY Ventilator. My question is : Where can I find the Arduino software that is used in this solution? I have looked everywhere and I cant find where I can download it. Thanks

    • #3698
      John Anderson

      Here you go – a LOW COST, ZERO SOFTWARE simple ventilator controller, designed, built and tested. Uses the same 24VAC power transformer you’d use with sprinkler valves anyway. Could build some thousands, 10’s of thousands or more with just the stock available at various electronics distributors today.

      This is example is only one way to do it – there are countless other good ways also.

      Could make an entertaining class project as well.

      An example schematic, circuit description, demo video and oscilloscope screenshots are included.


Viewing 36 reply threads
  • The forum ‘Open Source Ventilator Project’ is closed to new topics and replies.