Full Disclosure. Everything in the video was quoted from Fictiv and accurate, but FDM prices do vary a good bit across various suppliers. As mentioned in the video you should shop around and find a vendor that fits your needs for whatever method is right for you. I do believe it is a bit odd to outsource FDM when the consumer machines are getting so good. You can print a part for a few dollars at home that would cost 5-50x more through a vendor. I don't sell services like this so I don't benefit if you buy a 3d printer or order parts through a vendor. Just my opinion on the topic of course so take it for what that is worth. I did some more quoting with other vendors and did find some FDM prices that started at around $100 per controller vs the $167 with Fictiv. Depending on the supplier you might also see greater volume discounts with certain types of fabrication. The best I have found still put Injection Molding about equal with FDM in the 250-500 controller range with an advantage to Molding quickly after that point. As we saw the difference between 50 and 250 molded controllers was very little. Check your vendors and quote around, but also talk to the places and question the prices. If something seems abnormally high they should be able to tell you why. I have found Fictiv to be very open about that when I have asked on other projects.
JLCPCB Can print the same controller in SLA for a fraction of the cost you were quoted, that's Laser Resin cured 3D Printer which is much superior when compared to FDM.
@@nicamarvin agreed. The pic in the thumbnail and the car and controller I went over were SLA prints. I am not sure if you watched the whole video but I did show sla, sls, and fdm prints in the video. Note that I have been providing feedback to fictiv on their FDM pricing relative to the competition to let them know where they are good and where they miss the mark. As mentioned you should shop around.
This is such a fantastic video, Matt! So much objective commentary that contrasts the piles of hyped-up content spinning FDM as a complete replacement for very well-established manufacturing processes that are the foundation of our modern lives and consumer experiences.
Thanks Matt, as a 3D printing service provider I’ve had to learn the skills of urethane casting and soon injection molding. One thing I would like to point out is the hybrid approach 3d printed mold to either urethane cast or injection molding. It’s not getting the backing I would like but I plan furthering the discussion with my clients and designer like yourself. I’m always hesitant to 3D print FDM parts for people and I make them sign a waiver saying these parts cannot be used in a functional application. Mainly because as you pointed out no matter the material there are still limited as an end use part.
Great point! I have used silicone based molds from an SLA print with good results. I think the issue we run into is the cost benefit ratio. In the cases I have personally used it I was able to get away with some pretty complex parts that couldn't be molded traditionally, but the silicone mold off the SLA part let me get by. So getting some nice ABS parts that way is great. The tricky part is how much better is that to say ABS like SLA or even an MJF print. In your experience is there a tipping point where it makes sense? like making 10-50 of something the print based mold is idea but past that you would go to say Urethane? My biggest issue with Urethane for my applications was UV. At the time the parts needed some funky colors to get the performance (for what reason i don't recall). If you want to send me an email support@caducator.com with your info I am happy to include it in the desc of this video if you like.
I think you made some great points about fdm and version changes being a negative for consumers. That is something that hasn't really crossed my mind. I'm usually happy with just being able to change stuff, but the constant changes of things can actually be quite draining for both me and the consumer. Being solid and confident on a design, and investing in a large run is a good skill to have.
Matt, excellent video and it’s fantastic you are teaching what it takes to make a product and show there is no silver bullet… especially FDM! We appreciate you and Motocilino’s work greatly, and I’m glad we could lend our controller to teach folks about manufacturing and design 🎮
I’m grateful for this discussion to be happening. It’s important for customers and engineering students to be aware of the range of fabrication methods at available to them. To push the objectivity further, and this is not to be a troll or compel you to be one, perhaps you can get a quote from “The Company That Shall Not Be Named” and Fictiv for a product that would be feasibly produced from each company. Perhaps TCTSNBN has advantages different from other FDM print farm competitors that make their prices lower. I’d be curious to know and I’d love to have the dialog about this continue. Thank you for your excellent content.
Thanks :) And so funny story about that. I did try to quote one of the parts through them. Was listed as 1-3 days and 'we will get back to you". I put everything I wanted in the initial contact form for material, quantity, layer height etc. Said orientation was at their discretion. 3 or 4 days later got an email with no quote numbers but asking me if I was sure about my material choice. Have I made it before, how do I want it oriented and yada yada. Compare that to Hubs.com or Fictiv or Xometry or whoever that has instant quoting. If it were me and I was running a print farm I would quote it just like the customer asked, then offer up some alternatives if that is your angle. My guess is they don't want to print ABS, which is fine. I was curious where the pricing would be but honestly waiting 3 days to get no numbers back was problematic. I do have some plans to do a company to company comparo in the future. I want to look at one process, same material and look at different quantities and turn around times. I will try to include them in that, but I honestly don't know where they fit into the spectrum. In a land of instant quoting, they are slower than quoting injection molding.
@@LearnEverythingAboutDesign Now that is interesting. I expected there to be an automated quoting platform. A strong advocate for a paradigm shift should reduce as many barriers to entry as possible to facilitate its adoption. Hopefully this dialog will lead to those changes. As someone who is familiar with SLA, FDM, and SLS, and an unabashed 3D printing prostyletizer, I am open to innovations that make 3D printing a more widely used fabrication method. I would also like clear-eyed analyses, comparisons, and critiques of fabrication methods so we can make knowledgeable design decisions and also understand where there is room for 3D printing innovation resulting in efficiency and reduced resource consumption.
@@lestersmith7748 Agreed. I wrote a lengthy response to his latest video but it won't take my comment for some reason. Maybe I have been blocked. I really can't stand the attitude and condescension while reading comments that oppose his view and not offering up any evidence or proof to support his position. Everyone can have opinions and that is good, but I have yet to hear an honest discussion with valid talking points about the process from him. For example the recent comparison between FDM and Molding saying that FDM has fewer inputs being power and material, while molding has power, material, and tooling. Glossing over the fact that FDM material has to be melted 2x and there is another machine needed to make the spool from the pellets. And completely misrepresenting the size and capacity of molding machines. Sure paper and pencil is fewer inputs than a computer so surely its a better option right? 3D printing nerd had a recent video at Formnext talking with nexa about 3d printed dissolvable molds for impossible to make geometry. A 3d printing company printing tooling for a molding machine for hard or impossible to mold parts seems like a great direction/improvement. To somehow claim that FDM would accel at printing an impeller blade where an SLA mold printed in 20mins and an injection mold machine which molds the parts in seconds (granted you have to dissolve the mold over night) is some how going to be a better solution. ruclips.net/video/4kIeqTK5iao/видео.htmlsi=ay_pFx2VQCMMvsqd FDM is the slowest, most inaccurate, least cosmetic, and weakest out of nearly all other processes available.
Hecking smeg you make great content. Definitely gets my mind find clarification on a project I’m trying to picture and how it could be created. I have a good old frozen sonic 4K and have been using TR250 high temp engineering resin. It is designed for prototype injection moulding. If you look after the mould, it can last. I’m wondering where that type of resin printed injection mould comes in on the of scheme things. If you calculated that the resin injection mould could last for 200 items. Just brainstorming. Great channel. Please RUclips put more of this channel in my “for you” stream. 😂😎🤜🏼🤛🏼🍀🍀🍀☮️☮️☮️
Thanks for the kind words! I believe the 3d printing nerd channel had a video with Nexa showing an mSLA printed mold and a desktop sized injection molding machine showing not only the power of that process but using resin based molds for impossible or extremely difficult to mold parts in a traditional sense. I think that is where the real value of the process is right now. Aiding the molding process rather than replacing it.
Everything has it's place. For me, FDM is an amazing scratch paper where you can quickly develop ideas and get a better sense of all the dynamics of the proceess, get feedback and work on locking down the design. After that, mock it up with SLA, then lastly CNC.
I get what you're saying with the versatility in terms of design changes with FDM being a negative when it comes to sending out a design to be manufactured by a company but when you consider that FDM printers are pretty common among people designing and building at home it becomes a different story. I'd think the FDM design test phase is done in-house (or at home) and then beyond that, another method is used
For sure. FDM is great for concepts and testing. One big thing to consider is you really have to design for FDM. You aren't going to design for molding and then print it with an FDM. You could do that with SLS or SLA though. I love having an FDM at home, but I can't imagine selling and FDM product. Maybe the future will change my mind :)
Very soon! The intro to design for 3d printing has been edited, i just need to get it up on the site. For good or bad i just have way too much work on my plate at the moment.
Hi, sorry for the translation from French. Besides my work, I have a small company where I make parts in very small quantities for R&D in the food industry in FDM. (PETG) It's super cool, and my customers are really happy. but now some of these companies are asking me for spare parts normally molded in PEEK for machines in small quantities (2-6 units) (gears, outlet end etc.). I could reproduce them in CAD quickly and deliver them in record time with my SLS but I have a lot of questions about the resistance. In your opinion, would the resistance on a gear printed in SLS Nylon be as resistant as molded PEEK? If you have any information on this subject (comparative studies, durability test) I would be very grateful. Thank you for this video which was fascinating.
Sorry I don't have any specific info/papers/research on your specific question. Generally what I would say is to ask the manufacturer of your resin to provide technical information on that material. In general if we look at nylon vs PEEK there are some potential issues related to temperature. Nylon will get soft at a much lower temperature. So if you are looking at a gear, the friction from use could be enough to generate heat that would cause deformation. The root area of the tooth of a gear is generally small and if the surface was hot enough in a localized area that would be enough to fail. The surface finish on an SLS part (unless we are talking very large gears) is likely going to be a big performance factor vs a smooth molded part as well. Nylon in a cast or extruded application has a pretty low coefficient of friction, but that changes with an SLS printed part because of the powder and curing process used. It is hard for me to say for sure without knowing more specifics, but my initial thought is that a nylon SLS printed part would not be a good end use substitute for a molded PEEK part based on their mechanical properties and surface texture differences in an application like a gear.
Hey, I would like to ask a question, I think the price of controllers of what you are saying ahould be much much less, you know why? Because ABS is one of the hardest material to print in 3D printing, you should've chosen PLA, it's much cheaper and much easier to print. Correct me if I am wrong. Second, I would like to ask that do they have to make different molds, for different parts? Because if yes aren't molds like crazy expensive, I saw a video on RUclips that said a cube mould costs 5k, so a lot of small parts and a controller body should cost more than that, shouldn't it? Really excited for your response, coz I am really interested in this topic and loved your vid!
Hey there. I will do my best to answer but if something is missed just let me know. I chose ABS because that is the typical material used in molding. Yes it is more expensive to print BUT other vendors ( i think the pinned comment has some numbers) will vary pretty wildly on prices so its good to shop around!. Abs was used to try and do an Apples to Apples. SLA doesn't have ABS but ABS like material and SLS doesn't have an ABS, but there is a lot of content talking about FDM replacing molding. PETG would have been a good 3d print option but PLA not as much if you were trying to print and sell something. PLA is easy to print and great for one-off and at home designs but I wouldn't consider it for anything that would say be in someone's hands being gripped/twisted and heated up. PLA tends to warp and has a lower melting temp than PETG or ABS so if you left it on a desk in the sunlight it would be much worse off. So if you are making your own for your own use, PLA or PETG would be good and somewhat easy to print. Molds - The cost of molds isn't as high as you think but they do vary a good bit. The molds i quoted were tool still and 2 cavity. meaning the small parts, buttons, were in a single mold, and the controller bodies were in a single mold. "Family" molds are used when items have a similar volume and can be nested. Putting the buttons and the housings in a single mold generally doesn't work because the volume needed for the housing is much bigger. So you have temp and pressure issues. The molds i quoted were over seas (for us here in the US), and made from tool steel good to 10k units. The price of the mold is fixed but how many parts you make in a run will wildly change the price. If you set up a mold to make 50 parts vs 5000 parts for example. at 5000 parts the price per part is very low because the setup time associated is a lot of the cost. 3d printing doesn't offer much in the way of volume discounts (for fdm). Printing SLA or SLS does have benefits when you can fill up a build area. If one FDM part takes 1 hr, 2 take 2hrs. in SLA, if you can fit 50 buttons in a single build it will take the same amount of time as 1 button (very close) because it cures the entire layer at a time. So in this case the 2 molds were around $9000 i think. If you had an aluminum mold made with no cooling it could be a bit cheaper BUT you would have a much lower quantity you could make before it needed to be refurbished. It is all a trade off and you have to look at how many parts you want to make, the material, where you want to have it produced and so on. You could easily spend hundreds of thousands getting molds made, but if you are producing hundreds of thousands of parts you have a lot to consider. Not just the cost of tooling, but where the parts are and where that is in proximity to other components needed to build your design. Do the plastic parts get delivered 1mile down the road to the electronics parts/assembly line, or do they need to be on a boat for a month and clear customs before you can assembly and sell your design. That is just one example.
Thanks for the Sub! Which process/materials specifically are you talking about? The ABS like materials for SLA are pretty good. They tend to be a little more brittle than a true ABS. For example a screw boss is much more fragile in an ABS like SLA material than in a true molded ABS part. ABS fdm materials are pretty good but you still have the layer line issues so its never going to be structurally the same as a molded part. For at home 3d printing for more structural parts I prefer a PETG material. Prints better than ABS on most desktop machines, even without an enclosure than something like ABS, and is much more mechanically sound than a PLA print. For SLS I like to use Glass filled Nylon for most applications that need a bit more strength. You can even use Carbon filled.
For my uses I'm thinking about one off high quality parts that are durable/tough and accurate (capable of thin walls). I've been looking at sls carbon filled nylon, urethane casting; but the tolerances and properties on the processes are hard to find without testing Ok, I didn't know the sla abs type materials were good for structural parts I'll have to look into them. Always good to have a professional opinion so thank you.
@@qwertyqwerty9587 yeah the thing you want to look for is low shrinkage. www.amazon.com/ELEGOO-ABS-Like-UV-Curing-Standard-Photopolymer/dp/B0B12HX7CT?th=1 Thin walls are going to be tough but for the end use applications I would say an SLA process is going to get you closest but shrink/warpage are your enemy. MJF produces some really nice parts but depends on what you mean by "thin wall" really. Something like a carbon nylon sls can be great but again if you are talking about really thing stuff like 1-2mm you might find the surface finish/resolution isn't quite there. A lot of the bigger companies will send you sample prints/parts so its always worth asking a supplier. I know a lot of them have small samples they can send you to keep. Especially if you tell them you are going to be injection molding. For urethane casting you will generally be making more parts to justify the cost. Not something you would do for a 1-off, but the ABS like material for that should be close to the SLA stuff. Best thing is to find a supplier you want to work with and ask about their materials and samples. They should send you spec sheets and you can see numbers on shrinkage and mech properties.
@@LearnEverythingAboutDesign a lot of good information here, I'll definitely see about getting a couple of sample pieces so I can see how the parts feel in hand along with the specs; I think I lean more towards the function than the finish, so carbon nylon sls is starting to make sense to me. Thanks for the help.
One thing that I dont really see addressed in these discussions is *why* are these services quoting relatively simple FDM parts at such high prices? The material cost for ABS filament can be as low as 12 USD per kilogram just on the consumer end, and energy cost per part wont be especially high for an industrial printer. Is the rest purely labor cost for manual part removal? Having made low run or custom production parts on 3d printers before for exoskeletons and the like I'm simply not understanding where the extra cost is coming from. Even with exorbitantly expensive filament materials like GF40 nylon the cost per 200 gram part was sitting at around 50-60 USD per part after labor and electricity. it doesnt seem to make sense that a mid-quality ABS part would cost 167 dollars even at low volume
"Is the rest purely labor cost for manual part removal?" Labour, plus the cost of hiring the building, fixing/replacing broken printers, the time lost due to failed prints, insurance (3D printers are a fire hazard), 3D prints require a metric F-ton of post-processing, the list goes on. Don't forget that if you want to run a proper farm, you're loking at printers that cost several thousand a piece, and they can probably all be scrapped in a few years, you need to pay for them aswell. Also, 3D printing is a prototyping tool and these companies exist as a service to those who can't / don't want to do the work themselves. If you are working on a commercial product then this is a cheao way of finalizing the design. But when you are going into production then 3D printing is not even remotely an option.
@@vinny142 most 3D prints when designed correctly require little or no post processing and the said 60 USD cost includes maintenance and wastage. If you wanna be pessemistic even, add 10 for building rent and management costs. The 4 printers we had for these parts were 13K each and we made that back by far just by making our own parts for machines we sold. Even if you double the cost I quoted for even PA6-GF parts it doesn't touch most of the absurd quotes I see thrown around for sets of printed ABS parts from these print farm companies. Generally I favor injection molding or machining by far, but I'm just confused how the break even for these quotes isn't significantly higher
Industrial-grade 3d printers are eye-wateringly expensive. Stratasys, 3D Systems, HP, EOS, etc are literally tens and hundreds of thousands of dollars. The overhead a printing service carries is enormous. To some extent "print farms" ride the coat tails of AM service bureaus ssaddled with heavy invelstment on the basis of false equivalency: "FDM is FDM", (which I don't completely disagree with, depending on the specific case). Service bureaus invest in those industrial solutions for uptime, reliability and repeatability; it's a benefit to them, but it influences the cost that gets past on to the consumer.
Because I don't work for these companies it is all speculation on my part, so take that for what it is worth. Reconfiguring quotes to different materials change the prices considerably which leads me to believe that some of it is the difficulty of the process. For example one of the oval buttons from the controller was 3400 for 500 of them in an ABS with 85% infill. Moving to an ASA was $15000!!!!! Going to a PETG with 10-25% infill dropped it to about $2000. So material has a big effect on things and i would guess its based on common materials that area already in machines and ready to go and maybe the complexity or even consistency of the printing process. Other factors I think come in to play. When using an FDM process, these big companies can't stack the build volumes like they can with SLS and SLA. In those other processes they can literally 3d arrange to fill it up and not take a hit on print time. With FDM that isn't really the case. If I order a part that takes 2 hrs, and Joe also orders a part that takes 2hrs. They either put both at the same time and spend 4hrs on that machine, or if they are different material they spend 2hrs, then swap materials, then another 2hrs. There is also a warm up and cool down time. It really isn't a very efficient process for churning out a lot of parts like sls or sla. So from a business perspective that one machine is tied up for a long time which has a cost associated with it. Industrial FDM machines are very expensive. They aren't running a farm of Prusa machines. They invest thousands of dollars for industrial solutions for their quality and repeatability. Now with FDM that isn't saying much if Joe could go and buy a Bambu X1 and print everything he needs at home. Why use a print farm running the same printer you have at home? I think FDM as a service really doesn't make much sense, especially when you consider the print farm machines are ones you can buy yourself, unlike an SLS or MJF machine where you might go to a service provider. Years ago when I looked into some of these industrial machines I remember some of them you couldn't buy, they had lease only options with a large up front training cost and something like $15k/mo to lease. That means these things need to be running 24/7 or the cost are likely higher per print. Since FDM is so common I would imaging it is a lesser used process. On the topic of the quotes and $167 per controller. That is a top ($65) bottom ($59), 4 round buttons ($7ea) and 2 oval buttons ($7 ea). all rounded numbers here. So its not like they were way out of the ballpark. But those numbers didn't drop much even jumping up to printing 1000 round buttons, 500 oval, and 250 of each housing part. Round went from $7.12 to $6.72 for example. The housing top went from $64ish to $61. Where injection molding at 50 units the controller top was about $4.80 (in a nested mold with the bottom each being $4.80). Making 250 of them dropped the cost to $1.50 each. SO, i don't really know why the costs are what they are but those are my guesses/opinions on the matter. As far as outsourcing prints I have always outsourced DMLS, micro sla, sla, and sls. I have never paid to order an FDM print....
In FDM you must calculate bad batches and printer wear. I charge 2€ per hour + material cost x20%. One controller would be around 10-15€ in ABS I think. 5h print time.
Thanks for the comment. Did I say something incorrect about the process in the video? I think the only mention of silicone was for the molds. For example a silicone mold is UV stable, agree there, but the Urethane cast part from the mold generally is not UV stable. If something is incorrect I will make sure to pin the comment and address the time stamp. thanks!
I know right :) this one has been on my list for awhile and just finally got it up today and sure enough a few mins later....... I sort of hit his comment section hard on that video. It really bugs me that he never responds properly to clear comments or ever backs his opinions up. I think it does a big disservice to his image and business. We know that FDM isn't going to replace molding and we know that FDM isn't even the best 3d printing solution...
@LearnEverythingAboutDesign yeah he's getting told what's what pretty clearly in the comments. Ill agree with him printing is a prototype leader, but will never in terms of mass production be able to compete with molds. His twisting of cost analysis to make his point is interesting in terms of how hard the truth can be denied to support ego.
@@LearnEverythingAboutDesign Yea, bothered me too. I've unsubscribed from his channel, the design tips and tricks videos were great, the broken record it's become is too painful to watch.
$20 each still ain’t cheap. $1000 for 50 is pretty wild once you go to 1000 items at $20,000, where the 3d prints would be at over $100,000 at the cheapest. 🇦🇺😎🤜🏼🤛🏼🍀☮️☮️☮️ Not to mention how homogeneous and strong injection moulding is. Along with the material spectrum available. Wild “old” technology really. Haha😂
For sure. In the pinned comment on this video i did post some other numbers from a different 3d print vendor. Fictiv doesn't seem to have good volume discount on the 3d prints but others do. It just takes the "break even" point out a bit more but still Molding comes out way ahead for any real quantities. And note I didn't factor in a discount on the molding prices. Those would drop more if you ordered 5000 or 10000
There has to be cheaper fdm print farms 168$ per controller is too expensive it would cost 1-2 dollars to print yourself. This has to be more of the price for a 1 off 3d print that doesn't scale much with the amount of units. The final cost of the game controller should be around 20$ so none of the quotes are viable at only 50 units. Cheap fdm should work until you can scale up too make injection molding worth it but is 50-100 units really mass production? Theses methods would only work with a specialized product like that flight controller with high margins not just a basic game controller at a low amount of units. I also want to say that as a consumer I would only be ok with fdm if its part of the branding. Like its a small kickstarter project or someone is making parts out of his garage or small print farm. I would never be ok with fdm from a medium or large business.
I agree that $168 is too much! The numbers I used for 50 and 250 total controllers were quotes, not extrapolated for single unit prices. I did extrapolate them out from there, but those initial numbers were firm quotes from Fictiv. Happy to share them if you want to drop me an email. Not all big suppliers even offer FDM as a solution which was a reason i went with Fictiv. Protolabs for example offer FDM through a network of suppliers but not directly. The cost of an industrial FDM machine vs a print farm using consumer machines influences the cost, and a lot of these services that have a "network" of print farms are not using industrial machines and hence their prices are higher. If Fictiv is using an industrial FDM machine that costs thousands of dollars vs a print farm using a $500 machine for example. The layer height, material, color, and lead time all influence it as well. The intent of the video was to show the affordability of molding and how its not wildly out there and scales very well, which i think is still valid even with the high FDM prices with Fictiv, do you agree? Going through a supplier like Hubs.com which uses a network of suppliers they quote FDM much cheaper. There are also other suppliers that quote molding cheaper and more expensive. Kind of why I went to a single source, but maybe I can do an update with other suppliers across the board. For comparison if I went through Hubs(A protolabs affiliate) for FDM you could print one controller for around $100 for all the parts(1 top, 1 bottom, 4 round buttons, 2 oval buttons). Their quantity discounts were more aggressive than Fictiv to a point. Printing 250 total controllers in the same lead time as injection molding put the price around $5000 vs $11000 for molding. That is about $20/controller for FDM at that quantity. I also agree that these quantities don't represent mass manufacturing, but most of the quote systems flake out around 250-500 for 3d printing. Even at 250 units the price per part is cheaper for molding, but you haven't amortized the cost of the mold yet. But even at those deep rates the prices are very close at 500 units factoring in the mold cost. 10k vs 12k. So printing vs molding 500 controllers makes a $4/controller difference even paying for the mold and at 1000 units the molding cost drops to $16/controller and I haven't noticed much of a price drop for FDM past that point. That is usually where the quoting systems go "is this guy serious, please call our support" :) And I do think if you are making a few hundred of something and you find FDM as a viable solution for your product that it makes sense to buy a few machines and print the parts yourself vs even the best FDM quote at the small scale. At 1000 units, molding quickly becomes more beneficial. So even at 10k units we are talking a big difference in price for printing vs molding. But if you never plan to make 10k of the same thing and you want the small scale offering, but a few Bambu P1s machines and run them 24hrs a day and quickly scale. Run the AMS 4 spool system with 4 of the same materials and just have it swap over when it runs out. The machine will run for a long time without much human help. Thanks for the comment and chiming in on the topic!
I believe you about the quotes it just seems those services price to not lose money on a single unit and don't give big discounts with scale. I think you did a great job of showing that injection molding is much more affordable than I expected it to be. Would love to see more videos where you just talk about a topic like this too.
Unequivocally, what you have proven with this video is that is no longer feasible to have parts made in the U.S. At $82,040 for 10K units translates to $8.20 each for game controller housing. When you go through the math that a marketing person would do, you find that the cost of the housing burdens the product with a $28 price tag without the electronics guts( (COG + burdened labor) * 3.5). Also, I am assuming the tooling is aluminum, so the tool will be at or near it's end of life at 10K pieces, so the price will never get cheaper with that approach. Time to learn to speak Chinese.
In the video I didn't say these molds were made in the US. I did say they were tool steel. 738H which are good for 10k parts without refresh. The aluminum molds I have dealt with are good for around 500 parts before a refresh. These molds are "Overseas". $8.20 each for the plastic parts isn't bad for a low volume part. If it were millions of them it should be pennies but the molds are made for higher run machines. There are other vendors and I have had molds made in the US for various projects, Mexico, and China. In the video I did mention that I quoted 50 and 250 parts and extrapolated. There are higher volume discounts with molding but not as much with FDM. The point of the video was to highlight and expel the myth that low volume molding is somehow astronomically more expensive than 3d printing and that just isn't the case. This video had nothing to do with US vs overseas comparison, but cnc machining, and molding are much cheaper in most cases when outsourced sadly.
Full Disclosure. Everything in the video was quoted from Fictiv and accurate, but FDM prices do vary a good bit across various suppliers. As mentioned in the video you should shop around and find a vendor that fits your needs for whatever method is right for you. I do believe it is a bit odd to outsource FDM when the consumer machines are getting so good. You can print a part for a few dollars at home that would cost 5-50x more through a vendor. I don't sell services like this so I don't benefit if you buy a 3d printer or order parts through a vendor. Just my opinion on the topic of course so take it for what that is worth.
I did some more quoting with other vendors and did find some FDM prices that started at around $100 per controller vs the $167 with Fictiv. Depending on the supplier you might also see greater volume discounts with certain types of fabrication. The best I have found still put Injection Molding about equal with FDM in the 250-500 controller range with an advantage to Molding quickly after that point. As we saw the difference between 50 and 250 molded controllers was very little. Check your vendors and quote around, but also talk to the places and question the prices. If something seems abnormally high they should be able to tell you why. I have found Fictiv to be very open about that when I have asked on other projects.
JLCPCB Can print the same controller in SLA for a fraction of the cost you were quoted, that's Laser Resin cured 3D Printer which is much superior when compared to FDM.
@@nicamarvin agreed. The pic in the thumbnail and the car and controller I went over were SLA prints. I am not sure if you watched the whole video but I did show sla, sls, and fdm prints in the video.
Note that I have been providing feedback to fictiv on their FDM pricing relative to the competition to let them know where they are good and where they miss the mark. As mentioned you should shop around.
This is such a fantastic video, Matt! So much objective commentary that contrasts the piles of hyped-up content spinning FDM as a complete replacement for very well-established manufacturing processes that are the foundation of our modern lives and consumer experiences.
Thanks Matt, as a 3D printing service provider I’ve had to learn the skills of urethane casting and soon injection molding. One thing I would like to point out is the hybrid approach 3d printed mold to either urethane cast or injection molding. It’s not getting the backing I would like but I plan furthering the discussion with my clients and designer like yourself. I’m always hesitant to 3D print FDM parts for people and I make them sign a waiver saying these parts cannot be used in a functional application. Mainly because as you pointed out no matter the material there are still limited as an end use part.
Great point! I have used silicone based molds from an SLA print with good results. I think the issue we run into is the cost benefit ratio. In the cases I have personally used it I was able to get away with some pretty complex parts that couldn't be molded traditionally, but the silicone mold off the SLA part let me get by. So getting some nice ABS parts that way is great. The tricky part is how much better is that to say ABS like SLA or even an MJF print.
In your experience is there a tipping point where it makes sense? like making 10-50 of something the print based mold is idea but past that you would go to say Urethane? My biggest issue with Urethane for my applications was UV. At the time the parts needed some funky colors to get the performance (for what reason i don't recall).
If you want to send me an email support@caducator.com with your info I am happy to include it in the desc of this video if you like.
The strength of the part is huge. Bring a molder and having 5 injection molding machined in house, everything you stated is absolutely correct.
Thanks Brian!
I think you made some great points about fdm and version changes being a negative for consumers. That is something that hasn't really crossed my mind. I'm usually happy with just being able to change stuff, but the constant changes of things can actually be quite draining for both me and the consumer. Being solid and confident on a design, and investing in a large run is a good skill to have.
Thanks for sharing!
Hah this showed up side by side with Slant3D's video in my feed. Well played sir.
Crazy timing right. Also love the old Mario pic, brings back some memories(of watching it recently lol)
Matt, I really love this format and presentation. Well done brotha🤘
Thanks Austin! If there are other topics like this feel free to suggest them. My video list is long but I try to get to them all.
Matt, excellent video and it’s fantastic you are teaching what it takes to make a product and show there is no silver bullet… especially FDM! We appreciate you and Motocilino’s work greatly, and I’m glad we could lend our controller to teach folks about manufacturing and design 🎮
Thanks so much Thomas!
I’m grateful for this discussion to be happening. It’s important for customers and engineering students to be aware of the range of fabrication methods at available to them. To push the objectivity further, and this is not to be a troll or compel you to be one, perhaps you can get a quote from “The Company That Shall Not Be Named” and Fictiv for a product that would be feasibly produced from each company. Perhaps TCTSNBN has advantages different from other FDM print farm competitors that make their prices lower. I’d be curious to know and I’d love to have the dialog about this continue.
Thank you for your excellent content.
Thanks :) And so funny story about that. I did try to quote one of the parts through them. Was listed as 1-3 days and 'we will get back to you". I put everything I wanted in the initial contact form for material, quantity, layer height etc. Said orientation was at their discretion. 3 or 4 days later got an email with no quote numbers but asking me if I was sure about my material choice. Have I made it before, how do I want it oriented and yada yada. Compare that to Hubs.com or Fictiv or Xometry or whoever that has instant quoting. If it were me and I was running a print farm I would quote it just like the customer asked, then offer up some alternatives if that is your angle. My guess is they don't want to print ABS, which is fine. I was curious where the pricing would be but honestly waiting 3 days to get no numbers back was problematic.
I do have some plans to do a company to company comparo in the future. I want to look at one process, same material and look at different quantities and turn around times. I will try to include them in that, but I honestly don't know where they fit into the spectrum. In a land of instant quoting, they are slower than quoting injection molding.
@@LearnEverythingAboutDesign Now that is interesting. I expected there to be an automated quoting platform. A strong advocate for a paradigm shift should reduce as many barriers to entry as possible to facilitate its adoption. Hopefully this dialog will lead to those changes. As someone who is familiar with SLA, FDM, and SLS, and an unabashed 3D printing prostyletizer, I am open to innovations that make 3D printing a more widely used fabrication method. I would also like clear-eyed analyses, comparisons, and critiques of fabrication methods so we can make knowledgeable design decisions and also understand where there is room for 3D printing innovation resulting in efficiency and reduced resource consumption.
@@lestersmith7748 Agreed. I wrote a lengthy response to his latest video but it won't take my comment for some reason. Maybe I have been blocked. I really can't stand the attitude and condescension while reading comments that oppose his view and not offering up any evidence or proof to support his position. Everyone can have opinions and that is good, but I have yet to hear an honest discussion with valid talking points about the process from him. For example the recent comparison between FDM and Molding saying that FDM has fewer inputs being power and material, while molding has power, material, and tooling. Glossing over the fact that FDM material has to be melted 2x and there is another machine needed to make the spool from the pellets. And completely misrepresenting the size and capacity of molding machines. Sure paper and pencil is fewer inputs than a computer so surely its a better option right? 3D printing nerd had a recent video at Formnext talking with nexa about 3d printed dissolvable molds for impossible to make geometry. A 3d printing company printing tooling for a molding machine for hard or impossible to mold parts seems like a great direction/improvement. To somehow claim that FDM would accel at printing an impeller blade where an SLA mold printed in 20mins and an injection mold machine which molds the parts in seconds (granted you have to dissolve the mold over night) is some how going to be a better solution. ruclips.net/video/4kIeqTK5iao/видео.htmlsi=ay_pFx2VQCMMvsqd
FDM is the slowest, most inaccurate, least cosmetic, and weakest out of nearly all other processes available.
Hecking smeg you make great content. Definitely gets my mind find clarification on a project I’m trying to picture and how it could be created.
I have a good old frozen sonic 4K and have been using TR250 high temp engineering resin.
It is designed for prototype injection moulding. If you look after the mould, it can last.
I’m wondering where that type of resin printed injection mould comes in on the of scheme things.
If you calculated that the resin injection mould could last for 200 items.
Just brainstorming.
Great channel.
Please RUclips put more of this channel in my “for you” stream.
😂😎🤜🏼🤛🏼🍀🍀🍀☮️☮️☮️
Thanks for the kind words!
I believe the 3d printing nerd channel had a video with Nexa showing an mSLA printed mold and a desktop sized injection molding machine showing not only the power of that process but using resin based molds for impossible or extremely difficult to mold parts in a traditional sense. I think that is where the real value of the process is right now. Aiding the molding process rather than replacing it.
Everything has it's place. For me, FDM is an amazing scratch paper where you can quickly develop ideas and get a better sense of all the dynamics of the proceess, get feedback and work on locking down the design. After that, mock it up with SLA, then lastly CNC.
100%
I will say I make a lot of "little" things around the house with FDM. Custom drawer organizer trays for example. Great use of it.
I get what you're saying with the versatility in terms of design changes with FDM being a negative when it comes to sending out a design to be manufactured by a company but when you consider that FDM printers are pretty common among people designing and building at home it becomes a different story. I'd think the FDM design test phase is done in-house (or at home) and then beyond that, another method is used
For sure. FDM is great for concepts and testing. One big thing to consider is you really have to design for FDM. You aren't going to design for molding and then print it with an FDM. You could do that with SLS or SLA though.
I love having an FDM at home, but I can't imagine selling and FDM product. Maybe the future will change my mind :)
Hi Matt! When do you plan on rolling out the next set of courses? Cant wait to start.
Very soon! The intro to design for 3d printing has been edited, i just need to get it up on the site. For good or bad i just have way too much work on my plate at the moment.
Hi,
sorry for the translation from French.
Besides my work, I have a small company where I make parts in very small quantities for R&D in the food industry in FDM. (PETG)
It's super cool, and my customers are really happy.
but now some of these companies are asking me for spare parts normally molded in PEEK for machines in small quantities (2-6 units) (gears, outlet end etc.).
I could reproduce them in CAD quickly and deliver them in record time with my SLS but I have a lot of questions about the resistance.
In your opinion, would the resistance on a gear printed in SLS Nylon be as resistant as molded PEEK?
If you have any information on this subject (comparative studies, durability test) I would be very grateful.
Thank you for this video which was fascinating.
Sorry I don't have any specific info/papers/research on your specific question. Generally what I would say is to ask the manufacturer of your resin to provide technical information on that material. In general if we look at nylon vs PEEK there are some potential issues related to temperature. Nylon will get soft at a much lower temperature. So if you are looking at a gear, the friction from use could be enough to generate heat that would cause deformation. The root area of the tooth of a gear is generally small and if the surface was hot enough in a localized area that would be enough to fail.
The surface finish on an SLS part (unless we are talking very large gears) is likely going to be a big performance factor vs a smooth molded part as well.
Nylon in a cast or extruded application has a pretty low coefficient of friction, but that changes with an SLS printed part because of the powder and curing process used.
It is hard for me to say for sure without knowing more specifics, but my initial thought is that a nylon SLS printed part would not be a good end use substitute for a molded PEEK part based on their mechanical properties and surface texture differences in an application like a gear.
Hey,
I would like to ask a question, I think the price of controllers of what you are saying ahould be much much less, you know why? Because ABS is one of the hardest material to print in 3D printing, you should've chosen PLA, it's much cheaper and much easier to print. Correct me if I am wrong.
Second, I would like to ask that do they have to make different molds, for different parts? Because if yes aren't molds like crazy expensive, I saw a video on RUclips that said a cube mould costs 5k, so a lot of small parts and a controller body should cost more than that, shouldn't it?
Really excited for your response, coz I am really interested in this topic and loved your vid!
Hey there. I will do my best to answer but if something is missed just let me know.
I chose ABS because that is the typical material used in molding. Yes it is more expensive to print BUT other vendors ( i think the pinned comment has some numbers) will vary pretty wildly on prices so its good to shop around!. Abs was used to try and do an Apples to Apples. SLA doesn't have ABS but ABS like material and SLS doesn't have an ABS, but there is a lot of content talking about FDM replacing molding. PETG would have been a good 3d print option but PLA not as much if you were trying to print and sell something. PLA is easy to print and great for one-off and at home designs but I wouldn't consider it for anything that would say be in someone's hands being gripped/twisted and heated up. PLA tends to warp and has a lower melting temp than PETG or ABS so if you left it on a desk in the sunlight it would be much worse off.
So if you are making your own for your own use, PLA or PETG would be good and somewhat easy to print.
Molds - The cost of molds isn't as high as you think but they do vary a good bit. The molds i quoted were tool still and 2 cavity. meaning the small parts, buttons, were in a single mold, and the controller bodies were in a single mold. "Family" molds are used when items have a similar volume and can be nested. Putting the buttons and the housings in a single mold generally doesn't work because the volume needed for the housing is much bigger. So you have temp and pressure issues. The molds i quoted were over seas (for us here in the US), and made from tool steel good to 10k units. The price of the mold is fixed but how many parts you make in a run will wildly change the price. If you set up a mold to make 50 parts vs 5000 parts for example. at 5000 parts the price per part is very low because the setup time associated is a lot of the cost. 3d printing doesn't offer much in the way of volume discounts (for fdm). Printing SLA or SLS does have benefits when you can fill up a build area. If one FDM part takes 1 hr, 2 take 2hrs. in SLA, if you can fit 50 buttons in a single build it will take the same amount of time as 1 button (very close) because it cures the entire layer at a time.
So in this case the 2 molds were around $9000 i think. If you had an aluminum mold made with no cooling it could be a bit cheaper BUT you would have a much lower quantity you could make before it needed to be refurbished. It is all a trade off and you have to look at how many parts you want to make, the material, where you want to have it produced and so on. You could easily spend hundreds of thousands getting molds made, but if you are producing hundreds of thousands of parts you have a lot to consider. Not just the cost of tooling, but where the parts are and where that is in proximity to other components needed to build your design. Do the plastic parts get delivered 1mile down the road to the electronics parts/assembly line, or do they need to be on a boat for a month and clear customs before you can assembly and sell your design. That is just one example.
Whats your opinion on the strength of the materials?
Definitely got a new subscriber 😂
Thanks for the Sub!
Which process/materials specifically are you talking about? The ABS like materials for SLA are pretty good. They tend to be a little more brittle than a true ABS. For example a screw boss is much more fragile in an ABS like SLA material than in a true molded ABS part. ABS fdm materials are pretty good but you still have the layer line issues so its never going to be structurally the same as a molded part.
For at home 3d printing for more structural parts I prefer a PETG material. Prints better than ABS on most desktop machines, even without an enclosure than something like ABS, and is much more mechanically sound than a PLA print.
For SLS I like to use Glass filled Nylon for most applications that need a bit more strength. You can even use Carbon filled.
For my uses I'm thinking about one off high quality parts that are durable/tough and accurate (capable of thin walls).
I've been looking at sls carbon filled nylon, urethane casting; but the tolerances and properties on the processes are hard to find without testing
Ok, I didn't know the sla abs type materials were good for structural parts I'll have to look into them.
Always good to have a professional opinion so thank you.
@@qwertyqwerty9587 yeah the thing you want to look for is low shrinkage. www.amazon.com/ELEGOO-ABS-Like-UV-Curing-Standard-Photopolymer/dp/B0B12HX7CT?th=1 Thin walls are going to be tough but for the end use applications I would say an SLA process is going to get you closest but shrink/warpage are your enemy. MJF produces some really nice parts but depends on what you mean by "thin wall" really.
Something like a carbon nylon sls can be great but again if you are talking about really thing stuff like 1-2mm you might find the surface finish/resolution isn't quite there.
A lot of the bigger companies will send you sample prints/parts so its always worth asking a supplier. I know a lot of them have small samples they can send you to keep. Especially if you tell them you are going to be injection molding.
For urethane casting you will generally be making more parts to justify the cost. Not something you would do for a 1-off, but the ABS like material for that should be close to the SLA stuff. Best thing is to find a supplier you want to work with and ask about their materials and samples. They should send you spec sheets and you can see numbers on shrinkage and mech properties.
@@LearnEverythingAboutDesign
a lot of good information here, I'll definitely see about getting a couple of sample pieces so I can see how the parts feel in hand along with the specs;
I think I lean more towards the function than the finish, so carbon nylon sls is starting to make sense to me.
Thanks for the help.
One thing that I dont really see addressed in these discussions is *why* are these services quoting relatively simple FDM parts at such high prices? The material cost for ABS filament can be as low as 12 USD per kilogram just on the consumer end, and energy cost per part wont be especially high for an industrial printer. Is the rest purely labor cost for manual part removal? Having made low run or custom production parts on 3d printers before for exoskeletons and the like I'm simply not understanding where the extra cost is coming from. Even with exorbitantly expensive filament materials like GF40 nylon the cost per 200 gram part was sitting at around 50-60 USD per part after labor and electricity. it doesnt seem to make sense that a mid-quality ABS part would cost 167 dollars even at low volume
"Is the rest purely labor cost for manual part removal?"
Labour, plus the cost of hiring the building, fixing/replacing broken printers, the time lost due to failed prints, insurance (3D printers are a fire hazard), 3D prints require a metric F-ton of post-processing, the list goes on.
Don't forget that if you want to run a proper farm, you're loking at printers that cost several thousand a piece, and they can probably all be scrapped in a few years, you need to pay for them aswell.
Also, 3D printing is a prototyping tool and these companies exist as a service to those who can't / don't want to do the work themselves. If you are working on a commercial product then this is a cheao way of finalizing the design. But when you are going into production then 3D printing is not even remotely an option.
@@vinny142 most 3D prints when designed correctly require little or no post processing and the said 60 USD cost includes maintenance and wastage. If you wanna be pessemistic even, add 10 for building rent and management costs. The 4 printers we had for these parts were 13K each and we made that back by far just by making our own parts for machines we sold. Even if you double the cost I quoted for even PA6-GF parts it doesn't touch most of the absurd quotes I see thrown around for sets of printed ABS parts from these print farm companies. Generally I favor injection molding or machining by far, but I'm just confused how the break even for these quotes isn't significantly higher
Industrial-grade 3d printers are eye-wateringly expensive. Stratasys, 3D Systems, HP, EOS, etc are literally tens and hundreds of thousands of dollars. The overhead a printing service carries is enormous. To some extent "print farms" ride the coat tails of AM service bureaus ssaddled with heavy invelstment on the basis of false equivalency: "FDM is FDM", (which I don't completely disagree with, depending on the specific case). Service bureaus invest in those industrial solutions for uptime, reliability and repeatability; it's a benefit to them, but it influences the cost that gets past on to the consumer.
Because I don't work for these companies it is all speculation on my part, so take that for what it is worth. Reconfiguring quotes to different materials change the prices considerably which leads me to believe that some of it is the difficulty of the process. For example one of the oval buttons from the controller was 3400 for 500 of them in an ABS with 85% infill. Moving to an ASA was $15000!!!!! Going to a PETG with 10-25% infill dropped it to about $2000. So material has a big effect on things and i would guess its based on common materials that area already in machines and ready to go and maybe the complexity or even consistency of the printing process.
Other factors I think come in to play.
When using an FDM process, these big companies can't stack the build volumes like they can with SLS and SLA. In those other processes they can literally 3d arrange to fill it up and not take a hit on print time. With FDM that isn't really the case. If I order a part that takes 2 hrs, and Joe also orders a part that takes 2hrs. They either put both at the same time and spend 4hrs on that machine, or if they are different material they spend 2hrs, then swap materials, then another 2hrs. There is also a warm up and cool down time. It really isn't a very efficient process for churning out a lot of parts like sls or sla. So from a business perspective that one machine is tied up for a long time which has a cost associated with it.
Industrial FDM machines are very expensive. They aren't running a farm of Prusa machines. They invest thousands of dollars for industrial solutions for their quality and repeatability. Now with FDM that isn't saying much if Joe could go and buy a Bambu X1 and print everything he needs at home. Why use a print farm running the same printer you have at home?
I think FDM as a service really doesn't make much sense, especially when you consider the print farm machines are ones you can buy yourself, unlike an SLS or MJF machine where you might go to a service provider. Years ago when I looked into some of these industrial machines I remember some of them you couldn't buy, they had lease only options with a large up front training cost and something like $15k/mo to lease. That means these things need to be running 24/7 or the cost are likely higher per print. Since FDM is so common I would imaging it is a lesser used process.
On the topic of the quotes and $167 per controller. That is a top ($65) bottom ($59), 4 round buttons ($7ea) and 2 oval buttons ($7 ea). all rounded numbers here. So its not like they were way out of the ballpark. But those numbers didn't drop much even jumping up to printing 1000 round buttons, 500 oval, and 250 of each housing part. Round went from $7.12 to $6.72 for example. The housing top went from $64ish to $61. Where injection molding at 50 units the controller top was about $4.80 (in a nested mold with the bottom each being $4.80). Making 250 of them dropped the cost to $1.50 each.
SO, i don't really know why the costs are what they are but those are my guesses/opinions on the matter. As far as outsourcing prints I have always outsourced DMLS, micro sla, sla, and sls. I have never paid to order an FDM print....
In FDM you must calculate bad batches and printer wear. I charge 2€ per hour + material cost x20%. One controller would be around 10-15€ in ABS I think. 5h print time.
silicone is not exposed to UV rays, this may be useful to you in the future
Thanks for the comment. Did I say something incorrect about the process in the video? I think the only mention of silicone was for the molds. For example a silicone mold is UV stable, agree there, but the Urethane cast part from the mold generally is not UV stable. If something is incorrect I will make sure to pin the comment and address the time stamp.
thanks!
Weird that you and slant3d put out a video on kinda a similar thing at the same time.
I know right :) this one has been on my list for awhile and just finally got it up today and sure enough a few mins later....... I sort of hit his comment section hard on that video. It really bugs me that he never responds properly to clear comments or ever backs his opinions up. I think it does a big disservice to his image and business. We know that FDM isn't going to replace molding and we know that FDM isn't even the best 3d printing solution...
@LearnEverythingAboutDesign yeah he's getting told what's what pretty clearly in the comments. Ill agree with him printing is a prototype leader, but will never in terms of mass production be able to compete with molds. His twisting of cost analysis to make his point is interesting in terms of how hard the truth can be denied to support ego.
@@LearnEverythingAboutDesign Yea, bothered me too. I've unsubscribed from his channel, the design tips and tricks videos were great, the broken record it's become is too painful to watch.
$20 each still ain’t cheap.
$1000 for 50 is pretty wild once you go to 1000 items at $20,000, where the 3d prints would be at over $100,000 at the cheapest.
🇦🇺😎🤜🏼🤛🏼🍀☮️☮️☮️
Not to mention how homogeneous and strong injection moulding is. Along with the material spectrum available.
Wild “old” technology really. Haha😂
For sure. In the pinned comment on this video i did post some other numbers from a different 3d print vendor. Fictiv doesn't seem to have good volume discount on the 3d prints but others do. It just takes the "break even" point out a bit more but still Molding comes out way ahead for any real quantities. And note I didn't factor in a discount on the molding prices. Those would drop more if you ordered 5000 or 10000
There has to be cheaper fdm print farms 168$ per controller is too expensive it would cost 1-2 dollars to print yourself. This has to be more of the price for a 1 off 3d print that doesn't scale much with the amount of units. The final cost of the game controller should be around 20$ so none of the quotes are viable at only 50 units. Cheap fdm should work until you can scale up too make injection molding worth it but is 50-100 units really mass production? Theses methods would only work with a specialized product like that flight controller with high margins not just a basic game controller at a low amount of units.
I also want to say that as a consumer I would only be ok with fdm if its part of the branding. Like its a small kickstarter project or someone is making parts out of his garage or small print farm. I would never be ok with fdm from a medium or large business.
I agree that $168 is too much! The numbers I used for 50 and 250 total controllers were quotes, not extrapolated for single unit prices. I did extrapolate them out from there, but those initial numbers were firm quotes from Fictiv. Happy to share them if you want to drop me an email. Not all big suppliers even offer FDM as a solution which was a reason i went with Fictiv. Protolabs for example offer FDM through a network of suppliers but not directly. The cost of an industrial FDM machine vs a print farm using consumer machines influences the cost, and a lot of these services that have a "network" of print farms are not using industrial machines and hence their prices are higher. If Fictiv is using an industrial FDM machine that costs thousands of dollars vs a print farm using a $500 machine for example. The layer height, material, color, and lead time all influence it as well. The intent of the video was to show the affordability of molding and how its not wildly out there and scales very well, which i think is still valid even with the high FDM prices with Fictiv, do you agree? Going through a supplier like Hubs.com which uses a network of suppliers they quote FDM much cheaper. There are also other suppliers that quote molding cheaper and more expensive. Kind of why I went to a single source, but maybe I can do an update with other suppliers across the board.
For comparison if I went through Hubs(A protolabs affiliate) for FDM you could print one controller for around $100 for all the parts(1 top, 1 bottom, 4 round buttons, 2 oval buttons). Their quantity discounts were more aggressive than Fictiv to a point. Printing 250 total controllers in the same lead time as injection molding put the price around $5000 vs $11000 for molding. That is about $20/controller for FDM at that quantity. I also agree that these quantities don't represent mass manufacturing, but most of the quote systems flake out around 250-500 for 3d printing. Even at 250 units the price per part is cheaper for molding, but you haven't amortized the cost of the mold yet. But even at those deep rates the prices are very close at 500 units factoring in the mold cost. 10k vs 12k. So printing vs molding 500 controllers makes a $4/controller difference even paying for the mold and at 1000 units the molding cost drops to $16/controller and I haven't noticed much of a price drop for FDM past that point. That is usually where the quoting systems go "is this guy serious, please call our support" :)
And I do think if you are making a few hundred of something and you find FDM as a viable solution for your product that it makes sense to buy a few machines and print the parts yourself vs even the best FDM quote at the small scale. At 1000 units, molding quickly becomes more beneficial. So even at 10k units we are talking a big difference in price for printing vs molding. But if you never plan to make 10k of the same thing and you want the small scale offering, but a few Bambu P1s machines and run them 24hrs a day and quickly scale. Run the AMS 4 spool system with 4 of the same materials and just have it swap over when it runs out. The machine will run for a long time without much human help.
Thanks for the comment and chiming in on the topic!
I believe you about the quotes it just seems those services price to not lose money on a single unit and don't give big discounts with scale. I think you did a great job of showing that injection molding is much more affordable than I expected it to be. Would love to see more videos where you just talk about a topic like this too.
Unequivocally, what you have proven with this video is that is no longer feasible to have parts made in the U.S. At $82,040 for 10K units translates to $8.20 each for game controller housing. When you go through the math that a marketing person would do, you find that the cost of the housing burdens the product with a $28 price tag without the electronics guts( (COG + burdened labor) * 3.5). Also, I am assuming the tooling is aluminum, so the tool will be at or near it's end of life at 10K pieces, so the price will never get cheaper with that approach. Time to learn to speak Chinese.
In the video I didn't say these molds were made in the US. I did say they were tool steel. 738H which are good for 10k parts without refresh. The aluminum molds I have dealt with are good for around 500 parts before a refresh. These molds are "Overseas". $8.20 each for the plastic parts isn't bad for a low volume part. If it were millions of them it should be pennies but the molds are made for higher run machines.
There are other vendors and I have had molds made in the US for various projects, Mexico, and China.
In the video I did mention that I quoted 50 and 250 parts and extrapolated. There are higher volume discounts with molding but not as much with FDM. The point of the video was to highlight and expel the myth that low volume molding is somehow astronomically more expensive than 3d printing and that just isn't the case. This video had nothing to do with US vs overseas comparison, but cnc machining, and molding are much cheaper in most cases when outsourced sadly.