Recreating a Libre/Open Source Otoscope in FreeCAD

[tarek]

Hello friends,

Our group has created a Libre/Open Source Otoscope that is clinical-grade and is Health Canada licensed as a Class I device. The goal of our project is to make high quality, low cost devices that are open source so as to make medical devices more accessible.

However, Frankie, the amazing audiologist behind the design and testing, made it using TinkerCAD. There are a number of reasons for this, but essentially they come down to a lack of familiarity with FreeCAD and easy availability of TinkerCAD at the time for a person who was trying to make a prototype quickly. This has created a situation in which we can't make the 'source' of the otoscope available for others to modify the design as needed, since TinkerCAD makes it essentially impossible to do so.

We are seeking help from the community to recreate the object/STL designs as FreeCAD, hopefully done to good standards. As per this Reddit thread, @mutrax_be is willing to help, and @luzpaz (I think @Kunda1 here) has suggested we reach out here for help.

@Luzpaz opened a Github issue on the matter, and we are asking for community engagement.

We value everybody's contribution and have some money for the modelling aspect of this project that we are happy to either pay to the contributors or to donate to FreeCAD if that's what the contributors desire. I'm not sure how to properly value the contribution (i.e., how much it it should cost), so perhaps we can gain consensus on that through the this thread too.

Thank you so much for your help on this, and hopefully we can soon realize the full potential of this project by honouring the Freedom to make the source code of this otoscope available.

Thank you, and please join this thread or the github issue to discuss further.

tarek : )

[Kunda1]

Greets @tarek
Welcome to the FC forum.

This was originally posted in the Job Offers subforum but because the idea that if the community contribute their time to build these models in FreeCAD using one of the newer Assembly paradigms that perhaps some funds could be donated to FC or even be done Pro Bono, I moved it to the Open Discussion subforum for the time being

[kisolre]

You should provide some drawings or dimension/angles of mounting surfaces because extracting those from the STLs is not easy and not accurate. I already have two of the smaller pieces recreated and one had an elliptical surface and 24.24 deg wedges... I can't be sure that is correct. Maybe link to the Tinkercad source will also help.

[eddriesen]

EDIT: To get straight to business, go to my post about current organisation
So yeah,
Like tarek mentioned, I'm starting on the freecad versions. Using freecad as my private cad tool is long overdue, so I'm jumping into the deep end on this one...

The reply I posted on Github;

I'm creating the files as we speak. Please do mind that freecad is new to me. So bare with me.

Things i'm figuring out atm.;

• part / body / project structure

• ok to use assembly2 module (prime time?/ Getting an error on 0.18.3)

• seperate proj file for each component vs project file with multiple parts or bodies vs seperate files and assembly2 project file

• any topdown / associativity possible?

Question; any possibility gaining the real source files (not meshed / with dims?)

This means I'm drawing as we speak, I'm just figuring out what the best practice is for an assembly, and if top down/associative design is allready feasible with freecad (I have year of creo and inventor experience). Will be lurking the fora for the next hour or so

[eddriesen]

So here's my first structure and MO;

• Partcontainer with body container

• Import of obj tinkercad data

• attach/move data to zero of body origin

• build part

Afterwards I'll move the parts around so they appear properly assembled.
I was trying to use assembly2, but I got a errors ( cloning github version manually, and using the addon manager) . Looking at the "upstream" versions, they require some tinkering. While I'm not that afraid of tinkering, I fear getting assembly to work might be to daunting for some users that "just want to contribute". Also, if done with a deprecated version of assembly,things get broken, unusable.

I'm trying to copy them literally (rounding off some dims here and there, measuring on mesh is a pita). No optimising whatsoever, so we have a baseline of the original designs. The improvements are up to the community when Freecad files are up!

Anyone wants to chip in on advice?

[openBrain]

Looks like a cool project. I think many regulars here will help by modeling, reviewing or commenting.
The most important thing is IMHO that we have a clear pointer on a spot where last modeling can be found, and eventually where modifications can be submitted.
Regarding modeling, my opinion is that we should take care that it is as most parametric as possible, and also robust to future modifications. Also think about aptly naming each value so that design is highly maintainable.

[eddriesen]

Looks like a cool project. I think many regulars here will help by modeling, reviewing or commenting.
The most important thing is IMHO that we have a clear pointer on a spot where last modeling can be found, and eventually where modifications can be submitted.
Regarding modeling, my opinion is that we should take care that it is as most parametric as possible, and also robust to future modifications. Also think about aptly naming each value so that design is highly maintainable.

I'll post it on a forked github when a few things are done.
Collaborative, working with a single project file doesn't look practical. But see my assembly thoughts above.

[kisolre]

Anyone wants to chip in on advice?

We are seeking help from the community

I will have to repeat myself. If you need help with the actual modeling and recreating of the design STL and OBJ would not be enough. Taking mesurements from them is PITA and having to guess is a curve is an arc, elitical arc, hyperbola,... is too. If there is an angle is it important wor the design or just for looks? In the "neck" piece there is a 24.1531 deg angle between the outer surfaces and 24.24 deg angle between the cut surfaces. Are those important or are they derived from some other dimensions? Top surface is created from eliptical arc with one redius 10mm but the other measures around 6.95 - is that 7.0 or is it derived from other dimensions?
So any info of the required dimensions will be helpfull.

[eddriesen]

Yup,

+1 on the original files. Reverse engineering meshed parts is always a major pita.

Any help is appreciated. If I can copy the tinkercad project i can have a peek at dimensions.

[vocx]

...
I was trying to use assembly2, but I got a errors ( cloning github version manually, and using the addon manager) .
...

Do not use Assembly2. This workbench is no longer maintained nor developed. Its successor is A2plus Workbench, which is derived from the former, but is not backwards compatible. A2plus is relatively stable and is recommended by most users. However, if you don't need true assembly capabilities you may just position the bodies in space instead, with the help of other external workbenches.