Thoughts on one way we can learn from the 800 lb gorillas of the industry, without betraying our ideals.
Originally shared by Whosa whatsis
While DRM for feedstock is definitely scummy (it’s a dick move when 2d printer companies do it with their ink too), I have been giving some thought recently to the cartridge thing. The printer should always be able to use anyone’s filament, but inconsistency between vendors is one of the things holding us back from “just click to print”.
We need better than 99% reliable feeding (I’d say most of us are closer to the 90-97% range, and getting there a thousand different ways) and we need better ways to match pre-calibrated profiles with plastics from various suppliers (and in different colors). We should really come up with a standard using some kind of UPC/QR/RFID/NFC/ETC codes with a corresponding printer peripheral to automatically set things like extrusion temperature and exact diameter for the specific plastic. With an open standard, we could make hobbyist printers with hobbyist-grade filament as reliable as the ones that cost 10x as much.
Only thing i fear is: http://xkcd.com/927/
I’d suggest removing “exact diameter” from the list. “Exact” is always defined within a given set of tolerances, i.e. variable.
We need smarter extruder controller designs that track filament diameter variation and closed-loop control the flow-rate to compensate. This will be really important as more people start to explore DIY filament creation and related options.
It should be possible to make a small device that can measure the diameter of the filament or feedstock before it enters the extruder and with the available io pins would tell the firmware how to adjust its settings on the fly…
Optics? Sensors? A pressure system wrapped around the diameter of the filament?
Feasible? Or to complicated?
People have been talking about real-time filament measurement for years, but nobody has found a way to make it work. I know this measurement can be done with a laser micrometer, but it’s not an inexpensive printer add-on.
What I’m thinking is that each spool should have its stated diameter (not just “3mm” or “1.75mm”, but accurate to 10 microns) and tolerance, ideally measured while spooling each individual spool. This process would also catch bad spools, like a 2.87mm spool that has one spot where it balloons up to 3.32mm and will jam your extruder (I’ve seen several of these now), though per-spool measurement could be a value-add for the better suppliers.
It would also be nice to have an objective measurement of the color (need input from the photography folks on how to keep them honest about this, and not just giving idealized values).
@Nils_Hitze That’s why we should come up with a standard as a community before any of the vendors start coming up with their own.
Kinda like an ANSI/ISO standard for feedstock. Shouldn’t be too difficult. You just have to get everyone or the majority to agree upon a set standard.
We really just need to come up with a good enough standard to make it not worth making alternatives like in the xkcd linked above. We don’t have to get everyone to use it, as long as they don’t start making alternative standards. This information will be a value-add, and vendors that include the information will be preferred over those who don’t by people with machines that can read it (adding a human-readable form for machines that don’t support it might be a good idea).
So how do you propose we start this endeavor?
We would have to find a manufacture that is willing to accept some sort of value-add standard and put it through a trial run. I’m sure if it got them more demand over the competition they would have no issue continuing such a standard.
Sorry, aside from coming up with a good enough standard.
We need to come up with a format for the data (in addition to what has been discussed, there should be data about formulation (at a minimum, ABS (are there designations for different formulations of ABS?), PLA 4032D/4042D/4042D, Nylon 618/645, etc.), and a URI for fetching more machine-readable data as well as for re-purchasing. We need to decide on a way of reading the data from the list above, and we need to interface a peripheral for it. Really, getting vendors to start using it is the last step, as the system can be tested with data tags made by the user.
Also, if we use a type of tag that is rewriteable like NFC, we can put on a measurement of the amount of filament that is updated while printing, so that you will know if a spool doesn’t have enough left to finish.
Well, there is already a ICC standard for color. One thing that is hard to capture is the color shift that occurs that could be temperature/dwell time dependent. In an ideal world, we don’t cook the filament at all. In reality, there could be noticeable shifts.
Filament diameter measurements that are “close enough” is already doable with low cost components - here’s one http://www.thingiverse.com/thing:70775 - I’ve been meaning to work on my own version using an Avago optical mouse sensor.
Plastic formulation alone isn’t enough, because the colorants and any other modifiers that could have been added will affect the plastic’s characteristics. Ideally, this is all measured and tested after each batch is produced, and the actual data used – but that’s probably not practical. I’ve said last year that an XML/binary format for plastic parameters would be useful. But to get people to actually buy into the data format (and suppliers to produce useful data) is a big marketing effort. This is where commercial consortiums usually have a better luck at getting industry wide adoption of a standard.
The problem with NFC or barcode or anything is that you’d have to get everyone to agree to standardize the spool design to make it all fully interoperable. No small task. It’s already a headache getting a universal spool holder! 
I’m sure this idea comes up again and again (https://groups.google.com/group/makerbot/msg/ca55d857644c8535?dmode=source) – the problem is that until there’s enough community interest to standardize, and money to help drive the adoption, it’s not likely to happen anytime soon. BTW, this is one of the reasons why I don’t think the Cube “DRM” cartridges are pure evil. I think there’s some value to the assurance that the filament is made to a specific parameter that is controlled – and that’s part of the argument for buying an official filament cartridge. Overpriced? Yes. Evil? Not so sure. Do I like it? No.