I was shocked when my scientific director allowed me to print my own project on this thing. There are ~10k of them all over the world and only 2 in Russia, so it was incredible possibility for me
Surely at 160 or more layers you'd be getting into the territory where you'd want to embed components IN the PCB itself? Which... Should theoretically be possible using a method like this, really. (Though you'd likely end up with thermal management issues)
Surely printing resistors would be "straight forward" - then caps, then transistors. That would be insane. Can you imagine designing a circuit, and specifying all the embedded, printed components, then just hitting "go" and in a few hours you have your board all done? Maybe it will take the fun out of making stuff...
A maskless photolitography stepper with a cheap (used?) spin coating machine can be very good, if the stepper can optically detect alignment after stepping, to then shift the pixels to fix the positioning error you get from cheap-ish mechanics. Combined with e.g. silver nitrate based electroless plating, or even atmospheric plasma cleaning/deposition (gold ions through plasma, condensing on the pcb?), and used in semi-additive pcb manufacturing (e.g., plating a few dozen nm onto the substrate/previous layer, coating with photo resist, exposing, developing, electroplating in revealed parts of the electroless plating, stripping resist, short wet etch to remove the base that is bridging the plated traces, followed by "soldermask" spin coating/exposing/developing to create the next layer's isolating substrate. Repeat. Due to spin-coating naturally evening out any surface roughness, and the optical alignment in the stepper, there is no waging between precision and layercount or such. The only real downside is that copper requires electroless copper plating or a different, suitable base for more than a single layer, and for adhesion the surface below the base plating has to be _very_ clean. Traces down to 5 micrometer should be feasible without exotic optics in the stepper. BOM for the stepper would be
It enables interesting routing options on intermidiate layers, including ability to move between layer without vias, and doing arbitrary plating and thicknesses even on the same "layer". The soldering problem must be solved tho. I wonder what are resistance and capacitance properties of the materials used, resistance to high power elements, heatsinking abilities and moisture resistance.
Thought it was quite cool, but the low temperature kind of ruins the advantages. Never mind the totally different dielectric and high frequency properties of the material. That makes it kind of useless for the more fancy pancy applications (which are incidentally the ones who would fork out that amount of money for such a device).
This technique would also provide all sorts of features that aren’t physically possible with the common fiberglass PCB tech... for example, no need for via stitching adjacent layers... just don’t insulate between the conductive layers (solid conductive “bridge planes” between layers/components where desired). Vias could be made much smaller since they could be printed solid rather than being plated holes. Also, in theory you could actually design in-board capacitors because layer (conductive and dielectric) thickness could be controllable. Practicality might be questionable, but the idea is interesting. Of course this assumes you’re building small-batch production boards with it, because only this machine could produce said features - which kind of negates these concepts since buyers are unlikely to use this machine for production boards. Rolling my eyes at the “higher temps coming soon” claim, sounds like typical vapor ware... same old hollow marketing promises. They don’t even say what the “higher temps” will be... not much of a promise if it only provides an additional 10°. Still, VERY cool bit of tech, and I can see lots of companies springing for this.
Printed Circuit Board printer. Well. Sounds obvious. Seriously once refined (to allow higher temperature soldering), this is a revolution. Because it is a *3D* printer. And boards does not need to be flat anymore.
Amazing potential for this. If you can, it would be cool to see what kind of 'maker' spaces that they have there in your area of Australia. These kind of hacker / maker work spaces where they are tinkering around.
I spoke with these guys a couple of years ago. I remember it being expensive, but not nearly that expensive. The deal breaker was really the performance of the finished result and it being very non representative of a manufactured board. If they can improve the boards it builds I can see it being really useful. As it is you would always be remembered as the engineer who bought the most expensive white elephant in the department.
So... $300k for a machine that give you a prototype board that seems to be nothing at all like the final board. Unless their "secret sauce" is the ability to change the print so it matches the physical parameters of the designed circuit board materials. I can't think of a time when a 16 layer board would be useful, but not having exact high frequency performance matching to the final version is acceptable.
Also don't follow. He was repeating IP a few times. But that's just a matter of getting a confidential agreement. Most local manufacturers can deliver next day or two. So it's only interesting for companies who do an awful lot of last minute prototyping.
I got curious if the dielectric properties are the same as FR4 or other commonly used substrate. Otherwise any trace impedance matching, signal integrity and etc can be impacted by this technology and won't behave as the actual board should. Also defeats rapid prototyping for signal integrity tweaking
Do you think this kind of future tech will ever transition from plastics to ceramics? From my rather basic understanding, it seems like a prototype design system could be made using ceramics and sintering to produce a board with positive deposition tech. -Jake
3d printing with metals and ceramics is already being done commercially :). ...desktop metal is a few 10k $ to make custom 3d printed metal parts in house. ....I think it might be problematic for many different layers though, but perhaps in several steps you could readily adopt this same process to making ceramic boards. ...potentially that's very interesting for very high temperature applications(maybe with some crazy SiC silicon or even diamond in the future) or more economical very very high frequency boards.
We are slowing reaching that Star Trek utopia everyday! You gotta wonder though why there is such a demand for this kind of tech - I get the speed and the ability to keep it in house, I suppose those that "need" it can justify the cost... I wonder will we ever see this thing hooked up to a pick & place machine. That puts the tech more into a Terminator movie lol!
What kind of argument is IP for this? Why would a PCB manufacturer steal a clients designs? It is not even their business to design products and would be a gigantic risk. And if it really was a concern Im sure that for that price you can get some prototype grade manufacturing equipment. Moreover those inks are probably insanely expensive.
Imagine you are Tesla and design your own custom chips and each chip has accelerators that are specially designed to run neural networks. Would you like Chinese to steal the IP? 300k is nothing TSLA can buy 50 of them easily. They have billions of dollars.
@@artstationideas6479 ... Firstly this is about the PCB, not the chips. There is way less IP in PCB designs, from just a PCB it is pretty difficult to reverse the schematic. If the Chinese want to steal IP they'll just buy the product and reverse it after it is assembled, making it much easier. Adding to that is that even if you spun this prototype in house, what are you doing for series? Go to China for that, since you don't have the other machine in house? If you'd already have a PCB line in house, just use that one. That also makes a lot more sense from a 'prototyping' point of view. You don't want to switch to a different manufacturing process for series as that will require redoing a lot of measurements to make sure it meets the spec. PS I'd be surprised if Tesla actually spin their chips in house. That would be a huge waste of time since there is other parties that can do it way better than Tesla ever will (a big reason Apple still does not spin their own chips).
If it was a household appliance with huge consumers base you'll find this at $1K in less than 15y. Maybe $5K in a restricted market like the hobbyists. Time will tell.
Imagine you are Tesla and design your own custom chips and each chip has accelerators that are specially designed to run neural networks. Would you like Chinese to steal the IP? 300k is nothing TSLA can buy 50 of them easily. They have billions of dollars.
I can’t see how it’s there yet, or saleable. Leaded solder, no RoHS compliance, no automation, so where is the commercial value right now to be able to put a price on it?
Brek Martin ROHS, Lead free and fast assembly lines are only required for mass production. This machine is only for one-off boards. But it would still be nice to have a single-run machine place tiny 0201 parts and reflowing at the component's expected temp. profile
Whats the resistance over distance of the extruded traces and what is the dialectric strength? Those always seem to be rough numbers. Resistance especially is high for signal traces.
Hi, Yes ChipQuik is one option and it isn't actually limited to 16 layers, it is limited to 3mm which can be many more than 16 layers depending on layer thicknesses.
I like the future for tech like this, but I don't understand the purpose of the current machine at that price point. Obviously, as stated in the video, it seems the target market is government agencies that wish to 'keep their IP in house' but I guess I'm unsure what usefulness a board that does not meet the specs of the final production run is to such agencies, where typically everything must meet some standard or other. If such agencies have access to a board manufacturer they trust enough to make the production run, why wouldn't they be used for prototyping work. As another poster said, it's all fun and games, until you have to hand solder the thing, and at 140 degrees at that. I just don't understand who the market would be.
It most likely would be used for stuff that they really would want to stay in house and meeting a military/government spec would be the least of their worries.
@John 140°C is current status, other materials (inks) are in development.Any new machine of such capabilities is gonna have a high price point at first, if it takes off, price can easily go to 1/4 or less in a few years.
Ok so it does 16 layers but there is only so many components that will physically fit on the two outside surfaces so do you really need all those other layers to connect the components ?
if you can 3d print this the real advantage would be that it can be in ANY shape. ...probably even have 3d traces. .....and if you could 3d print caps and resistors ....that would open up a new world in extremely compact boards.
Kevin Miedema Ok i see ! It looks like it works a bit like a bubble jet printer and uses a uv cured ink . Strange thing is i was thinking of something similar but having it work more like a laser printer and using a heat source to fuse each layer of plastic instead of using filament or liquid polymers like current 3D printers - never thought of using a printer to make circuit boards ! Circuit boards with almost no external components that can be made to fit in any space , maybe even flexible !
I think I saw a video demo where the operator imported a design file to be printed, then selected and arranged layers, including the top layer which was symbols, shapes and letters. I assume that is the stencil you are referring to?
Why would I buy a $300,000 PCB "prototype" printer that prints PCB's I can't solder when I can just have a company like Bay Area Electronics do real PCB of any type for penny's in a few days? Haha, prototype... Ever hear of a bread board? lol Israel.
I just wish that these companies would stop with this technology because eventually it's going to make the electronic repairs engineer go redundant so why can't make companies just leave this stuff alone
I'd really rather they didn't this kind of technology opens up many new technical possibilities for far advancing what we can do. Repair engineers are already going out the window as product prices are low enough that the time spent & billed is often more then the product itself. We need to double down and accelerate developing this technology, there are huge implications for space travel repairs, more compact electronics and many other things not even thought of.
So... Like about a billion other professions? Yeah, sorry technology has never stopped for the sake of any other career. Don't hold your breath it will stop for that one... Only question is, will there be anything left for humans in the long run.
eh things change ...and repair has already changed. this wouldn't be harder to repair per se. ...and here is the thing, time is valuable - maybe more valuable then recycling and just making a new one for many common objects once you have very very flexibe and good automation. ......our economy has not caught up to this kind of post scarcity society but if it will it will be less of a issue. ....but the diagnostic part of that repair will remain a usefull and powerfull skill way way, way into the future - and even far into the future we will need to maintain alot of older equipment. so its not disappearing; its changing.
Keeping IP in house is a huge deal.
I was shocked when my scientific director allowed me to print my own project on this thing. There are ~10k of them all over the world and only 2 in Russia, so it was incredible possibility for me
I'm feeling 16 layers of old, right now. Awesome tech.
16 layers ? Did I heard that right ?!
60
I want a 160-layer PCB, just lay everything out vertically and jam components on the ends
why not on the sides at that point ? PCB CUBE YAY
Surely at 160 or more layers you'd be getting into the territory where you'd want to embed components IN the PCB itself?
Which... Should theoretically be possible using a method like this, really.
(Though you'd likely end up with thermal management issues)
I've been thinking that components inside of PCBs is the next era of electronics manufacturing.
Thats totaly insane, we are able to print multi layer boards, now we need a way to print components and we're able to print working devices.
Imagine internal bypassing caps built into the layers.
I am sure I read that is already being done.
Surely printing resistors would be "straight forward" - then caps, then transistors. That would be insane. Can you imagine designing a circuit, and specifying all the embedded, printed components, then just hitting "go" and in a few hours you have your board all done? Maybe it will take the fun out of making stuff...
Andre Gulbis Resistors are printed and then laser trimmed IIRC.
Then everything can be as impossible to repair as an Apple product.
Awesome tech. Once temperatures are sorted out this will be a great toy for the big Universities and developers. . .
A maskless photolitography stepper with a cheap (used?) spin coating machine can be very good, if the stepper can optically detect alignment after stepping, to then shift the pixels to fix the positioning error you get from cheap-ish mechanics. Combined with e.g. silver nitrate based electroless plating, or even atmospheric plasma cleaning/deposition (gold ions through plasma, condensing on the pcb?), and used in semi-additive pcb manufacturing (e.g., plating a few dozen nm onto the substrate/previous layer, coating with photo resist, exposing, developing, electroplating in revealed parts of the electroless plating, stripping resist, short wet etch to remove the base that is bridging the plated traces, followed by "soldermask" spin coating/exposing/developing to create the next layer's isolating substrate. Repeat. Due to spin-coating naturally evening out any surface roughness, and the optical alignment in the stepper, there is no waging between precision and layercount or such. The only real downside is that copper requires electroless copper plating or a different, suitable base for more than a single layer, and for adhesion the surface below the base plating has to be _very_ clean. Traces down to 5 micrometer should be feasible without exotic optics in the stepper. BOM for the stepper would be
It enables interesting routing options on intermidiate layers, including ability to move between layer without vias, and doing arbitrary plating and thicknesses even on the same "layer".
The soldering problem must be solved tho. I wonder what are resistance and capacitance properties of the materials used, resistance to high power elements, heatsinking abilities and moisture resistance.
Im jealous of my kids generation
I don't have kids yet but I'm already jealous too.
@@e4r281 me too! I meant future ones!!
Only technologically. Everything else not so much.
Mahdi T.Ranjbar What they gain in technology they lose in just about everything else.
mumble mumble ... hibernation... mumble ... stasis... sigh...
Thought it was quite cool, but the low temperature kind of ruins the advantages. Never mind the totally different dielectric and high frequency properties of the material. That makes it kind of useless for the more fancy pancy applications (which are incidentally the ones who would fork out that amount of money for such a device).
Gotta walk before you can run
so we will all have one of these in 5 years.
Bloody hope so!
More like 15 unfortunately ...
maybe not with with 16 layers, but maybe 2 or 4 layers?
This technique would also provide all sorts of features that aren’t physically possible with the common fiberglass PCB tech... for example, no need for via stitching adjacent layers... just don’t insulate between the conductive layers (solid conductive “bridge planes” between layers/components where desired). Vias could be made much smaller since they could be printed solid rather than being plated holes. Also, in theory you could actually design in-board capacitors because layer (conductive and dielectric) thickness could be controllable. Practicality might be questionable, but the idea is interesting. Of course this assumes you’re building small-batch production boards with it, because only this machine could produce said features - which kind of negates these concepts since buyers are unlikely to use this machine for production boards. Rolling my eyes at the “higher temps coming soon” claim, sounds like typical vapor ware... same old hollow marketing promises. They don’t even say what the “higher temps” will be... not much of a promise if it only provides an additional 10°. Still, VERY cool bit of tech, and I can see lots of companies springing for this.
Printed Circuit Board printer.
Well. Sounds obvious.
Seriously once refined (to allow higher temperature soldering), this is a revolution.
Because it is a *3D* printer. And boards does not need to be flat anymore.
Do not forget that IC components are still flat
Amazing potential for this. If you can, it would be cool to see what kind of 'maker' spaces that they have there in your area of Australia. These kind of hacker / maker work spaces where they are tinkering around.
Just as soon as there's ink! I'll believe it when I see it
I spoke with these guys a couple of years ago. I remember it being expensive, but not nearly that expensive. The deal breaker was really the performance of the finished result and it being very non representative of a manufactured board. If they can improve the boards it builds I can see it being really useful. As it is you would always be remembered as the engineer who bought the most expensive white elephant in the department.
"Keeping IP in-house." Holy Batman; Every day, that's becoming a better reason to have one.
Meanwhile Dave sends the Gigatron files to JLCPCB.
Okurka It could be that Gigiatron had their boards made by JLCPCB or a similar service in China.
So... $300k for a machine that give you a prototype board that seems to be nothing at all like the final board. Unless their "secret sauce" is the ability to change the print so it matches the physical parameters of the designed circuit board materials.
I can't think of a time when a 16 layer board would be useful, but not having exact high frequency performance matching to the final version is acceptable.
Also don't follow. He was repeating IP a few times. But that's just a matter of getting a confidential agreement. Most local manufacturers can deliver next day or two. So it's only interesting for companies who do an awful lot of last minute prototyping.
300k can buy an awful lot of prototype boards even if the process worked perfectly. Which it probably doesn’t.
Just the ability to do it.
Dont forget the big about not working in reflow ovens.
@@rasz that's a tiny issue. You just use an ultra low melting point solder
awesome stuff, future is looking awesome
I got curious if the dielectric properties are the same as FR4 or other commonly used substrate. Otherwise any trace impedance matching, signal integrity and etc can be impacted by this technology and won't behave as the actual board should. Also defeats rapid prototyping for signal integrity tweaking
Do you think this kind of future tech will ever transition from plastics to ceramics?
From my rather basic understanding, it seems like a prototype design system could be made using ceramics and sintering to produce a board with positive deposition tech.
-Jake
3d printing with metals and ceramics is already being done commercially :). ...desktop metal is a few 10k $ to make custom 3d printed metal parts in house. ....I think it might be problematic for many different layers though, but perhaps in several steps you could readily adopt this same process to making ceramic boards. ...potentially that's very interesting for very high temperature applications(maybe with some crazy SiC silicon or even diamond in the future) or more economical very very high frequency boards.
Very cool tech!
We are slowing reaching that Star Trek utopia everyday! You gotta wonder though why there is such a demand for this kind of tech - I get the speed and the ability to keep it in house, I suppose those that "need" it can justify the cost... I wonder will we ever see this thing hooked up to a pick & place machine. That puts the tech more into a Terminator movie lol!
How long before hobby version available for $1K or less. I can dream cant I🤔.
Dear Santa...
What kind of argument is IP for this? Why would a PCB manufacturer steal a clients designs? It is not even their business to design products and would be a gigantic risk. And if it really was a concern Im sure that for that price you can get some prototype grade manufacturing equipment. Moreover those inks are probably insanely expensive.
You should remember that there are people for literally everything.
Humans are flawed, there will be someone who steals something.
Imagine you are Tesla and design your own custom chips and each chip has accelerators that are specially designed to run neural networks. Would you like Chinese to steal the IP? 300k is nothing TSLA can buy 50 of them easily. They have billions of dollars.
@@artstationideas6479 ... Firstly this is about the PCB, not the chips. There is way less IP in PCB designs, from just a PCB it is pretty difficult to reverse the schematic. If the Chinese want to steal IP they'll just buy the product and reverse it after it is assembled, making it much easier.
Adding to that is that even if you spun this prototype in house, what are you doing for series? Go to China for that, since you don't have the other machine in house? If you'd already have a PCB line in house, just use that one. That also makes a lot more sense from a 'prototyping' point of view. You don't want to switch to a different manufacturing process for series as that will require redoing a lot of measurements to make sure it meets the spec.
PS I'd be surprised if Tesla actually spin their chips in house. That would be a huge waste of time since there is other parties that can do it way better than Tesla ever will (a big reason Apple still does not spin their own chips).
If it was a household appliance with huge consumers base you'll find this at $1K in less than 15y. Maybe $5K in a restricted market like the hobbyists. Time will tell.
Electronex..a wealth of content for EEVblog2!
Heat resistance is a big problem. Probably hard to solve using printable materials. Otherwise good concept.
So next is 3D printed ICs??🙄
i wonder when we will see a diy version for 300 $...alright, i'll settle for a few K....
Imagine you are Tesla and design your own custom chips and each chip has accelerators that are specially designed to run neural networks. Would you like Chinese to steal the IP? 300k is nothing TSLA can buy 50 of them easily. They have billions of dollars.
Everything is fun and games until you realize you have to solder all the stuff by hand :'(
He said there is special leaded solder. Lower melting point for use in ovens.
To survive hand soldering. Still not reflow.
Actually, the website says it reflow compatible.
I can’t see how it’s there yet, or saleable. Leaded solder, no RoHS compliance, no automation, so where is the commercial value right now to be able to put a price on it?
Brek Martin ROHS, Lead free and fast assembly lines are only required for mass production. This machine is only for one-off boards. But it would still be nice to have a single-run machine place tiny 0201 parts and reflowing at the component's expected temp. profile
Whats the resistance over distance of the extruded traces and what is the dialectric strength? Those always seem to be rough numbers. Resistance especially is high for signal traces.
i shell stick to jlc and perfboard
140°? what do you use as solder? ChipQuick?
And why is it necessarily limited to 16 layers?
Layer/board thickness I guess
Hi, Yes ChipQuik is one option and it isn't actually limited to 16 layers, it is limited to 3mm which can be many more than 16 layers depending on layer thicknesses.
Don't turn it on, take it apart!
I wonder how long will it be before they will be able to print components into the board ????
Now that cool.😮
Can you please suggest an alternative to this printer with similar specs?
Who solders 16-layer PCBs by hand? I'm just asking for a friend.
So at the rate 3d Printing is moving, in a few generations of Prusa machine we could have this!?
I like the future for tech like this, but I don't understand the purpose of the current machine at that price point. Obviously, as stated in the video, it seems the target market is government agencies that wish to 'keep their IP in house' but I guess I'm unsure what usefulness a board that does not meet the specs of the final production run is to such agencies, where typically everything must meet some standard or other. If such agencies have access to a board manufacturer they trust enough to make the production run, why wouldn't they be used for prototyping work. As another poster said, it's all fun and games, until you have to hand solder the thing, and at 140 degrees at that. I just don't understand who the market would be.
It most likely would be used for stuff that they really would want to stay in house and meeting a military/government spec would be the least of their worries.
@John 140°C is current status, other materials (inks) are in development.Any new machine of such capabilities is gonna have a high price point at first, if it takes off, price can easily go to 1/4 or less in a few years.
now is it cheaper than jlb pcb for a couple of boards
But can it print me a coffee?
Ok so it does 16 layers but there is only so many components that will physically fit on the two outside surfaces so do you really need all those other layers to connect the components ?
if you can 3d print this the real advantage would be that it can be in ANY shape. ...probably even have 3d traces. .....and if you could 3d print caps and resistors ....that would open up a new world in extremely compact boards.
Kevin Miedema
Ok i see ! It looks like it works a bit like a bubble jet printer and uses a uv cured ink .
Strange thing is i was thinking of something similar but having it work more like a laser printer and using a heat source to fuse each layer of plastic instead of using filament or liquid polymers like current 3D printers - never thought of using a printer to make circuit boards ! Circuit boards with almost no external components that can be made to fit in any space , maybe even flexible !
that would be a nightmare to repair though
@@km5405 auto repair.. crypto talk language.. fix me digitally
Can it print a stencil for soldering, too, or would you have to get that manufactured externally?
I think I saw a video demo where the operator imported a design file to be printed, then selected and arranged layers, including the top layer which was symbols, shapes and letters. I assume that is the stencil you are referring to?
ST's isolinier chips are now real?
They should have found the right polymer that can withstand reflow temperature before announcing the product. It sounds like a half-baked product.
300k is great! if only they where high temp boards.
I bet that you are looking for a 3D PCB printer.
I'm not being funny about this new technology but I would not trust it as far as I could throw it
darn.... why do the rich get all the fun toys...
Because they have enough money to buy them...
:o
gets my pants wet and its not pee^^
Why would I buy a $300,000 PCB "prototype" printer that prints PCB's I can't solder when I can just have a company like Bay Area Electronics do real PCB of any type for penny's in a few days? Haha, prototype... Ever hear of a bread board? lol Israel.
With added Mossad...
I just wish that these companies would stop with this technology because eventually it's going to make the electronic repairs engineer go redundant so why can't make companies just leave this stuff alone
I'd really rather they didn't this kind of technology opens up many new technical possibilities for far advancing what we can do. Repair engineers are already going out the window as product prices are low enough that the time spent & billed is often more then the product itself. We need to double down and accelerate developing this technology, there are huge implications for space travel repairs, more compact electronics and many other things not even thought of.
So... Like about a billion other professions?
Yeah, sorry technology has never stopped for the sake of any other career. Don't hold your breath it will stop for that one...
Only question is, will there be anything left for humans in the long run.
eh things change ...and repair has already changed. this wouldn't be harder to repair per se. ...and here is the thing, time is valuable - maybe more valuable then recycling and just making a new one for many common objects once you have very very flexibe and good automation. ......our economy has not caught up to this kind of post scarcity society but if it will it will be less of a issue. ....but the diagnostic part of that repair will remain a usefull and powerfull skill way way, way into the future - and even far into the future we will need to maintain alot of older equipment. so its not disappearing; its changing.
I love Israel :)