yaobls case

A cheap logic analyzer. Get one for $50, including worldwide shipping. A collaboration between the Gadget Factory and Dangerous Prototypes.

yaobls case

Postby dsm » Sun Jul 10, 2011 4:31 pm

yaobls case

You might want to skip ahead to the "clamshell" yaobls case postings that start 13Sep11 on page three of this topic.

I've noticed a number of Open Bench Logic Sniffer case designs on the Dangerous Prototypes site. I recently decided to design the yaobls (Yet Another Open Bench Logic Sniffer) case in order to gain some experience with the tool chain that I want to use to fabricate case prototypes for another project.

As an experiment I created eight yaobls case variants by changing some parameters. These were uploaded to the Shapeways site under /shops/dsm . If you are interested, you can order your own copy in a variety of materials. The yaobls case in red strong & flexible SLS nylon was quoted to arrive in just eight days. The Shapeways site allows you to view and rotate the yaobls case variants in 3D.

The yaobls case variants include the following:
  • TF ~ yaobls-case-typeA-tall-full-dsm-v4
  • TFS ~ yaobls-case-typeA-tall-full-skel-dsm-v4
  • TP ~ yaobls-case-typeA-tall-partial-dsm-v4
  • TPS ~ yaobls-case-typeA-tall-partial-skel-dsm-v4
    These cases have been replaced with yaobls-case-dsm-v5 (link).
    See 05Aug11 posting for more details.

  • SF ~ yaobls-case-typeA-short-full-dsm-v4
  • SFS ~ yaobls-case-typeA-short-full-skel-dsm-v4
  • SP ~ yaobls-case-typeA-short-partial-dsm-v4
  • SPS ~ yaobls-case-typeA-short-partial-skel-dsm-v4
    These cases have been replaced with yaobls-tray-dsm-v5 (link).
    See 05Aug11 posting for more details.
The following handy dandy decoder explains the features of the above yaobls case variants.
  • typeA ~ boards slide in from side.
  • Tall ~ one lower slot for the OBLS board plus one upper slot for a laser cut acrylic cover (or another board). The height of the Tall case means that an 8.5 mm high female wing connector will be 1.0 mm above the acrylic cover (providing better access) and 1.0 mm below the top frame border (providing some protection).
  • Short ~ one slot for the OBLS board. The height of the Short case is limited by the height of the USB connector opening.
  • Full ~ top frame border on four sides. Provides some additional support at the open end of the case.
  • Partial ~ top frame border on three sides. Makes it easier to slide tall components into the case.
  • Skel ~ material removed from bottom surface to reduce fabrication costs (roughly $4.46 per case).
  • v4 ~ increased margin on the USB connector opening versus the previous v3 series.
  • feet ~ All units have mounting wells for feet on the bottom surface. Use 5/16" x 3/32" clear bumper pads (rubber, polyurethane, cork, felt, etc.).
  • top cover ~ The top cover is intended to be made out of 1/16" laser cut clear acrylic. The top cover is the same size as the OBLS board and positioned directly above the OBLS board to make it easier for you to design your own top cover.
Any feedback on the yaobls case design you can provide would be greatly appreciated.

Thanks for your time.

dsm
Last edited by dsm on Tue Sep 13, 2011 9:36 am, edited 9 times in total.
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby arhi » Sun Jul 10, 2011 6:12 pm

In the spirit of open source that OBLS is all about you might want to share all the designs (here or for e.g. on thingiverse) so that ppl can make the case themselves using your design
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: yaobls case

Postby ian » Mon Jul 11, 2011 4:42 am

Here's the direct link, the spam filter will let you post links after 24 hours:
http://www.shapeways.com/shops/dsm

Do you plan to release the design (as open source)? Any pictures of a completed design :)
User avatar
ian
Crew
Crew
 
Posts: 10803
Joined: Mon Jul 06, 2009 6:14 am

Re: yaobls case

Postby dsm » Mon Jul 11, 2011 11:56 am

Errata
The URL addendum in the second line of the second paragraph of my initial post was initially incorrect.
Ian's response points out the correct complete URL.



Arhi,

I do intend to release the next version yaobls case design under one of the Creative Commons licenses, but I have to consider the current "-v4" series version to be an intermediate draft of the design that I have uploaded to Shapeways mostly to solicit feedback and new feature requests. I would like to introduce any relevant suggestions / feature requests into the "-v5" yaobls case design, build some prototypes to verify the changes, and then release the results. The fact that there are eight members in the "-v4" series suggests that I have some very conflicted ideas about what people might find to be useful.

As I mentioned in my initial post, the rationale for the yaobls case design was to gain some experience with the tool chain I want to use to fabricate case prototypes for another project - from my viewpoint the yaobls case design is just an experiment in partially automated parametric case design.

There are several things that would be very helpful to me.
    1) feedback and new feature requests for the "-v5" series...

    2) suggestions about places to release the results...
      Shapeways, Thingiverse, Ponoko, etc. ~ 3D prototyping services?
      Seeed Studio, SparkFun, Gadget Factory, etc. ~ places that sell the OBLS board?
      design repositories?
      See 28Aug11 posting for more details.
    3) suggestions about which Creative Commons license to use...
      See 28Aug11 posting for more details.
Arhi, your post on 11Jul10 titled OBLS - case - print your own got me thinking about what example to use to experiment with my tool chain. My recollection is that you own or have access to an FDM (fused deposition modeling) 3D printer. I would be interested in getting some FDM prototypes in ABS built to compare against the SLS (selective laser sintering) prototypes in nylon that I am getting through Shapeways.

I currently do not own a MakerBot, RepRap, or other 3D printer. Since a MakerBot (as a kit) costs roughly 50 yaobls cases (not including the ABS material) and I have access to 3D printers through the local TechShop, I have to give this purchase lower priority.



Ian,

I would be more than happy to post some pictures of the actual yaobls case (including the laser cut top cover) and a family portrait of the "-v4" series yaobls variants after the 24 hour hold expires.

Thanks again for your helpful comments.

dsm
Last edited by dsm on Sun Aug 28, 2011 7:19 pm, edited 5 times in total.
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby dsm » Mon Jul 11, 2011 7:22 pm

yaobls pictures

actual "-v3" series yaobls case prototype (on flickr)
Shown in Shapeways "black strong & flexible" SLS (selective laser sintering) nylon material. OBLS board in lower slot and laser cut top cover in upper slot. The main difference between "-v3" series shown and "-v4" series currently on Shapeways is an increased dimensional margin on the USB connector. Note that the little detail over the logic connector is only 2.0 mm by 5.0 mm, but it seems completely robust in SLS nylon.

family portrait of yaobls variants (on flickr)
    top row
      TF ~ yaobls-case-typeA-tall-full-dsm-v4
      TFS ~ yaobls-case-typeA-tall-full-skel-dsm-v4
      TP ~ yaobls-case-typeA-tall-partial-dsm-v4
      TP ~ yaobls-case-typeA-tall-partial-skel-dsm-v4
      These cases have been replaced with yaobls-case-dsm-v5 (link).
      See 05Aug11 posting for more details.
    bottom row
      SF ~ yaobls-case-typeA-short-full-dsm-v4
      SFS ~ yaobls-case-typeA-short-full-skel-dsm-v4
      SP ~ yaobls-case-typeA-short-partial-dsm-v4
      SPS ~ yaobls-case-typeA-short-partial-skel-dsm-v4
      These cases have been replaced with yaobls-tray-dsm-v5 (link).
      See 05Aug11 posting for more details.
Enjoy!

dsm
Last edited by dsm on Sun Aug 28, 2011 7:26 pm, edited 6 times in total.
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby arhi » Tue Jul 12, 2011 7:30 am

dsm wrote:I do intend to release the next version yaobls case design under one of the Creative Commons licenses,


You do not have to do it. It'l be cool if you do (in the spirit of open source whole project is about).

dsm wrote:solicit feedback and new feature requests.


the board is fairly simple and you need access to 2 or even 3 sides of the board + buttons and led's hence the "cover the bottom" approach to "case" design seems to be best. I designed it to be "super easy to print with any 3d printer" and to satisfy single requirement I had - to protect the bottom of the board from short circuiting with junk on my desktop. Your design for e.g. can't be printed on a printer that do not have support ability (sintering process have the "powder" as a support so can print overhangs, fdm process can't lay plastic layer in the air) the way you positioned your object so it would have to be printed by positioning usb port on the print bed (so vertically) - of course only the version with "full" back could be printed as again the holes would look ugly (overhangs).

dsm wrote:experience with the tool chain I want to use to fabricate


What do you use to design the objects? The target file format in almost all cases is STL. Then you slice the STL on the software for the particular printer you are going to print it on. The source file format differs. Currently the most popular in the open source community (and perfect for objects like this case and any case for that matter) is OpenSCAD. I know that for a regular designer writing "code" to make object ain't fun nor "normal" but the learning curve is super steep and it is incredibly great tool for objects like project cases :). Other free popular tools are Art of Illusion and Blender while from commercially available tools most popular ones are CoCreate PE (free personal edition, they btw changed the name and now are called differently) and Alibre design (not free but they have some cheap entry level licence, 100$ iirc). I personally use openSCAD for cases and other "single objects" and SolidWorks for complex designs.

dsm wrote:suggestions about places to release the results...

put them everywhere :)

dsm wrote:Arhi, your post on 11Jul10 titled OBLS - case - print your own got me thinking about what example to use to experiment with my tool chain. My recollection is that you own or have access to an FTM (fused deposition modeling) 3D printer. I would be interested in getting some FTM prototypes in ABS built to compare against the SLS (selective laser sintering) prototypes in nylon that I am getting through Shapeways.


I am actually a member of a reprap core team. I'm involved in reprap project for many years, own 3 printers (one darwin, one mendel and one of my own design, designing two new ones). I am actually on "vacation from my hobbies" for few months already and will go back to finish designs of the new printers when I come back from this few months long "sabbatical" in August. If I can help in any way, let me know. Btw FTM/FDM "print type" is a registered term and stratasys is very quick on trigger to sue anyone who use that term so try to avoid using it if you are not talking about stratasys.

Anyhow as I mentioned, if you put your design on thigiverse, if the design is any good you will soon see pictures attached to your object of "ppl who printed the design". For e.g. if you check things tagged obls on thingiverse: http://www.thingiverse.com/tag:obls you will see my original design and Madox's design as printable designs. Then you will see a laser cut design and derivative/upgrade of that laser cut design. Not too many designs but still ... more then one :) ... if you on the other hand check on bus pirate: http://www.thingiverse.com/tag:buspirate you will find few interesting cases there too :)

I don't see ppl purchasing cases from RP sites in near future. The service is imho too expensive for what it offers in this case because you do not require a special case for these devices. A single piece of old plexi, veroboard.. and a piece of double sided foamed tape will solve the "issues" one might have (shortcuts etc), not to mention that one that play with electronics know where to get plastic cases for his devices and you can get a nice plastic case for few bucks (looking trough some cases on my desk now - I could put obls in a 4$ plastic box without a problem) and I doubt you can get a RP service to print you a part for 4$ (who knows maybe you can).
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: yaobls case

Postby dsm » Thu Jul 14, 2011 5:01 pm

FDM versus SLS Comments

I have to preface the following comments by saying that I am collecting parts to build my own reprap mendel 3D printer since I've been extremely impressed by what I've seen on the reprap site. Because I don't currently own a 3D printer, however, I've been using an online SLS 3D fabrication service provided by Shapeways.

Using an online SLS (selective laser sintering) 3D fabrication service has both advantages and disadvantages over using an FDM (fused deposition modeling) 3D printer (online service or personal commercial or personal open source). In my opinion, the advantages of using an online SLS 3D fabrication service include the following:
    Material choice ~ FDM printers use one of several thermoplastics (ABS, PPSF, PC, PEI, etc.).
    SLS printers can use thermoplastics like ABS and other materials such as nylon, alumide, high detail plastics, steel, silver, glass, sandstone, ceramics. Shapeways materials
    Material shrinkage ~ The thermoplastics that FDM printers use shrink as they cool. Even with a heated platform, determining the final dimensions of a part can be more of an art than a science.
    SLS printers use media that cools before the next layer is added, so the dimensions of a finished part can be accurate, predictable, and repeatable.
    Material support ~ Thermoplastics over openings need support until the material cools.
    These support structures then need to be removed in a later manual post-processing step.
    SLS printers have no such requirement since the unfused SLS media provides the support structure.
    Material strength ~ While the thermoplastic materials are quite strong, the bond between layers is usually less so. This weak bond makes for parts that tend to break at the layer boundary, but substantial structures with more bonding surfaces have fewer problems in this regard.
    SLS materials are strong and the bond between SLS layers also appears to be quite strong.
    I have been quite impressed with the tumble-polished SLS nylon materials I have been using.
    Material feature size ~ My understanding is that the minimum FDM printer feature size is determined by the size of the extruded bead, while the minimum SLS printer feature size is limited by the size of the laser beam and the SLS media.
    No hardware costs ~ Being able to produce 3D parts without first spending money to build or buy a 3D printer makes it easier to get started in 3D fabrication.
    Fewer design limitations ~ Being able to design parts without thinking so much about process (e.g. support structures, design orientation, shrinkage, strength, etc.) makes it easier to get started in 3D fabrication.
In my opinion, the disadvantages of using an online SLS 3D fabrication service include the following:
    Material cost ~ SLS materials can be quite expensive. SLS nylon costs 2 or 3 times what FDM ABS costs. This disadvantage is partially ameliorated by being able to use stronger materials allowing thinner parts.
    Response time ~ Personal FDM printers produce parts in minutes. Online SLS printer services produce parts in days or weeks. This is a huge disadvantage.
In summary, there are tradeoffs between paying someone to perform 3D fabrication for you and spending the money to build an open source 3D printer which vary for each person. Online SLS fabrication services can produce small, strong, accurate, detailed parts in a wide range of materials with very little overhead, but the length of time these online SLS 3D fabrication services take is a huge problem.

I think that using a online SLS 3D fabrication service allows me to design parts to be just the way I envision them to be without as much concern about the limitations of specific 3D fabrication processes. Sometimes a person just wants to order a nicely designed part and not think about process. My initial designs have been uploaded to Shapeways/shops/dsm to enable me to obtain the parts I want to build. If my designs are of use to anyone else, that's just a positive side effect.

My $0.02 worth.

dsm
Last edited by dsm on Fri Jul 29, 2011 9:53 pm, edited 1 time in total.
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby arhi » Thu Jul 14, 2011 5:54 pm

I used SLS (not online service but a local shop) for some parts when I needed some level of details I cannot print with my own printers and they absolutely have many benefits in prototyping industry compared to FDM but

1. the problem with layer adhesion really does not exist if you know how to use your machine (and you have a good machine) - there is no issue there - it is something you have full control over (strength of your design is very repeatable and is you who decides how strong the part will be)
2. strength also depends only on the material used, I print a lot of HDPE .. there's hardly a plastic to match it, unfortunately apart from being super strong and resistant material and from being 20-30 times cheaper then ABS/PLA .. it is incredibly hard material to print with (shrinkage, warping ..). If you think about your object while designing it you can make it much stronger then when using same thermoplast with SLS process. (of course SLS can sinter metal and that's another story all together :) )
3. repeatability - this does not exist as a problem, yes some materials shrink but always exactly the same so again if you have a good printer repeatability is not an issue
4. warping/shrinkage - if you do not need parts that need to work in high temp env (90C+) there are materials (PLA and some other) that do not shrink at all
5. materials - all thermoplast you can SLS you can also print, the SLS benefit comes when you use metal, that you can't print (yet :D)
6. support - I can print "any" object in reallity as I print soluble support on my reprap so that's really just about what printer you use.

Again, SLS is more precise then FDM (it is order of magnitude more precise then best fdm printers out there and 2 or more orders of magnitude more precise then Mendel) but is also order of magnitude (or more) more expensive then FDM. That is basically only real difference. If you need high precision fdm is not your technology, but if you need "fast and cheap" FDM can't be beaten. I get 10kg roll of polypropylene for 10$ ... do you have any idea how many boxes, brackets, corners, joints... you can make from 10kg of polypropylene ... now if I need more precision (PP shrinks so sharp corners are almost impossible to print) I can use ABS (with heated chamber and heated bed - no shrinkage - no warping) or PLA (transparent trough semitransparent to opaque - no warping, no shrinkage even when printing on room temperature), unfortunately we are talking now 50eur for 3kg of filament. Again, do you have a clue how much stuff you can print with 3kg :D.

Your box is for e.g. a great example of what I'm talking about. For my OBLS I don't really care if the edges are perfect nor if the surface is super smooth, but I do care if a box cost me 10c or even less (the box is few grams only, made out of abs that is under 1.5c per gram) compared to 20E + postage that I'd have to pay to shapeways.

On the other hand this (CCTV surveillance camera cufflinks) is a perfect example of where FDM isn't technology of choice.

As for the "price for the machine", if you go with mendel mod's like prusa mendel you can make one for under 500EUR (depends on where you live, in usa you can get it way below 500$ especially if you join the seeding mendel group)
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: yaobls case

Postby dsm » Thu Jul 14, 2011 6:40 pm

CAD Tools
arhi wrote:What do you use to design the objects?

In order to create 3D models, lately I've been using
  • Autodesk Inventor Professional 2011 (mostly due to a local TechShop promotion).
  • PTC Creo/PE 3.0 (mostly because I have a friend who is an expert).
  • Dassault Systèmes SolidWorks (mostly because that's what my schematic package likes).
I've been also playing with
  • Google Sketchup with a plugin from CADspan
  • OpenSCAD from link (because of your suggestion)
  • Blender 2.5 from link (mentioned by Adrian Bowyer)
  • TinkerCAD from link
Generally I've been going down the list of programs on the reprap, MakerBot, and Thingiverse sites to find something I like. Your suggestion about OpenSCAD was useful because I'm interested in partially automated parametric case design and OpenSCAD allows a simple way to tweek the internal parameters of the model.

Thanks for your help.

dsm
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby dsm » Sat Jul 16, 2011 11:48 pm

reprap 3D printer accuracy questions

arhi,

Thank you for your comments about FDM materials (strength, cost, accuracy, other properties, etc.) and FDM printers. As I may have mentioned, I've been collecting parts to build my own reprap 3D printer and your comments were very helpful. Somehow I am not at all surprised that you have been contributing to the reprap open source effort and helping to design improved 3D printers.

Although it's a bit off the topic of this message thread, I was particularly interested in your comments about the implied differences in accuracy between the best commercial FDM 3D printers and reprap 3D printers.
Do you have any ideas about the root cause of these differences in accuracy? Like any moderately complicated mechanical device, I would expect a reprep 3D printer to have sources of error such as the following:
    overall frame stiffness and strength
    belt drive stretch and slippage
    bearing type
    track smoothness
    extruder design
    temperature control
    media size and type
    NEMA stepper motor size and type
    platform heater
    printed versus machined parts
    firmware issues
    hand-built versus factory built
No doubt there are many other error contributors other than the ones listed above, but do you have any idea what factors predominate? Or perhaps the differences in accuracy are due to some concatenation of factors?

Assuming that I wish to build as accurate an open source 3D printer as possible, are there component choices I can make at this stage of construction that will make a significant difference in accuracy?
Is there a significant difference in accuracy between the reprap Mendel and the reprap Prusa designs?

Nominal Mendel Specifications (from reprap site) (Corrected)
  • Diameter of nozzle ~ 0.5 mm ~ 0.020”
  • Minimum feature size ~ 2.0 mm ~ 0.080”
  • Positioning accuracy ~ 0.1 mm ~ 0.004"
  • Layer thickness ~ 0.3 mm ~ 0.012”
Regards,

dsm
Last edited by dsm on Fri Jul 29, 2011 9:41 pm, edited 5 times in total.
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby dsm » Sun Jul 17, 2011 1:27 am

parametric case design

As I mentioned in my initial post, the rationale for the yaobls case design was to gain some experience with the tool chain I want to use to fabricate case prototypes for another project - from my viewpoint
dsm wrote:the yaobls case design is just an experiment in partially automated parametric case design.

I own a couple of the Saleae logic analyzers. One of the things I like about their design is the nicely CNC machined black anodized aluminum case that protects the electronics.

I mention this because I believe that many Dangerous Prototype designs would benefit from a little more thought about the overall packaging scheme during the specification, component selection, initial design, and board layout processes. For example, setting aside a little space on the board for mounting bosses (or other mounting schemes) would be useful. The high-level result would be better protected designs that would be easier to use so that more people would use them.

As an example of what I mean by partially automated parametric case design, suppose you could specify a few clam shell case parameters...
    Board length
    Board width
    Board thickness
    Space above board
    Space below board
    Wall thickness
    Edge radius

    Draft angle (optional)
    Lip detail (optional)
    Lip offset (optional)
    Snap finger detail (optional)
    Boss details (optional)
    Rib details (optional)
    Texture (optional)
    Feet well detail (optional)
and expect most of the hard part of creating a custom case design to be completed. Manually extruding holes for connectors, buttons, labels, battery doors, and display screens is comparatively simple. An STL file of the resulting design ready to be fabricated would be generated.

Regards,

dsm
Last edited by dsm on Tue Sep 13, 2011 4:44 am, edited 5 times in total.
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby honken » Sun Jul 17, 2011 11:36 am

I was out in the same errand as you when I did my interpretation of a BusPirate case design http://dangerousprototypes.com/forum/viewtopic.php?f=28&t=2515.

I decided to make it parametric with the distance between mounting holes as the main parameter.
It worked out pretty good, as the final exterior size of the box most likely is a secondary requirement.

My other parameters are excess from mounting hole to inside edge, material thickness and height.
Below the "landing pads" for the pcb is only a space of equal height to the material thickness.
User avatar
honken
Full Member
Full Member
 
Posts: 190
Joined: Fri Feb 26, 2010 10:34 am
Location: Stockholm/Sweden

Re: yaobls case

Postby dsm » Mon Jul 18, 2011 6:26 pm

Bus Pirate case comments

honken,

I started thinking about typeA cases (where the board slides in from the side) because I had been looking at arhi's tray design and later madox's tray design. I used an SLS printer via Shapeways instead of an FDM printer because
    I currently don't own an FDM printer.
    I wanted to implement some fine design details that would difficult with an FDM printer.
    I wanted to implement some design details that would difficult with injection molding.
    I wanted to experiment with partially automated parametric case design.
I've since decided that typeB clam-shell cases (where the board is held between top and bottom case shells) provides a better solution. As recently as early June 2011, Seeed Studio carried a small inexpensive Bus Pirate compatible case of this type, but I don't see it on their website now. Comment. I think of your Bus Pirate case as similar in concept, but more of a complete case shell and flat top cover (with integrated mounting bosses).

Honken wrote:In my head I see this milled out of aluminum and anodized black, perhaps with some silk-screened white ink...

Before you go to the expense of CNC machining and black anodization and since you have the relevant STL files, you might want to try fabricating your case prototype using an SLS printer in black nylon. You can even slightly indent any lettering and fill the letters with white ink (seal the letters first to prevent leakage since some materials are porous).

Once you have your initial prototype in hand, you'll probably think of other details and features that could have added. One of the really powerful concepts about additive machining is that adding more features and details is less important than the amount of material used. The situation is reversed with subtractive machining where each additional feature and detail means more processing steps (and thus cost).

Regards,

dsm
Last edited by dsm on Fri Jul 29, 2011 1:39 pm, edited 2 times in total.
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby dsm » Tue Jul 19, 2011 12:15 am

possible yaobls "-v5" series features

The yaobls case designs that I uploaded to Shapeways at /shops/dsm were from my yaobls case "-v4" series. One of the results I had hoped to get out of this action were
dsm wrote:feedback and new feature requests for the "-v5" series...

Although ian, arhi, and honken have been kind enough to respond to my posting and I can infer a little bit from what yaobls cases people seem to be ordering (see below), I am still left wondering what features people might find useful for OBLS and other Dangerous Prototypes designs. If anyone has any additional suggestions, now would be a good time to make them.

The following is unordered list of some of the ideas that I've been considering:
    typeA versus typeB ~ Although I believe that typeB cases (where the board is held between top and bottom case shells) provide a better solution in most situations than typeA cases (where the board slides in from the side), I want to finish up my typeA efforts first.
    increase slot thickness ~ Since there is some run-to-run and material-to-material dimensional variation using an online SLS 3D fabrication service, the slot thickness should be increased from 1.7 mm to 1.8 mm for more margin.
    alternate slot retention schemes ~ The OBLS board and the laser cut acrylic cover are currently held in place by friction (possibly aided by a discrete drop of glue). Since there is some run-to-run and material-to-material dimensional variation using an online SLS 3D fabrication service, I would like provide an additional way to retain the OBLS board and the acrylic top cover.
    a) Reducing the slot thickness (currently 1.7 mm) or slot width seems like a bad idea (see previous).
    b) I experimented with crush ribs, but didn't like the results.
    c) I could provide two bosses that match the two mounting holes near the external wing connector. Unfortunately this might interfere with currently unpopulated connectors.
    d) I experimented with little flexible grabber fingers and liked the results. The grabber finger support, although easily strong enough, would need to be a bit stiffer (by changing the width and length of the support) than my experiment (which was only 1.5 mm by 1.5 mm by 15 mm).
    clear case ~ While Shapeways provides a translucent frosted ultra detail material, FUD cost is quite high and FUD is not very transparent. On the other hand, FUD might be a good material for custom LED light pipes which would allow a completely closed case. This would eliminate the laser-cut clear acrylic cover fabrication step.
    The laser-cut clear acrylic cover is probably the best choice for now.
    A future injection molded case could use a transparent plastic cover.
    thinner walls ~ While Shapeways specifies a minimum wall thickness of 0.7 mm when using the SLS nylon material, I've been using a minimum wall thickness of 1.5 mm for the slot. Thus the case wall thickness could be reduced somewhat.
    wall features ~ More bottom case holes, indented text (example), and indented texture (example) could reduce the material used.
    case versus tray ~ arhi pointed out that developers might prefer a tray design to a full case design for greater access and to reduce costs. Replace all "short" cases with a "tray" design.
    eliminate all "non-skel" cases ~ Eliminate all cases that don't have holes in the bottom of the case to save material costs.
    top frame ~ The top frame around the acrylic cover is currently 5.0 mm wide largely for aesthetic reasons. The inside edge of the "-v3" series top frame had a 1.0 mm chamfer while the inside edge of the "-v4" series had a 1.0 mm radius. Using a 1.0 mm chamfer for the inside edge of the top frame would make it easier to modify a "full" top frame to a "partial" top frame since the chamfer could be used to guide a razor blade.
Thanks for your time.

dsm

By the way...
    The yaobls-case-typeA-tall-full-skel-dsm-v4 design in red strong & flexible SLS nylon
    (which is also my favorite) has been the most purchased "-v4" series yaobls case so far.
Last edited by dsm on Fri Aug 05, 2011 9:02 pm, edited 6 times in total.
dsm
Full Member
Full Member
 
Posts: 216
Joined: Sun Jul 10, 2011 3:04 pm

Re: yaobls case

Postby arhi » Mon Jul 25, 2011 2:18 pm

sorry for the long delay but I today returned from beautiful island of Crete with my batteries recharged :) and there I tried to have as little to do with computers as I could :D (the only thing running on electricity was the fridge) :D

dsm wrote:Somehow I am not at all surprised that you have been contributing to the reprep open source effort


hahaha, they even made me a core team member some years back :D
I am now back from a few months long sabbatical from all time consuming hobbies (including reprap) so I should be publishing more data soon :)

dsm wrote:I was particularly interested in your comments about the implied differences in accuracy between the best commercial FDM 3D printers and reprap 3D printers.
Do you have any ideas about the root cause of these differences in accuracy?


It is not a mistery why the "best commercial" ones are more precise. There are 2 sets of issues
1. design
- they use 1.7mm filament by default compared with 3mm default filament reprap uses. This provides much more accurate way of feeding filament. Just compare to soldering SMD with 1.5mm solder and 0.5mm solder wire :)
- they use stiff machines made from quality steel that don't "bend" nor "deform" with heat
- they use heated bed + heated chamber (I do too with my machines but "regular" reprap like darwin or mendel, and regular repstraps like makerbot or shapercube don't) hence they do not have inbuilt stress in the object

2. calibration
- you need to calibrate your software (that generates G-Code) to your material. Compared to reprap users that use many different materials from many different sources (I for e.g. use PP, LDPE, HDPE, ABS, PLA on regular bases and have more then 5 different suppliers, I also tested many more thermoplasts) where even from same source two batches behave differently (especially with ABS) "professional" machines use always the same material. You are obligated to purchase material from the supplier of the machine at huge prices but the material is "always the same" and behaves "identical" every time. The software you get with those machines is aware of that thermoplast properties so output is always the same no matter what tray of material you put into machine. With "classic reprap" thermoplast repeatability is linked to the plastic you are using, so same G-code will not produce same part if you use ABS from BFB or from ReprapSource.
- ease of use, since the material on the "proff" machines is always the same all the advanced material stuff is hidden from user while on the reprap tools you need to expose all this to the customer hence you really need to understand both process and materials to setup and use reprap "best". Of course you can use "defaults" and get "average quality" parts but getting great parts really requires way more knowledge then with those proff apps

dsm wrote:[*]Diameter of nozzle ~ 0.5 mm, 2.0 mm ~ 0.020”, 0.080”
[*]Minimum feature size ~ 0.1 mm ~ 0.004"
[*]Positioning accuracy ~ 0.3 mm ~ 0.012"
[*]Layer thickness ~ 0.3 mm ~ 0.012”[/list]


can't really comment on the prusa vs sell's mendel variants as I really dislike both of the designs. On the other hand they are both impressive designs, super simple and they do produce impressive quality prints for such simple design. Be sure to get them printed in ABS as if you print them in PLA you can't really use them to print ABS later.

I use nozzle diameters of:
0.25mm (not easy to setup and operate)
0.35mm (experimental attm, works nice for me)
0.4mm (mine recommendation for the beginners, best quality vs ease of use)
0.5mm (classic nozzle, I almost not use these any more)
0.9mm (for stronger bigger parts that can be printed faster)
1.5-2mm (experimental)

Not sure what you mean by "minimum feature size", minimal layer height I print is 0.125mm, minimal wall thickness I print is 0.175mm but with this wall thickness minimal angle that will stay "sharp" will be around 50 degrees. If I want to go down to 20 degrees then I have to push minimal wall thickness up to 0.2mm. Everything else is dependent on these limitations.

Positioning accuracy .. hm .. this is not really an issue. Problem is repeatabitlity and it depends on the stiffness of your machine. My machine for e.g. has positioning accuracy of below 0.01mm for X and Y and below 0.004mm for Z

layer thickness, well 0.3mm is not nearly what you are looking at with high precision parts. For brackets and parts like mendel joints yes, 0.45 is even good enough but if you need real precission you need to get down to 0.25 or below. As I mentioned I print at 0.125mm layer height. Major problem here is not the Z positioning but extrusion control as at this height you reeeeeeeeeeeeeeally need to control your filament flow perfectly.

dsm wrote:parametric case design ...I mention this because I believe that many Dangerous Prototype designs would benefit from a little more thought about the overall packaging design during the design specification


The cnc'd anodized alu case you mention cost more then most of DP products so I must say I strongly disagree with you here :). If DP would to concentrate on making profit from few products (like saleae is) then yes, packaging matters a lot, but with the concept I see DP working on (new project every month, make projects where everyone learn something by making something usable, open ...) I don't see a place for a case that cost more then a project. For e.g. ordering DP products from SEEED cost almost the same price as if you make the boards yourself (I for e.g. got logicshrimp pcb from Ian and I ordered parts from local store, parts were only few euros cheaper then ordering the assembled and tested part from SEEED) - this show the profit margin they are working with and it show imho a lot about who and how is running DP - and that's why I like it (way more then saleae that encased a single mcu inside fancy box and are selling it super expensive)

honken wrote:I decided to make it parametric with the distance between mounting holes as the main parameter.

openSCAD is really perfect tool for making cases. I design almost all my cases in openSCAD. I believe you will find this interesting: MCAD. It is library of useful functions for openSCAD (like predefined objects, gears, rounding functions etc etc).

dsm wrote:possible yaobls "-v5" series features


it really depends on how you use the part. For e.g. for me any hole on the bottom is pointless as the whole idea of the closed bottom is to prevent shorts from junk on the table and having holes on the bottom actually allows junk to get stuck between board and the case. On the other side I don't see any reason for covering any part of the top of the board as you have points of interest all around the board (buttons, leds, ICSP, SPI, External Clock, Inside shield ..) and you want that fpga to stay cool. Now if you don't use any of the inside connectors you could close it up and expose only the buffered inputs but then you are not using the "whole device" :) and opening the device to access "advanced" functions every time you want them will soon lead to "losing the top cover" .. mine 0.02
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Next

Return to Open Bench Logic Sniffer