[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/diy/ - Do It Yourself

[Advertise on 4chan]


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


I want to 3D model the Genesis 6-button controller so I can make a customized version of the enclosure/buttons.
I looked into 3D scanning but the guys bloody want to charge me over $1000 to entirely scan the thing.

So without resorting to 3D scanning, is there any concievable way I can measure out the dimensions and aspects of this controller? I think I can get the rough max width+height by setting up some squares on a flat surface. But no idea how I can accuractely determine arc lengths, radii, etc. without some significant guesswork involved.

Better yet, is there any reliable 2D sketches/3D models floating around of this controller?

[spoiler]Would a Sega employee share these design details if I agreed to sign an NDA?[/spoiler]
>>
>>1940284
get a pair of calipers and start measuring.
>>
File: Gen_Width.jpg (72 KB, 768x768)
72 KB
72 KB JPG
>>1940286
Easier said than done.
Pic related is literally all I could measure with a set of calipers.
>>
>>1940293
NGMI.
learn about how to use references while measuring.
Like think outside the box.
You take a bunch of measurements with respect to a reference point and interpolate the curves.
You could trace on graph paper
You could do so many things to get this done.
>>
Thousands of fags have probably printed gay shit like that, you'd surely be able to find the measurements online somewhere
>>
>>1940297
I studied electronics mate, I have only the most utter basic understanding of how to measure shit.
I don't even know how to measure from a reference point.
I thought of tracing onto graph paper. But that just introduces another layer of human error (how close can the pencil get to the edge, etc).
>>1940308
The models I found are either 3-button controllers or just roughly modelled via guesswork.
I want to model something that is as true to the orignal design as possible.
>>
AFAIK you can buy phone apps to 3d scan objects.
>>
File: totally happening.jpg (13 KB, 480x358)
13 KB
13 KB JPG
Xbox One Kinect scanner
>>
>>1940284
Give up. Trying to reverse engineer aesthetic design like that is incredibly complex. seeing as you’re asking how to do it I can tell you that you don’t have the CAD skills to pull it off even if I told you how.
>>
File: CAD Test.png (54 KB, 857x448)
54 KB
54 KB PNG
>>1941959
>you don’t have the CAD skills to pull it off even if I told you how
I know how to use lofts and sweeps in Autodesk Inventor. I have modeled products made up of more recognizable geometric shapes, hell I can even make my own controller designs based on my own measurements, but measuring this genesis controller is a mystery to me.

I have tried tracing the controller as this anon >>1940297
suggested.
Imported the image into Inventor and slapped together some arcs (pic related). I'm not at all confident this is an accurate way to do this obviously and is just blind guesswork.
I'm assuming many if not all the measurements should be some round metric number?

I've thought of cutting the controller into slices and making measurements on the pieces?
I don't particularly care how complex it is, that's really for me to determine.
>>
>>1940869
shut the fuck up, dont spoon feed this retard.
>>
>>1941967
Why am I a retard?
>>
>>1941964
Why are you trying to get it 100% accurate?
> Imported the image into Inventor and slapped together some arcs (pic related). I'm not at all confident this is an accurate way to do this obviously and is just blind guesswork.
Yes, that's how this works. You make a guess, test it, find the differences, correct those, try again, and iterate.
Tracing an image and then fixing the scale using references is a good way to proceed.
>>
File: wireframe concept.jpg (51 KB, 823x497)
51 KB
51 KB JPG
>>1942176
>Why are you trying to get it 100% accurate?
Autism mainly. I want to be able to compare it side by side to an orignal OEM controller and have it look as if Sega themselves made the controller.
>Tracing an image and then fixing the scale using references is a good way to proceed.
I will go about doing it that way then I suppose.

No clue how I'm going to emboss this thing. I think I'm gonna have to get some sacrifical controllers off ebay and saw them into pieces and trace out a wireframe so to speak..
>>
>>1940293
>The guy charges a lot to 3d scan it
>Measuring it with calipers is harder than it looks
"Why am I paying so much for a plane ticket when I can just walk?".
>>
>>1942203
Is it really THAT difficult to measure a controller with a set of calipers/manual measurement tools?
>>
>>1940284
you'll probably be better off making a mold and modifying that. Or you could make it of clay, then slice it up and take measurements at several points and interpolate.
>>
>>1942208
it's the curves, not the fact that it's a controller, and the fact that OP didn't have the experience doing something like it before.
>>
>>1942395
What I'm not getting is why some anons on this thread are so apprehensive about OP's question.
I sure as hell don't know how to measure odd curves like that and I don't know what "experience" would teach you something like that. We come here to ask questions to do things ourselves, not to be told "lol get a job in a field that does that".
>>
>>1940284
Why don't you pay for the 3d scanning then release the results online so that the next person like you will have a resource? Pay it forward.
>>
>>1940312
You're probably more likely to find s Saturn controller or a knock off one. Just get the shell from one.
>>
Take it apart completely and cover each face in tape, then remove the tape sheet and you will have the surface area. Measure the height of all curves from a flat surface and you have all your dimensions.

Don't be a fucking retard.
>>
>>1940284
>>1942524

Is it that hard to 3D scan something by yourself? I've seen Shankmods et. al. use a regular 2D paper scanner to scan controllers and other objects to make CAD drawings. Is there software you can get that will let you 3D scan from a regular scanner?

OP, if you're gonna do this by tracing, you can use a 2D scanner to get high rez images of the controller, then use Inventor or Fusion 360 to trace over the features you want. Should be more accurate than hand tracing.
>>
>>1940284
Photogrammetry should give you a decent rough model
>>
>>1940284
Take a picture from all sides, trace it in Inkscape, import to 3D CAD, figure out scale, add depth. T. Don’t have a fucking clue, but that’s what I’d try.
>>
Anyone have good 3D scanner recommendations or resources to start learning about them in general? I'm planning on learning CAD for 3D printing. My brother is majoring in mechanical engineering and my dad is modeling shit for patients missing parts of their jaw. So I'll get plenty of info there. But they are at a loss when it comes to hardware (I've built them their PCs). What's a good place to start with getting a 3D scanner? Need one that's about 10-12 inches in diameter.
>>
ITS TIME TO GOTO THE MARIO
>>
somebody has modelled the controller fully, just use that ya dingus
>>
>>1940284
I think you're thinking of this the wrong way around. If you're intending to use the original board & buttons, you should be designing inside out; get the spacing of the board mount holes and button placement first, then work your way out. Don't worry too much about making it 1:1, just make it functional first.
>>
Are you sure you can't find a model of it already?
I can find 3d models for a ROB the robot, surely you can find this.
>>
>>1940284
Could you use photogrammetry with some free apps? I saw something advertised recently in gamedev circles for scanning your own objects for use in 3D environments.

https://all3dp.com/2/5-best-3d-scanner-apps-for-your-smartphone/
>>
You could replicate it with a mold (silicone?), then make a model from the mold (plaster?) and cut it up into slices. Do two models and you can cut one up in y direction and one in x direction.

trace the outline of the slices, scan them in and create your CAD model from that.

kind of like this:
https://www.youtube.com/watch?v=-zOIAVSV2Rc
>>
File: PCG-908A_2.jpg (46 KB, 500x500)
46 KB
46 KB JPG
Read this OP;
https://en.wikisource.org/wiki/Measuring_Tools
Also buy pic related to help you measure contours.
You can transfer that onto some paper, smooth out the rough edges, then you can use your middle school geometry set to do the rest.
Even if you did laser 3D scanning to the nth decimal place, you will HAVE to do some well educated guesses to round off the measurements to something Sega themselves would have scribbled on their draft drawings, if your goal here is to make an OEM replication.
And its likely all their dimensions were done in metric (millimeters).

Weirdly enough, my gf bought me this book for my birthday a while back;
>https://readonlymemory.vg/shop/book/sega-mega-drive-genesis-collected-works/
Which has some Sega hardware design drawings (unfortunately no controller drawings).
Flipping through it now, I see dimensions are in millimeters to 1 decimal place at most.
Standby though. I will try see if I can some halfway readable scans of the drawings. Maybe some other anons can make use of them, they sure as hell will never be of any use to me.
>>
>>1945663
https://www.threeding.com/3d-printing-models/sega-mega-drive-2-controller
>>
ill scan it for you I have a 5 axis point recorder. Ill do it for 25 euros.
>>
>>1942203
pretty retarded comparison.
>>
File: mdcontroller.jpg (3.44 MB, 2988x2742)
3.44 MB
3.44 MB JPG
>>1940284
OP I have a 6 button Mega Drive controller model MK-1653-50 and access to a 3D scanner. I'm not 100% certain but they should be the same shape
>>
>>1940293
take a top down picture of it next to a ruler, then try tracing over it in whatever drafting software you got
>>
>>1941964
how accurate does it need to be? whats your tolerance?
>>
>>1942198
oh, just thought of another idea, you could, since you're already willing to slice up the controller layer by layer, make a plaster or sillicoln mold of it instead.

If you want to keep it you could wrap it in plastic wrap before doing so, or even paper mache to make a mold, then make a copy from the mold, then slice that up
>>
>>1947111
99 cents. Anon delivers.
>>
>>1940284
How come 2 much money to 3d scan, 4chan do it 4 free

> Okay I 3d print and sell game controller

> What pla? Pla not good enough for controller

> Injection mold cost how much?!?!? I just want to bring product to market and nobody is helping me by working for me for free
>>
>>1940284
>measure some major and minor dimensions of components
>take high res pics from all sides
>turn pics to vector
>use measurements to scale it properly
>use 2D pics to make 3d model
>3d print prototype
>see what fits
>adjust
>3d print prototype
>see what fits
>adjust
>3d print .....
I guess you can figure out rest
after 3d prints max you should have all referance points figured out
stuff like screw holes for circuit boards etc can be measured
you can print outlines on paper and see if they match to
>>
File: StuCh7u.gif (1.13 MB, 195x229)
1.13 MB
1.13 MB GIF
>>1940312
>I don't even know how to measure from a reference point.
jesus anon
>>
>>1948809
The 3d viewer on that page downloads the STL, you can get it out of cache
>>
why are you all making sound impossible to measure it??
>>
Check out this video for an example of 3d scanning a similar controller.

https://youtu.be/iNqCLwoMBCk
>>
>>1940284
I have an actual controller, how may I help?
>>
>>1940284
figure out what diamter teh radii are and you're most of the way there
>>
>>1940284
I have that controller, it's in mint condition i love it. Do you want pictures? I could take them over graph paper of your like.
>>
>>1940284
Put the controller in a bowl of milk
Take a picture from the top down
Add a little more milk
Take another picture
Repeat
Congratulations you just scanned it in 3d
>>
OP here, thanks for all the feedback. I will address what I can.
To elaborate slightly;
I need to modify the controller to have a socket to hold a battery. I also need to add a cutout for a small LCD, and I want to make the letters on the buttons transparent so I can make the buttons glow RGB.
No idea how I'm gonna go about making the modified buttons tbqh.

>>1942394
>>1947578
Have considered this.
Moreover, considered modifying the parts as necessary by hand, then making a mold based off that.
Good call making duplicates so I can hack them though. I was considering buying a bunch of non functional ones off ebay.
>>1942524
Thought about it. Blowing $1000 is a little much for me though.
>>1942535
>Measure the height of all curves from a flat surface
I don't get what you mean by this.
>>1942896
A 2D paper scanner may work actually with tracing in CAD. I could lay a grid transparant sheet so I have a point of reference for scaling.
Thing is that the cotroller has some portrusions that prevents it from laying nice and flat.
>>1945653
I am making my own PCB from scratch. I'm adding a bunch of bells and whistles to "modernize" the controller in a sense.
>>1942898
>>1946476
Have tried this. Does not give me the accuracy I'd need. I guess it'd serve as a reference point ultimately.
>>1945550
>>1945663
>>1947111
Virtually what I could get from my own phone scans. It's very rough. But I digress, may be a point of reference at the very least.
>>1946667
Yes I have thought about the whole nearest decimal to round off to. If you're saying they measure up to 0.1mm accuracy, that would at least help me for rounding things off that may come off weird.
If you're still around, I welcome those scans of those drawings.
>>1947166
>>1947181
>>1952185
>>1952367
If you guys could 3D or 2D scan on your end, I more than welcome it.
I don't want to be a burden on you guys for my autistic ventures.
>>1952475
Honestly not a bad idea. I might actually try this out.
>>
Character limit.

>>1946484
I will look into trying this out as
>>1947578
>>1942394
Have pointed out. Forgot to add you to the first reply chain.
>>1948851
Guess and check does seem like the way to go it seems until I can get something remotely convincing.
>>1948904
The vernacular just threw me off.
The point I was misunderstanding was "what" that point of reference would be.
>>1950882
Watched the video.
Seems like despite even having access to a 3D scanner, he couldn't even get a perfect translation.
I guess those rough 3D scans don't seem so bad now if the tech isn't even up to snuff yet.
>>1952213
If I knew how to get radii, I agree this wouldn't be nearly as difficult as I see it to be.
>>
16
>>
>>1940284
>>1940293
>>1940312
>>1941964

Anon, this dude did the same thing you want to do, just check his vid.
https://www.youtube.com/watch?v=iNqCLwoMBCk
>>
>>1947575
This is exactly what I would do and then I would fit it to scale.

It wouldn't be perfect but it be good enough for a custom model.

3D modeling curves can be difficult sometimes but in reality it's not that hard if you're willing to try. I have a lot of experience though.

You might be able to grab model online and then modify that. But that's a big maybe considering all you'll probably find is STL files.
>>
>/diy/ doesn't know about 3DF Zephyr
I hope you have a good camera and strong computer OP.

Though, based on
>No idea how I'm gonna go about making the modified buttons
You sound completely clueless on 3D cad work while thinking a 3D printer is going to solve your problem. Have you even opened up the controller itself? Why are you trying to do it as one piece?
>>
>>1952481
I tryed but no luck :^(
The plastic is too shiny to get a good scan, the only solution being to spray it with primer. Not gonna happen I'm afraid
>>
Use a surface modeling software like Rhino 6. Take pictures of a straight on top view and side view to import orthographic views. Build curves based off of those views, match and connect them, and network surfaces. Looks like you only have to model one side and mirror it to get the form.

There are many lessons on youtube for surface modeling cars and other curvature complex products in Rhino using this method.
>>
>>1940284
Just make picture from 3 sides, put the pictures in the model space in the corresponding planes and try to make a surface model. You can resize it if you have the correct outline.
>>
File: A Button.png (183 KB, 854x689)
183 KB
183 KB PNG
>>1953391
I did see this posted earlier in the thread.
As it seems, 3D scanning won't get you a picture perfect model regardless of your setup. It would help with approximating everything though.
>>1954456
>I hope you have a good camera and strong computer OP
Decent computer, but no cameras aside my phone's. I will check out 3DF Zephyr.
>You sound completely clueless on 3D cad work while thinking a 3D printer is going to solve your problem.
3D modeling is really not my issue at all. It's determining the arc lengths/radii to a reasonably accurate degree.
I want to avoid 3D printing for the finished product entirely. I'll probably opt for a silicone mold and use a resin close enough to ABS.
With regards to me being clueless about the modded buttons; I was referring about the concept of keeping the letter clear, but the rest of the button black.
Might make it a 3 piece (button, clear insert, then some black insert to build the rest of the letter).
>Have you even opened up the controller itself? Why are you trying to do it as one piece?
Of course I'm modeling each piece individually. I'm not that retarded.
>>1954608
Dang. Thanks for trying though.
>>1954626
Haven't used Rhino before, only had experience with SolidWorks and Inventor.
An anon posted a video of modelling a car, I intend to take a look at that and more or less try apply this concept here.
>>1954687
Is one of the approaches I will try.
>>
>>1941964
>measuring this genesis controller is a mystery to me.
Take the shell apart, lay the shell half insider-down onto a flatbed scanner with a ruler in-frame.
Dump the outputted scan into CAD software and scale it to match the ruler in the image.

Trace the model to duplicate the internal features. If you want to compare your tracing work against the original just print the CAD 1-to-1 scale on paper and scan them both in the flatbed again and overlay one part of he image over the other. Or print the CAD onto a transparency.
So long as you model the inside of the parts correctly so that the internals will fit you can fudge the external surfaces and features a bit based on what you can measure. Measuring the outside accurately is less important and more difficult to accomplish without a CMM.
I've done the above process more than once and managed good results. One item was so large I had to scan it in three segments and stitch the scans together before doing the CAD work.
>>
>>1956412
Why not just make a silicone mold from the controller?



Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.