5 ms·

CTO of Fabric8 here - happy to answer any questions you may have about the technology or company. As you can tell we've got a process that's quite different tha

by dpain 2y ago

CTO of Fabric8 here - happy to answer any questions you may have about the technology or company. As you can tell we've got a process that's quite different than other metal AM techniques with some very unique benefits that we're excited to share!

Joel_Mckay 2y ago
How did your team handle the clogging problem with these designs? https://www.youtube.com/watch?v=B-UbDk7LrvU https://www.youtube.com/watch?v=B-UbDk7LrvU Best regards, =3
- justinclift 2y ago
  That's a cool video. They even show creating multi-metal stuff with it, which seems like it'd be pretty lined up for creating meta-materials when using a small enough tip resolution. :) https://youtu.be/B-UbDk7LrvU?si=1GeqL7aoLbG6PBCH&t=462 https://youtu.be/B-UbDk7LrvU?si=1GeqL7aoLbG6PBCH&t=462
  - Joel_Mckay 2y ago
    Someone probably has something better already sitting on a shelf somewhere... =3
adrian_b 2y ago
Which are exactly the metals that you can currently print by ECAM? The images suggest pure copper. Are there any alloys that can be printed at this time?
Nathanael_M 2y ago
Where does this fall on the hazardous/toxicity scale? What kind of off-gassing/risks are there, especially compared to existing high resolution powder based systems? Disclaimer: I don’t know much about this field, this may be a dumb question.
- dpain 2y ago
  Much safer than existing powder-based systems. The feedstock is effectively water based so none of the flammability risks of metal powder. No special gases required, no high powered lasers or thermal processing sytems. That being said it still is an industrial process and requires responsible handling of the feedstock and equipment to ensure safety to personnel and the environment.
  - Nathanael_M 2y ago
    Thanks for the response!
- tastyfreeze 2y ago
  Metallic ionic solutions are usually strong acids. They are made by dissolving metal in sulfuric or nitric acids. Looks like they are using copper sulfate as the electrolyte. After the copper is deposited you are left with sulfuric acid.
cjbgkagh 2y ago
It’s really cool and I wish I had one of these machines. On the assumption that I couldn’t afford such a machine myself for DIY, it would be really cool to have a send-cut-send like service. It seems like the relatively gentle process and I assume predictable results with minimal post processing would lend itself to an high degree of automation. Perhaps a scale out manufacturing set up where the speed of an individual machine is less important than the throughput of having many machines. There are a lot of little random things that I would design if I knew there was a capability to have it made at non-aerospace prices. Initially heatsinks, manifolds, heat exchangers, but possibly many random things depending on the pricing.
- Tossrock 2y ago
  You can get on-demand metal printing from Shapeways. Somewhat expensive compared to sendcutsend, but still far below aerospace.
  - cjbgkagh 2y ago
    Sure, there is a huge gamut of prices from aerospace to just above material costs. On the expensive side I would use it judiciously, on the economical side I would use it for just about everything and so would many others. I do a bit of DIY and a large amount of the design work is to work around limitations of cheaper materials and cheaper processes. Most of what I make is onesie-twosie where the design costs dominate so I can afford to spend more on a more expensive materials and processes to cut back on the design. But it would be extra amazing if there was an economical high quality process that is suitable for prototyping and would also be amenable to cost competitive mass manufacturing. That would be absolutely world changing. If I could just list designs somewhere and customers have them printed on demand, either via Amazon or some drop-ship, I could recoup a lot of the design costs so I could spend more time designing things - maybe even make a career out of it. Instead of needing a whole factory to support the low cost manufacturing of a single part it could just be a design document sitting on a server somewhere. I understand that it is not in Fabric8s interest to do this though, expensive machines are high margin and exclusivity is required to maintain a bit of a monopoly for their customers so that their customers can justify purchasing the expensive machines in the first place. If Fabric8 were to start out with expensive machines and then subsequently release cheaper machines they would burn their previous customers. That would make this one of the many cool things that I can't use until the patents expire which would be a damn shame and would also make the technology completely uninteresting to me as I would have to focus on alternatives in the interim. It took a long time to go from Rep-Rap to Bambu Labs for FDM but with increased interest in the space that process is speeding up. Hopefully Micronics will manage a consumer Nylon SLS printer. A ton of small scale manufacturing technology is being made cheaply enough for home use and services like PCBWay and Send-Cut-Send have really democratized high quality manufacturing. I can make many things production quality with near zero overheads and I can buy things from others who have done the same at similarly low costs. Something like send-cut-send for ECAM seems inevitable and it would be up to Fabric8 to decide if they want to cannibalize themselves instead of having someone else undercut them. I don't know how defendable their patents are, there does seem to be a fair amount of prior art. There is probably a bunch of trade secrets though. For them one of the downsides of having such a promising solution is that there is an even greater incentive for competitors to enter the market. I would posit that long term there is more money to be made with a scale-out low margin mass market solution where they could better leverage trade secrets gleaned from process experience than than selling individual high margin machines. If they kept overheads low enough it would never be in anyone else's interest to enter the market and they could leverage their trade secrets for above average returns ad infinitum.
philipswood 2y ago
This is so cool! What kind of currents do you need? I've considered the idea before and it seemed to me that since you need a fixed number of electrons for every metal ion you deposit that the currents end up being huge. Also deposit speeds tend to be slow. How fast can your process layer metal (say in grams per hour)?
- dpain 2y ago
  Yes, currents can be high as they directly correlate with build rate. But the voltages are typically low, making it a lower power process than one would think. 100-1000x faster than a typical electroplating process
cpitman 2y ago
Your website mentions printing on pcbs, do you expect this to be a competitive option for pcb prototyping or manufacture?
- dpain 2y ago
  certainly an area we're keeping an eye on.
- michaelt 2y ago
  PCB prototype manufacturing is already exceptionally cheap compared to pretty much all other forms of outsourced manufacturing. You can get 10 PCBs for $20 including worldwide shipping. Pricing for CNC or laser cutting or 3D printing just doesn't go that low.
jimbooonooo 2y ago
Tons of questions, this is incredibly cool! Can you discuss: 1.) How you compensate for anode consumption/geometry changes over the lifetime of the anode. For instance, does the center get worn away or do you try to uniformly use each "pixel". 2.) More details on anode "pixel" geometry and minimum feature size. 3.) Can you talk about the development process? Did you have in situ measurement, or post build analysis of the part and anode. 4.) Typical surface finish?
ipsum2 2y ago
Can we see a benchy?
InDubioProRubio 2y ago
What sort of crystallization behaviour can one expect from this process? Can one controll the crystallization formation?
FrostKiwi 2y ago
First time hearing about the technique, sounds exciting. I see a close comparison in features to SLM [1], which is already established as a core 3D Metal printing technique for a long time. SLM has precision down to the size of a mechanical pencil's lead. In what way is ECAM better? Is it more precision + no need to handle powder or shield gas + no need for laser source and containment, minus ECAM being slower. Am I missing some crucial feature? [1] https://en.wikipedia.org/wiki/Selective_laser_melting https://en.wikipedia.org/wiki/Selective_laser_melting
phkahler 2y ago
Has there been any work on photoelectric plating? You could then use DPL to project a pattern for an entire layer at once. I'm not sure how one could ensure uniform thickness. Just wondering if the basics of such an approach have been developed or if its even possible.
darksaints 2y ago
A couple of questions: 1) can you print on oxide, nitride, or carbide ceramics? 2) can you print on shaped materials? Like for example, printing a pcb directly onto the inner side of a rigid shell for an aerofoil/wing? 3) what sort of material restrictions come with electrochemical printing? Can you do aluminum? Titanium? 4) what is the minimum feature size? Could this, for example, print a controlled porous/wick structure with pore sizes around 50um? 5) for those of us without the capital to buy a printer, will you also offer printing services?
- dpain 2y ago
  1 - Yes, we are currently doing this 2 - Right now our focus is on flat substrates, curved surfaces are a little trickier 3 - Any pure metal or alloy system that can be electroplated would be compatible with our approach. Aluminum and Titanium are difficult but not impossible. We're focused on copper as our first commercial material but have other material systems in development. 4 - Our current pixel size is 33 microns, ~50micron negative features should be doable. Controlled wick structures are actually a really good application of the technology 5 - Our primary business model actually is to offer print services to our customers. Feel free to reach out to me via email in profile if there's an application you'd like to explore

by dpain 2y ago

Joel_Mckay 2y ago

How did your team handle the clogging problem with these designs? https://www.youtube.com/watch?v=B-UbDk7LrvU https://www.youtube.com/watch?v=B-UbDk7LrvU Best regards, =3

justinclift 2y ago
That's a cool video. They even show creating multi-metal stuff with it, which seems like it'd be pretty lined up for creating meta-materials when using a small enough tip resolution. :) https://youtu.be/B-UbDk7LrvU?si=1GeqL7aoLbG6PBCH&t=462 https://youtu.be/B-UbDk7LrvU?si=1GeqL7aoLbG6PBCH&t=462
- Joel_Mckay 2y ago
  Someone probably has something better already sitting on a shelf somewhere... =3

adrian_b 2y ago

Which are exactly the metals that you can currently print by ECAM? The images suggest pure copper. Are there any alloys that can be printed at this time?

Nathanael_M 2y ago

Where does this fall on the hazardous/toxicity scale? What kind of off-gassing/risks are there, especially compared to existing high resolution powder based systems? Disclaimer: I don’t know much about this field, this may be a dumb question.

dpain 2y ago
Much safer than existing powder-based systems. The feedstock is effectively water based so none of the flammability risks of metal powder. No special gases required, no high powered lasers or thermal processing sytems. That being said it still is an industrial process and requires responsible handling of the feedstock and equipment to ensure safety to personnel and the environment.
- Nathanael_M 2y ago
  Thanks for the response!
tastyfreeze 2y ago
Metallic ionic solutions are usually strong acids. They are made by dissolving metal in sulfuric or nitric acids. Looks like they are using copper sulfate as the electrolyte. After the copper is deposited you are left with sulfuric acid.

cjbgkagh 2y ago

It’s really cool and I wish I had one of these machines. On the assumption that I couldn’t afford such a machine myself for DIY, it would be really cool to have a send-cut-send like service. It seems like the relatively gentle process and I assume predictable results with minimal post processing would lend itself to an high degree of automation. Perhaps a scale out manufacturing set up where the speed of an individual machine is less important than the throughput of having many machines. There are a lot of little random things that I would design if I knew there was a capability to have it made at non-aerospace prices. Initially heatsinks, manifolds, heat exchangers, but possibly many random things depending on the pricing.

Tossrock 2y ago
You can get on-demand metal printing from Shapeways. Somewhat expensive compared to sendcutsend, but still far below aerospace.
- cjbgkagh 2y ago
  Sure, there is a huge gamut of prices from aerospace to just above material costs. On the expensive side I would use it judiciously, on the economical side I would use it for just about everything and so would many others. I do a bit of DIY and a large amount of the design work is to work around limitations of cheaper materials and cheaper processes. Most of what I make is onesie-twosie where the design costs dominate so I can afford to spend more on a more expensive materials and processes to cut back on the design. But it would be extra amazing if there was an economical high quality process that is suitable for prototyping and would also be amenable to cost competitive mass manufacturing. That would be absolutely world changing. If I could just list designs somewhere and customers have them printed on demand, either via Amazon or some drop-ship, I could recoup a lot of the design costs so I could spend more time designing things - maybe even make a career out of it. Instead of needing a whole factory to support the low cost manufacturing of a single part it could just be a design document sitting on a server somewhere. I understand that it is not in Fabric8s interest to do this though, expensive machines are high margin and exclusivity is required to maintain a bit of a monopoly for their customers so that their customers can justify purchasing the expensive machines in the first place. If Fabric8 were to start out with expensive machines and then subsequently release cheaper machines they would burn their previous customers. That would make this one of the many cool things that I can't use until the patents expire which would be a damn shame and would also make the technology completely uninteresting to me as I would have to focus on alternatives in the interim. It took a long time to go from Rep-Rap to Bambu Labs for FDM but with increased interest in the space that process is speeding up. Hopefully Micronics will manage a consumer Nylon SLS printer. A ton of small scale manufacturing technology is being made cheaply enough for home use and services like PCBWay and Send-Cut-Send have really democratized high quality manufacturing. I can make many things production quality with near zero overheads and I can buy things from others who have done the same at similarly low costs. Something like send-cut-send for ECAM seems inevitable and it would be up to Fabric8 to decide if they want to cannibalize themselves instead of having someone else undercut them. I don't know how defendable their patents are, there does seem to be a fair amount of prior art. There is probably a bunch of trade secrets though. For them one of the downsides of having such a promising solution is that there is an even greater incentive for competitors to enter the market. I would posit that long term there is more money to be made with a scale-out low margin mass market solution where they could better leverage trade secrets gleaned from process experience than than selling individual high margin machines. If they kept overheads low enough it would never be in anyone else's interest to enter the market and they could leverage their trade secrets for above average returns ad infinitum.

philipswood 2y ago

This is so cool! What kind of currents do you need? I've considered the idea before and it seemed to me that since you need a fixed number of electrons for every metal ion you deposit that the currents end up being huge. Also deposit speeds tend to be slow. How fast can your process layer metal (say in grams per hour)?

dpain 2y ago
Yes, currents can be high as they directly correlate with build rate. But the voltages are typically low, making it a lower power process than one would think. 100-1000x faster than a typical electroplating process

cpitman 2y ago

Your website mentions printing on pcbs, do you expect this to be a competitive option for pcb prototyping or manufacture?

dpain 2y ago
certainly an area we're keeping an eye on.
michaelt 2y ago
PCB prototype manufacturing is already exceptionally cheap compared to pretty much all other forms of outsourced manufacturing. You can get 10 PCBs for $20 including worldwide shipping. Pricing for CNC or laser cutting or 3D printing just doesn't go that low.

jimbooonooo 2y ago

Tons of questions, this is incredibly cool! Can you discuss: 1.) How you compensate for anode consumption/geometry changes over the lifetime of the anode. For instance, does the center get worn away or do you try to uniformly use each "pixel". 2.) More details on anode "pixel" geometry and minimum feature size. 3.) Can you talk about the development process? Did you have in situ measurement, or post build analysis of the part and anode. 4.) Typical surface finish?

ipsum2 2y ago

Can we see a benchy?

InDubioProRubio 2y ago

What sort of crystallization behaviour can one expect from this process? Can one controll the crystallization formation?

FrostKiwi 2y ago

First time hearing about the technique, sounds exciting. I see a close comparison in features to SLM [1], which is already established as a core 3D Metal printing technique for a long time. SLM has precision down to the size of a mechanical pencil's lead. In what way is ECAM better? Is it more precision + no need to handle powder or shield gas + no need for laser source and containment, minus ECAM being slower. Am I missing some crucial feature? [1] https://en.wikipedia.org/wiki/Selective_laser_melting https://en.wikipedia.org/wiki/Selective_laser_melting

phkahler 2y ago

Has there been any work on photoelectric plating? You could then use DPL to project a pattern for an entire layer at once. I'm not sure how one could ensure uniform thickness. Just wondering if the basics of such an approach have been developed or if its even possible.

darksaints 2y ago

A couple of questions: 1) can you print on oxide, nitride, or carbide ceramics? 2) can you print on shaped materials? Like for example, printing a pcb directly onto the inner side of a rigid shell for an aerofoil/wing? 3) what sort of material restrictions come with electrochemical printing? Can you do aluminum? Titanium? 4) what is the minimum feature size? Could this, for example, print a controlled porous/wick structure with pore sizes around 50um? 5) for those of us without the capital to buy a printer, will you also offer printing services?

dpain 2y ago
1 - Yes, we are currently doing this 2 - Right now our focus is on flat substrates, curved surfaces are a little trickier 3 - Any pure metal or alloy system that can be electroplated would be compatible with our approach. Aluminum and Titanium are difficult but not impossible. We're focused on copper as our first commercial material but have other material systems in development. 4 - Our current pixel size is 33 microns, ~50micron negative features should be doable. Controlled wick structures are actually a really good application of the technology 5 - Our primary business model actually is to offer print services to our customers. Feel free to reach out to me via email in profile if there's an application you'd like to explore