| General > General Technical Chat |
| Medtronics' "open source ventilator" bullshit PR |
| << < (9/10) > >> |
| SparkyFX:
--- Quote from: rstofer on April 06, 2020, 06:25:08 pm ---Releasing the design was a good move. It takes the heat off Medtronics and makes them look like a responsible corporate citizen. "Hey, we gave away a perfectly good design, some company just needs to step up and start building!". --- End quote --- As far as i understood it there went some serious tax money to develop these ventilators years ago because the next respiratory illness that makes it into a pandemic was kind of foreseeable. I mean no one keeps a manufacturer from hiring more people to build more of them, it is not rocket science to assemble one. So either they do not have the parts or some other bottleneck that someone that tries to make a clone will also face. But it is unfeasible to do so in three-six weeks time. --- Quote ---Worse yet, hospitals will be flooded with ventilators and the market won't recover for years. Medtronics could possibly run additional shifts (if possible) but expanding capacity through capital investment is a money losing proposition. --- End quote --- I am not sure if manufacturers of medical devices can allow to go such a route, their customers are hospitals after all and this type of machine goes only with intensive care, so the usual amount of devices they sell is coupled to the amount of intensive care units that are in the country. Might as well just have been produced on order with long lead times, they are just not prepared for this - no one is. It is not their only product as well. |
| amyk:
--- Quote from: rstofer on April 06, 2020, 06:25:08 pm ---Worse yet, hospitals will be flooded with ventilators and the market won't recover for years. --- End quote --- On the bright side, we can maybe look forward to some interesting teardown videos in the future... of ones that have hopefully been thoroughly sterilised, of course. |
| donotdespisethesnake:
--- Quote from: free_electron on March 31, 2020, 03:05:54 pm ---Also note that the cpu is probably long obsolete. if i recall the ST10 family was discontinued in 2008 ... --- End quote --- Well, that would explain why they released this particular product design. They probably have a dwindling stock of CPUs and plan to stop production soon anyway. Edit: Actually it appears the CPU used is still available but NRND, https://www.mouser.co.uk/ProductDetail/STMicroelectronics/ST10F276Z5T3?qs=SoAZXCQuyY%252BYJOKV3TS93A%3D%3D. Quite expensive, and seems to have some unusual features. Edit2: pretty daft releasing Open Source, but having to register to get it! Anyway, Seth Hilbrand, of KiCad Services, has set up https://gitlab.com/openventilator. Plan is to reimplement designs in KiCad. Edit3: to be fair, Medtronic are not calling it Open Source, merely "open". The "permissive" license reads: --- Quote ---Modified Permissive License Limited Term. This permissive license is effective from the time you download the Design Materials and Software until the earlier of (i) the final day the WHO’s PHEIC is in effect or (ii) October 1, 2024 (the “Term”). --- End quote --- They clearly have no f*** clue what permissive means, regardless of anything else! Completely pointless doing anything with this platform, with such a highly restricted license. So yeah, I'm calling bullshit on this too. |
| free_electron:
Has anyone scrubbed the bom of this thing ? 3/4 of the world is busy converting the files to whatever platform-du-your , lambasting medtronics for the license or other dilly-dallying. The supply chain is the most critical thing for these machines. And i'm not just talking electronics. Valves ? The display ? the membrane keypad ? There are some really tough to get parts in this kind of machine. All the 'conversion effort' is a waste if you can't get the parts to build it ! |
| Kilo Tango:
Perhaps instead of trying to trace down supplies of obsolete components and ending up with desoldered junk from you know where, as had been said else where, we should use the Medtronics information as a learning pack to tell us what is needed and redesign the control system from scratch. Its not difficult, provided the code is robustly written in say C, then you use a different processor, write a few drivers, assemble the lot and that's it, it will work the same if not better than the original. I was involved in redesigning part of the control system for a Pharma machine. It used a plug in AIM card with a 386 processor and obsolescence was a pain. I basically put down an LPC1768, and connected everything up. Chips like that make things very easy, its all I/O lines and serial interfaces, plus the speed of 100MHz means it will process faster than older stuff. The software engineer wrote drivers to replace the originals and that was it, it still ran the original code, and was better cheaper, faster and more reliable. As far as a membrane sealed keyboard, would this do ?. https://uk.farnell.com/multicomp/mcak304nwwb/keypad-3x4-array-plastic/dp/1182235 It doesn't need to have the exact key layout, just needs a method of signalling the system to do a certain function. Don't create problems when there is a simpler answer, Ken |
| Navigation |
| Message Index |
| Next page |
| Previous page |