Printed electronics

Terraformer · 2018-09-23 15:30:47

We might as well push for complete self-sufficiency, and see what the limits are...

So, printed electronics, in particular microprocessors. They're not as powerful as the ones modern chip fabs produce, by far, but they should be able to provide more power than the Apollo program had available - and hopefully enough to run the desktop chip printer, closing the loop and enabling the first desktop self-replicating industrial ecosystem.

https://www.printedelectronicsworld.com … nology-now
https://spectrum.ieee.org/tech-talk/sem … ransistors

It seems that Zinc Oxide can be suspended in an ink and used to print transistors onto a plastic backing. Zinc is relatively abundant, and simple enough to refine that we've been doing it since at least the 12th century. I don't know about the chemistry involved in forming the ink, though.

Terraformer · 2018-09-23 15:33:22

Also - http://www.3ders.org/articles/20160408- … -inks.html

SpaceNut · 2018-09-23 15:42:18

Have worked in the electronic industry and metal on metal oxides can be used to make semiconductor parts but printing them will make the devices much larger than the standard way of doing them. We have been able to print the carbon resistors and make the electrical connections as well with the technology. Though these do not handle the same amounts of current.
The metal on metal oxides do create capacitors at material change over layers. Its still going to be a period of time before this will be a main stay to making electronics.

kbd512 · 2018-09-23 19:54:45

SpaceNut,

The thin film discrete component electronics in question may be larger and consume more power than microelectronics providing equivalent function, but if the resultant systems are so simple, modular, and resistant to radiation and temperature effects that their fabrication or repair is essentially a one-and-done proposition requiring a single tool (like a specialized 3D printer), then perhaps critical systems like environmental controls and communications equipment would benefit from that type of architecture because they're comparatively simplistic to troubleshoot and more robust.

Remember how the Star Trek computers were basically a bunch of plastic semiconducting ribbons or boards that could be replaced to restore function and there was a specialized core computer system deep within the engineering section of the ship? That's what I think we need. The core functionality can be provided by specialized (and expensive) radiation-hardened system-on-chip type microelectronics, whereas all the sensors and ancillary electronics assemblies (similar to the old avionics subsystem boxes in fighter jets that all handled discrete functions, like a box to control the chaff or flare dispensers, for example) should be easily field serviceable / repairable without the requirement for a full electronics repair shop or special cleanliness requirements.

SpaceNut · 2018-09-23 20:03:35

Metal on metal oxides are easily damaged by static discharge, they are not radiation hardened, are easily damaged by repetition flexing, and would mean going back to a bus card structure to a systems computer which is not a bad thing. Just not where we are with single board computers which would be turning the clock back almost 30 years to when they were.

louis · 2018-09-24 16:50:23

Could be v. useful particularly in the early stages of colonial development when labour is such short supply.

But I don't see why a Mars colony couldn't, with the help of 3D printing of the machinery, build a sophisticated microprocessor manufacturing facility on a scaled down version of this sort of factory:

https://www.youtube.com/watch?v=qm67wbB5GmI

Terraformer wrote:

We might as well push for complete self-sufficiency, and see what the limits are...
So, printed electronics, in particular microprocessors. They're not as powerful as the ones modern chip fabs produce, by far, but they should be able to provide more power than the Apollo program had available - and hopefully enough to run the desktop chip printer, closing the loop and enabling the first desktop self-replicating industrial ecosystem.
https://www.printedelectronicsworld.com … nology-now
https://spectrum.ieee.org/tech-talk/sem … ransistors
It seems that Zinc Oxide can be suspended in an ink and used to print transistors onto a plastic backing. Zinc is relatively abundant, and simple enough to refine that we've been doing it since at least the 12th century. I don't know about the chemistry involved in forming the ink, though.

Terraformer · 2020-05-09 12:26:45

A company promising semiconductor printing

Printing of micro and nanoscale electronics and sensors (down to 20nm).

Hmm. The missing piece in the self-replicating tech stack?

tahanson43206 · 2023-08-24 08:41:31

I would like to see this topic continued and developed.

The topic of electronic reliability is top of mind for me, as one of my ancient systems has started acting flaky, and after a relatively new power supply began causing random system shutdowns due to an unreliable supply of a particular voltage, on another system.

The folks who set up shop on Mars are going to need reliable electronics that lasts longer than the 7 years that seems to be an average for system boards on Earth in 2023.

We have a member (marc) who was working on a PhD thesis project to develop an ultra reliable computer based upon TTL chips. The system would have been slow compared to modern integrated circuits, but in the context of Mars, it would have been repairable at the chip level, and fabrication of such primitive chips would be easier to set up on Mars than a modern high density integrated circuit manufacturing facility.

(th)

Calliban · 2023-08-25 16:25:07

I hadn't heard of printed electronics before. It could potentially be a very useful replacement for the low end electronics that are coming out of China. Those electronics are very likely to become unavailable as China collapses, which is looking increasingly likely.

On Mars, the ability to print simple electronic devices would provide a huge advantage. Simple linear controllers would useful. Temperatures controllers, sunlight tracking systems, things like that.

tahanson43206 · 2023-08-25 17:13:23

For Terraformer ... re topic ...

If you have time, please see if there might be updates available ...

It seems to me this is an important topic ... it is (or appears to be) an early step toward the StarTrek replicator, which assembles atoms into desired configurations at breath taking speed. We humans can see atom placement working all the time, in every plant and tree, but the process is languid in comparison.

A 3D printer for Integrated Circuits would be an important step forward, and I would think there would be a market for chips made that way right now.

As a reminder, NewMars member marc is working on a TTL based minicomputer designed to operate with 36 bit words, in a ** very ** robust manner. We haven't heard from marc for a while, but I assume he's been working on the system all this time.

The project involved writing an assembler, as well as designing all the hardware.

(th)

tahanson43206 · 2023-08-28 06:45:33

For Terraformer re topic ...

Here is a report on advances in the art of 3D printing of integrated circuits...

It appears that current technology can print at micron scale. The goal appears to be to print at nano scale.

https://www.nano-di.com/resources/blog/ … the-future

The capability described in this article would be helpful for a small community on Mars, where independence from Earth as soon as possible would be advantageous.

(th)

kbd512 · 2023-08-28 16:05:52

I think the idea of a simplified computing system, lasting for decades because it's not based upon the sort of microcircuitry that the universe is so unkind to (galactic cosmic rays, solar radiation, static discharge destruction over time), is greatly under-appreciated. What I view as a "future proof" fault-tolerant computing system, though, will ultimately be based upon long-lasting materials (glass or diamond or ceramic metals) and the use of very specific wavelengths of light both as input and for processing, rather than very precise electrical control.

For example, Microsoft's light-based AIM computing technology doesn't store information, it continuously processes it and reacts according to changes in light patterns. It requires very little power relative to Silicon-based electronics, it's focused on processing rather than storing data, and the volume of throughput it can process is wildly in excess of what integrated circuits with the same size and input power are capable of, because multiple wavelengths of light are not limited to a binary state. AIM doesn't store data in registers, operating more like an information sieve intended perform a particular computational task to generate usable output. AIM is very much like combining the function of the sensor and firmware of an IR-guided missile. The missile's sensor input is used for pattern matching something that represents the shape of an enemy aircraft and wavelength range of a jet engine exhaust, which is normally very different from a flare. AIM is a far more sophisticated implementation that doesn't clock-driven switching to match sensor input with database values to produce the output using more clock cycles. The processor / sensor / database are one and the same thing, not subject to voltage regulation, binary data storage, switching, and a clock controlling when the entire process can repeat. "System-on-Chip" is close in concept, if the sensor was also part of the same chip.

I had a hard time wrapping my head around how this could be used as part of a control system when Terraformer or Void or someone else here first presented the link. Raytheon's AIM-9X Sidewinder is similar to a digital camcorder with a modestly high-rate, given its speed, doing post-processing of the video and then responding according to what it "sees" (voltage changes and clock-driven switching, ultimately). Microsoft's AIM is what a human eyeball and brain combination would be capable of interpreting if they could operate at a frame rate nearing the speed of light reflectance from the target, while "seeing" visible spectrum / IR / UV at the same time. Without a clock cycle limitation telling a high speed switching machine (microchip) when it can process the next image, it responds as soon as the next photons strike the sensor. The actuators moving the control fins become the limiting factor.

Apart from weapons technology, this is exactly how we need control electronics for life support to function. A sensor determines CO2 content of the atmosphere, then turns on a fan motor to pump air through a filter to remove it, and doesn't require a highly specific voltage value to determine that cabin air is becoming unbreathable. A light-based sensor / computer knows a contaminant is affecting air molecule reflectivity and responds by turning on the fan motor. If electronics were the solution, then static discharge could prevent operation, GCR could fuse the switches in the microchip, an electrical connection failure from thermal cycling can occur, and the voltage-regulated sensor is dependent upon the generation of in-spec voltage deltas, rather than some intrinsic physical phenomenon directing light down a different path through the information sieve.

The voltage-dependent O2 sensors in a vehicle emissions control system can't actually discern the difference between an engine leaking coolant into the exhaust and increased O2 content signaling that too much uncombusted fuel is leaving through the exhaust. Both conditions produce voltage values that the engine control computer interprets as an emissions problem. Both conditions are probably bad for the engine, but variant voltage values aren't telling you what is actually happening. You can spend a lot of money chasing after a non-specific problem. An optics-based sensor and computer can easily tell the difference between water and fuel, as both reflect light in different ways, because they're different molecules. They used to do actual emissions tests to figure out what was coming out of your vehicle's tailpipe. Now they look at voltage readings from onboard sensors, which don't indicate that something specific has happened.

Why is this differentiated behavior so important for a long-term viable colony in space?

If you rely solely on a voltage-dependent microcircuit switching in response to very small voltage changes to decide what the atmospheric pressure or CO2 level should be, said control system ceases to be "life support" and becomes "death support" after "the unkind universe" (GCR, solar flare, etc) causes aberrant indirect sensor voltage readings or induces a voltage spike sufficient to fry the microchip. It's harder to cause a light-based sensor and computer to malfunction. The computer is rerouting light down a different path when its reflectance value changes in direct response to the reflectance value of the gas composition in the cabin changing.

SpaceNut · 2023-08-29 10:55:14

The use of materials in the build of circuit boards is not new and has been ongoing for at least 2 plus decades that I am aware of. The depositing of carbon to make resistors is not new as well as metal inks used to connect the circuits with in them. The use of metal on metal oxides as well have been known and used to make capacitors.
All this is just used to auto mate the build with hopeful lowering of costs but it also does rise the hazardous chemicals used in the processing.

New Mars Forums

Announcement

#1 2018-09-23 15:30:47

Printed electronics

#2 2018-09-23 15:33:22

Re: Printed electronics

#3 2018-09-23 15:42:18

Re: Printed electronics

#4 2018-09-23 19:54:45

Re: Printed electronics

#5 2018-09-23 20:03:35

Re: Printed electronics

#6 2018-09-24 16:50:23

Re: Printed electronics

#7 2020-05-09 12:26:45

Re: Printed electronics

#8 2023-08-24 08:41:31

Re: Printed electronics

#9 2023-08-25 16:25:07

Re: Printed electronics

#10 2023-08-25 17:13:23

Re: Printed electronics

#11 2023-08-28 06:45:33

Re: Printed electronics

#12 2023-08-28 16:05:52

Re: Printed electronics

#13 2023-08-29 10:55:14

Re: Printed electronics

Board footer