Kris Erickson | Meta: Is two-photon micro-printing the ultimate solution to the optical chip-to-fiber packaging bottleneck?
00:10:24 - 00:10:56
Other snippets from this talk
Summary of the clip:
Is two-photon micro-printing the ultimate solution to the optical chip-to-fiber packaging bottleneck?
Silicon photonics and high-bandwidth optical computing face a severe assembly bottleneck: aligning and interconnecting optical chips to output fibers with sub-micron accuracy. Traditional physical fiber coupling is expensive, mechanically sensitive, and limits the packaging density of high-speed co-packaged optics.
To solve this, additive manufacturing leverages ultra-high-resolution two-photon polymerization (2PP). This precise 3D lithography technique allows engineers to print specialized optical polymers directly on top of active photonic chips, effectively creating the optical equivalent of an electrical wire bond.
These 3D micro-printed waveguides bridge the physical gap between on-chip laser sources and external fiber arrays. By bypassing classical mechanical alignment limits, this approach enables compact, low-loss, and high-yield optical compute packages suitable for next-generation data centers and AI accelerators.
In this short video, you can learn:
* Why physical fiber-to-chip alignment acts as the primary barrier to high-bandwidth optical compute packaging.
* How two-photon polymerization (2PP) prints sub-micron 3D polymer structures directly on active silicon.
* The mechanics of "optical wire bonding" as an alternative to traditional optical interconnect assemblies.
đź“‹ **Clip Abstract**
Kris Erickson details a novel application of additive manufacturing for optical compute packaging. By using two-photon polymerization, engineers can print optical wire bonds that solve the high-bandwidth chip-to-fiber coupling bottleneck.
#TwoPhotonPolymerization, #OpticalWireBonding, #SiliconPhotonics, #3DPrintedWaveguides, #CoPackagedOptics, #PhotonicPackaging
This is a highlight of the presentation:
More Highlights from the same talk.
06:36 - 08:25
Why do 3D-printed circuits need a new CAD toolchain instead of existing EDA and mechanical CAD?
Why do 3D-printed circuits need a new CAD toolchain instead of existing EDA and mechanical CAD?
Conventional EDA tools are designed for flat two-dimensional PCBs, while mechanical CAD tools such as SolidWorks handle shape and form without electronics. As printed parts grow more complex, designing multilayer boards and complex traces inside a three-dimensional part becomes cumbersome across that divide. The mismatch appears early when traces, components and mechanical geometry must be co-designed rather than handled in separate environments.
The effort uses Rhino with a plugin in an open-source CAD approach. Circuit data from EDA tools such as Cadence and Eagle is brought in; users click where each component should sit and then auto-route the traces. When the three-dimensional design changes, they can regenerate the configuration, aiming for a result optimal both for electronic performance and for printability.
The key difference is three-dimensional routing. Traces can be produced in all three dimensions rather than staying on a two-dimensional plane with vias going down. The design can route organically throughout the part. That changes the design problem from placing flat PCB traces to generating a printable, electrically functional three-dimensional interconnect within the mechanical form.
In this short video, you can learn:
* Existing EDA tools target flat 2D PCBs, and mechanical CAD tools such as SolidWorks lack electronics, so complex 3D parts expose a toolchain gap.
* The workflow imports circuits from Cadence or Eagle into a Rhino plugin, places components by clicking, and auto-routes traces.
* Changing the 3D design allows regeneration of the trace configuration for electronic performance and printability, with routing in all three dimensions rather than 2D plus vias.
đź“‹ **Clip Abstract** Complex 3D-printed electronics expose a mismatch between 2D-oriented EDA tools and electronics-free mechanical CAD. A Rhino plugin effort imports Cadence or Eagle circuits, enables click placement and auto-routing, regenerates designs when the 3D geometry changes, and supports organic three-dimensional traces instead of 2D routing with vias.
About the speaker:
* Speaker: Kris Erickson
* Company: Hewlett Packard
* Event: Printed, Hybrid, Structural & 3D Electronics 2021
* Location: Online
Join TechBlick - the home of printed electronics, additive electronics, flexible hybrid electronics and 3D electronics
https://www.techblick.com/
#RhinoCADPlugin, #EDAToolchain, #AutoRouteTraces, #3DInterconnectRouting, #PrintedElectronics, #3DPrintedElectronics
16:29 - 18:27
Can automated pick-and-place reliably connect 24-pad QFNs inside 3D-printed sockets?
Can automated pick-and-place reliably connect 24-pad QFNs inside 3D-printed sockets?
The process creates sockets—regions where powder is later removed—so components can be placed into the printed part rather than only on its surface. The speaker notes that socket mounting can improve mechanical integrity, illustrating an LED placed in a socket and a resistor on the other side. This moves post-placement beyond manual assembly toward conventional automated pick-and-place and solder-paste deposition used in PCB manufacturing.
Using automated solder-paste deposition and automated component placement, the work pushed limits on component size and pad-to-pad spacing. The reported result: 24-pad QFNs were working most of the time, with well over 50% achieving all 24 pads successfully connected. The speaker frames this as still improvable, while two-terminal devices—resistors and LEDs—were successfully connected all the time after process optimization.
The print approach also allows a range of capacitors that can be printed, extending the discussion from interconnect and placement to printed components. The component workflow combines printed sockets, solder paste, and pick-and-place, with surface mounting as an alternative to socket mounting. Mechanical integrity is the stated reason sockets can be beneficial, especially when placing parts such as LEDs and resistors into a printed structure.
In this short video, you can learn:
* Printed sockets are regions where powder is removed so components can be placed into the part, with socket mounting optionally improving mechanical integrity.
* Automated solder-paste deposition and pick-and-place achieved well over 50% success connecting all 24 pads on 24-pad QFNs, though this remains improvable.
* Two-terminal resistors and LEDs were successfully connected all the time after process optimization.
đź“‹ **Clip Abstract** The clip explains how 3D-printed sockets allow components to be placed into printed parts, with socket mounting used for mechanical integrity and surface mounting also possible. It then reports automated solder-paste and pick-and-place results: 24-pad QFNs connected all pads well over 50% of the time, while two-terminal resistors and LEDs connected successfully all the time after optimization.
About the speaker:
* Speaker: Kris Erickson
* Company: Hewlett Packard
* Event: Printed, Hybrid, Structural & 3D Electronics 2021
* Location: Online
Join TechBlick - the home of printed electronics, additive electronics, flexible hybrid electronics and 3D electronics
https://www.techblick.com/
#AutomatedPickAndPlace, #SolderPasteDeposition, #24PadQFN, #PrintedComponentSockets, #PrintedElectronics, #3DElectronics
13:10 - 15:26
How wide a resistance range can you print by blending conductive and resistive agents in one layer?
How wide a resistance range can you print by blending conductive and resistive agents in one layer?
The printed resistor approach pairs a good conductive material with a less conductive, more resistive material that can be doped in. Design files print down as a composite rather than a single color, using both agents in different ratios within the same region. That ratio tunes high connectivity, low connectivity and the states between, with resistors shown in XY and Z orientations and at higher and lower conductivity.
The demonstration spans an entire nine orders of magnitude in bulk resistance, although tuning and tight resistor tolerances are described as earlier-stage work. With that range, geometry can extend performance further: a serpentine versus a smaller feature could theoretically go beyond it. That breadth supports the claim that the approach can meet any conventionally requested resistor.
The nine-order range also permits typical requested resistor values to be less than a square millimeter or a cubic millimeter in volume. Tight tolerances are still in progress, so the resistor work is presented as range-capable rather than fully dialed in. The next passive examined is the printed capacitor.
In this short video, you can learn:
* Blending conductive and resistive agents in different ratios within one region tunes connectivity from high to low.
* The printed resistor approach demonstrated a bulk-resistance span of nine orders of magnitude.
* Geometry, such as serpentine versus smaller features, could extend that resistance range further.
đź“‹ **Clip Abstract** A printed resistor approach blends a good conductive material with a more resistive material that can be doped in, using different ratios inside the same printed region to tune connectivity in XY and Z. It demonstrates a nine-order-of-magnitude bulk-resistance span, with geometry offering further range and tight tolerances still in progress.
About the speaker:
* Speaker: Kris Erickson
* Company: Hewlett Packard
* Event: Printed, Hybrid, Structural & 3D Electronics 2021
* Location: Online
Join TechBlick - the home of printed electronics, additive electronics, flexible hybrid electronics and 3D electronics
https://www.techblick.com/
#PrintedResistor, #ConductiveResistiveBlending, #BulkResistanceRange, #NineOrdersMagnitude, #PrintedElectronics, #AdditiveElectronics
Other talks by Kris Erickson
- Additive Manufacturing for Future High Volume Manufacturing of Electronic Devices · The Future of Electronics RESHAPED | Berlin 2023
- 3D Printing of Electronics using Multi Jet Fusion · Printed, Hybrid, Structural, & 3D Electronics 2021
More highlights from Kris Erickson
- ▶ Can additive manufacturing resolve the paradox of scaling sub-micron features without sacrificing high-volume throughput? · The Future of Electronics RESHAPED 2023 Berlin
- ▶ Why do micro-printed dielectric materials suffer catastrophic mechanical failure during standard package reflow? · The Future of Electronics RESHAPED 2023 Berlin
- ▶ Why do 3D-printed circuits need a new CAD toolchain instead of existing EDA and mechanical CAD? · Printed, Hybrid, Structural & 3D Electronics 2021
- ▶ How wide a resistance range can you print by blending conductive and resistive agents in one layer? · Printed, Hybrid, Structural & 3D Electronics 2021
- ▶ Can automated pick-and-place reliably connect 24-pad QFNs inside 3D-printed sockets? · Printed, Hybrid, Structural & 3D Electronics 2021
- ▶ Do printed conductive traces compromise Nylon 12 mechanical strength or sheet resistance? · Printed, Hybrid, Structural & 3D Electronics 2021


