PDA

View Full Version : Venus rovers?



lpetrich
2015-May-22, 04:38 AM
Of the Earth's two neighbors, Venus has been explored much less than Mars, for rather obvious reasons. The champion survivor on that planet has been Venera 13, at about 127 minutes of radio transmission after landing. However, some Mars landers and rovers have survived on that planet for years, with the champion being the Opportunity rover at 11 years.

Venus's surface has a pressure of about 92 bars (about 92 * Earth's atmosphere pressure) and a temperature of about 450 C (Atmosphere of Venus (https://en.wikipedia.org/wiki/Atmosphere%20of%20Venus)). Almost but not hot enough to glow (Draper point (https://en.wikipedia.org/wiki/Draper%20point), about 525 C). Its atmosphere is mostly CO2 with a few percent N2, with not much O2.

Windsurfing on a Wicked World | NASA (http://www.nasa.gov/directorates/spacetech/home/feature_windsurfing.html)


But one part of the Landis NIAC study is focused on using wind force on Venus as a propulsive nudge. While the winds at the surface of Venus are low (under one meter per second, or just a little over two miles per hour), at Venus pressure, even low wind speeds develop significant force, he explains.

"A sail rover would be extraordinary for Venus. The sail has only two moving parts-just to set the sail and set the steering position-and that doesn't require a lot of power. There's no power required to actually drive," says Landis.

Venus's surface is also fairly flat, something that also helps. Though it is perpetually clouded over, its clouds are above about 30 km / 18 mi, meaning good visibility near the ground.

Venus's surface temperature is rather mild by some industrial standards, and I've easily found lubricants capable of functioning at the planet's surface temperature. However, electronics is another story. One has to generate electricity, and then use it to move the sail, run the onboard computer, take pictures, make various measurements, and then radio the results back to an orbiter or directly to the Earth. Semiconductor components are rather obviously necessary, and preferably thousands or millions on a chip. One will have to demonstrate high-temperature versions of:

CPU chips, volatile-memory (RAM) chips, persistent-memory (Flash, etc.) chips, CCD (for cameras) chips, photovoltaic chips.

Design of high-temperature semiconductor electronic components is still in a very limited state (High-Temperature Electronics (http://www.analog.com/library/analogdialogue/archives/46-04/high_temp_electronics.html) at Analog Devices, Extreme-Temperature Electronics (Tutorial - Part 1) (http://www.extremetemperatureelectronics.com/tutorial1.html)). Some commercially-available components are rated for temperatures of 200 C or even 300 C, but they look like rather simple ones. There are various things that one can do to improve temperature tolerance, like trench isolation and silicon-on-insulator. Looking at alternative materials, silicon carbide goes up to over 600 C, though only simple components have been demonstrated with it, like power transistors.

Fun fact: a favorite way of torture-testing electronic components is to run them at high temperatures. That makes them fail much faster.

For storing electrical energy, one can use high-temperature batteries. A sodium–sulfur battery (https://en.wikipedia.org/wiki/sodium%E2%80%93sulfur%20battery) operates at 300 - 350 C, so a 450-C battery may be feasible.

For magnetic components, Venus's surface temperature is well under the Curie temperature (https://en.wikipedia.org/wiki/Curie_temperature) of several ferromagnetic materials, the temperature where the ferromagnetism disappears, so there should be no problem there.


So while a Venus rover is technically feasible, it requires semiconductor parts that have yet to be demonstrated.

joema
2015-May-23, 11:54 PM
...Venus's surface...temperature of about 450 C...One will have to demonstrate high-temperature versions of:...CPU chips, volatile-memory (RAM) chips....Design of high-temperature semiconductor electronic components is still in a very limited state....while a Venus rover is technically feasible, it requires semiconductor parts that have yet to be demonstrated.

It definitely does not require high temperature electronics that can endure 450 C. A major design priority of spacecraft, whether manned or unmanned, is maintaining thermal control of delicate electronics. Designers have never been limited by electronics which must endure an unregulated ambient space or planetary thermal environment. Whether the icy -290F cold that Cassini Huygens endured, or the 2,500 F the upcoming Solar Probe Plus will endure http://solarprobe.jhuapl.edu/, the electronics in those spacecraft are carefully insulated from ambient conditions. The same would be so for a future Venus rover.

cjameshuff
2015-May-24, 02:46 AM
As joema points out, active cooling can keep sensitive components cool. There's a hefty thermal gradient to deal with, but it's doable. Additionally, you only really need enough brains to communicate with probes in orbit or in the upper atmosphere.

High temperature SiC JFET integrated circuits have been constructed (http://spectrum.ieee.org/semiconductors/devices/silicon-carbide-logic-circuits-work-at-blistering-temperatures). If you have integrated circuitry, you can do ROM/PROM. Non volatile writable storage would be trickier, but not particularly necessary.

RTGs would be a bit less efficient, but would still provide useful amounts of power, and you wouldn't need much radiator area at all...the atmosphere would be a very effective heat sink.

Sodium-sulfur batteries are the result of research into reducing the operating temperatures of molten salt batteries, some older chemistries would require an additional heat source on Venus.

lpetrich
2015-May-24, 10:40 AM
As joema points out, active cooling can keep sensitive components cool.
However, on Venus's surface, one is surrounded with 92 bar of 450-C atmosphere. One can't use a sunshade, what one can do in space. One has to run the refrigerator continuously, and extra mechanism is something that a spacecraft does not need.


If you have integrated circuitry, you can do ROM/PROM. Non volatile writable storage would be trickier, but not particularly necessary.
Nonvolatile rewritable storage is useful for reprogramming the system.


RTGs would be a bit less efficient, but would still provide useful amounts of power, and you wouldn't need much radiator area at all...the atmosphere would be a very effective heat sink.
RTG's would require thermocouples capable of functioning at 450 C.


Sodium-sulfur batteries are the result of research into reducing the operating temperatures of molten salt batteries, some older chemistries would require an additional heat source on Venus.
Is that gap unbridgable? If it can be bridged, then one can get Venus-friendly batteries.


Glenn Research Center at Lewis Field Development of High Temperature Silicon Carbide (SiC) Electronics for Intelligent Engine Systems (http://www.eng.morgan.edu/~cibac/events/Day2/IVHM/1-Development%20of%20High%20Temp%20(Hunter).pdf) -- for jet engines.

Seems to be proposing:


Integration of sensor technology with high temperature wireless communications and energy harvesting to enable a smart systems operable at high temperatures.

High-temperature wireless communications based on SiC electronics and rugged RF passive components.

Energy harvesting systems focusing thermo-electric and photo-voltaic materials for generation of power for remote sensors.

High-temperature radio equipment and photovoltaic cells -- what a Venus rover will need. This presentation also mentions SiC capacitive pressure sensors.

Stability goals:
For testing in a jet engine on the ground: >1000 hr
For most aviation and space applications: >100,000 hr

Also mentions tests of high-temperature-tolerant chip packaging.

High Temperature SiC Electronics: Update and Outlook (http://www.grc.nasa.gov/WWW/cdtb/aboutus/workshop2012/Presentations/Session%203.%20Distributed%20Engine%20Control/DEC_04_Beheim.pdf)
It covers much of the ground of the previous one, but with additional numbers. Like these temperature upper limits:

Ordinary silicon chips: 150 C
Silicon-on-insulator chips: 300 C
SOI only for low-power applications; SiC better for high-power ones

The authors concede that JFET's are limited by their high power consumption, about 1 milliwatt per gate -- can't get very many of them onto a chip.

But SiC JFET's can function all the way down to -125 C. Good for space travel.

Also some stuff on multilevel interconnects -- integrated-circuit internal "wires".

Plans: 4-bit A/D and D/A converters, 2*2 static RAM, op amp, ring oscillator, binary AM radio transmitter


High Temperature Silicon Carbide CMOS Integrated Circuits - rsl_semi_published_hiten2011.pdf (http://www.raytheon.co.uk/rtnwcm/groups/rsl/documents/content/rsl_semi_published_hiten2011.pdf) -- including working on the insulator layers in the transistors. Successful high-temperature CMOS would make it possible to have thousands or millions of transistors on a chip.

Future high temperature applications for SiC integrated circuits - Zetterling - 2012 - physica status solidi (c) - Wiley Online Library (http://onlinelibrary.wiley.com/doi/10.1002/pssc.201100689/abstract) -- also mentions lower on-resistance and possibly being radiation-hardened. If one can get rad-hardness a result of high-temperature tolerance, then that will be an additional win.

cjameshuff
2015-May-24, 12:10 PM
However, on Venus's surface, one is surrounded with 92 bar of 450-C atmosphere. One can't use a sunshade, what one can do in space. One has to run the refrigerator continuously, and extra mechanism is something that a spacecraft does not need.

If it's necessary, then the rover will have one.



Nonvolatile rewritable storage is useful for reprogramming the system.

It's not at all necessary. Store enough to talk to spacecraft in orbit, and anything else can be put in volatile memory. Given the constant power available on the surface, it probably wouldn't have to be reloaded often.



RTG's would require thermocouples capable of functioning at 450 C.

GPHS-RTGs already have themocouples capable of operating at around 1000 °C. Bumping the cold side from ~300 °C to 460 °C would hurt efficiency, but it'd still be an effective power source.

Additionally, Stirling engines could be used to convert power at somewhat higher efficiency, and you could link Stirling machines together or use some high temperature variation of an Einstein refrigerator to provide constant cooling for the parts of the craft that need it, possibly with fewer losses than converting to electrical power to run a cooler.



Is that gap unbridgable? If it can be bridged, then one can get Venus-friendly batteries.

The point is that there isn't a gap. The challenge will be in selecting a chemistry that can perform for extended periods, not one that can perform at the temperatures experienced on the surface of Venus.



The authors concede that JFET's are limited by their high power consumption, about 1 milliwatt per gate -- can't get very many of them onto a chip.

I found that a bit of a strange statement. JFETs themselves aren't known for high power consumption, their gates are not generally forward biased under normal operation as BJTs are. It may be that they are doing some form of RTL logic, rather than using complementary N- and P-channel JFETs, or they simply haven't optimized things for power yet. In addition to the CMOS work, there's work being done on complementary JFET logic that wouldn't have the troublesome issues with making an insulating layer that can survive the temperature swings:
https://theses.ncl.ac.uk/dspace/handle/10443/2233

At any rate, 1 mW/transistor would let you build a simple single-chip CPU.

publiusr
2015-May-24, 07:14 PM
What about lubrication? Something with a higher melting point? I wonder if lubrication choices for wheels might actually be easier for Venus rovers than for those on Mars

cjameshuff
2015-May-24, 07:36 PM
What about lubrication? Something with a higher melting point? I wonder if lubrication choices for wheels might actually be easier for Venus rovers than for those on Mars

Ionic liquids (that is, molten salts) can be effective lubricants, there just aren't many that are liquid down to around room temperature. (There are some, though, and they are a target of active research as lubricants: http://www.insidescience.org/content/molten-salts-could-improve-fuel-economy/1492)

I wonder what sorts of high temperature elastomers might be hiding in plain sight as hard, brittle substances...

publiusr
2015-May-24, 07:44 PM
We need some Horta engineers

lpetrich
2015-May-24, 08:51 PM
What about lubrication? Something with a higher melting point? I wonder if lubrication choices for wheels might actually be easier for Venus rovers than for those on Mars
I searched for "high temperature lubricants", and I found a lot of hits for companies selling high-temperature lubricants for industrial use. Like Industrial Lubricants (http://www.kanolabs.com/indLub.html) by Kano Laboratories. That page advertises Pyrolube as being good up to 1800 F or 1000 C, and as having several big companies as satisfied customers.

lpetrich
2015-May-25, 09:08 AM
(An onboard refrigerator)

If it's necessary, then the rover will have one.
There's a reason why spacecraft are designed to have as few moving parts as possible. They are vulnerable to failure.


It's not at all necessary. Store enough to talk to spacecraft in orbit, and anything else can be put in volatile memory. Given the constant power available on the surface, it probably wouldn't have to be reloaded often.
A mission failure waiting to happen. What constant power? It ought to be as autonomous as reasonably feasible, and not expect an orbiter to be in radio range at all times.

This is another planet and its nearby space not a LAN (Local Area Network). Booting off of another computer is feasible for a LAN, because of its very good network connections, but it isn't widely done across larger-scale networks.


GPHS-RTGs already have themocouples capable of operating at around 1000 °C. Bumping the cold side from ~300 °C to 460 °C would hurt efficiency, but it'd still be an effective power source.
I concede here.


Additionally, Stirling engines could be used to convert power at somewhat higher efficiency, and you could link Stirling machines together or use some high temperature variation of an Einstein refrigerator to provide constant cooling for the parts of the craft that need it, possibly with fewer losses than converting to electrical power to run a cooler.
Stirling engines -- moving parts -- avoid using them.

Einstein refrigerator -- what properties of its working fluids determine its minimum and maximum temperatures? Can one select working fluids for getting from 450 C to 100 C or thereabouts?

(molten salt vs. sodium-sulfur: a temperature gap)

The point is that there isn't a gap. The challenge will be in selecting a chemistry that can perform for extended periods, not one that can perform at the temperatures experienced on the surface of Venus.
It'll have to be both.

(stuff on various sorts of transistors)

At any rate, 1 mW/transistor would let you build a simple single-chip CPU.
I agree. One can construct a simple 8-bit chip much like the earliest CPU chips.

Fun fact: Embedded systems often use 8-bit and 16-bit chips despite some people believing them to be obsolete.

However, a camera must be much like the Viking landers' cameras. Here is some stuff on the Viking ones:
Camera, Imager, Viking Mars Lander | National Air and Space Museum (http://airandspace.si.edu/collections/artifact.cfm?object=nasm_A19810661000)
NASA ADS: The Viking Mars lander camera (http://adsabs.harvard.edu/abs/1975SSI.....1..189H)
Those cameras had a more-or-less single-pixel photodiode detector, with light directed onto it by a rotating mirror that did the vertical scanning. This assembly rotated to do the horizontal scanning.

Furthermore, it would have to transmit as it goes, rather than store and transmit later.

I can't find details on the Venera landers' cameras, however.

cjameshuff
2015-May-25, 02:53 PM
(An onboard refrigerator)

There's a reason why spacecraft are designed to have as few moving parts as possible. They are vulnerable to failure.

Moving parts are necessary, and Stirling engines would be one of the lower risk applications that require them, with low mechanical loads and some types not even having any sliding surfaces. They would likely long outlive the wheel bearings, and can provide more power, reducing the amount of radioisotopes required for a mission and allowing more missions to be performed, more mass to be devoted to other areas, etc.



A mission failure waiting to happen. What constant power? It ought to be as autonomous as reasonably feasible, and not expect an orbiter to be in radio range at all times.

This is another planet and its nearby space not a LAN (Local Area Network). Booting off of another computer is feasible for a LAN, because of its very good network connections, but it isn't widely done across larger-scale networks.

You're talking about putting all the hardware required for autonomous operation in the environment most hostile to the function of that equipment, vastly increasing the complexity of the software by requiring it to manage non-volatile storage and perform all the processing tasks required for autonomy, requiring volatile and non-volatile storage for all the extra code, increasing the number of high-temperature transistors with relatively short MTBFs by several orders of magnitude, and relying on some currently-unknown flash process that can function at the surface for which the probe would be one of the first applications, and you're calling a system with a minimalist communications/control system in ROM a failure waiting to happen?

The point of a minimal read-only system with as much as possible being done remotely from orbit is risk and complexity reduction, by moving functionality to a less harsh environment and components with longer expected lifetimes. Exactly how could it cause mission failure? If the rover can't communicate, it's worthless anyway.

As for power, yes, I expect a Venus rover to have constant power. Why wouldn't it be constant? RTGs don't need to rest, and with the upper atmosphere circling the planet every four days, long periods without wind seem unlikely.



Einstein refrigerator -- what properties of its working fluids determine its minimum and maximum temperatures? Can one select working fluids for getting from 450 C to 100 C or thereabouts?

It depends on temperature dependent solubility of one fluid in another, which lets you manipulate partial pressure of a gas separately from the overall temperature of a gas mixture. There's numerous variations of the basic idea. I don't have any specific suggestions for working fluids that would perform at Venus, but the relatively high pressure should boost performance a bit due to the increased density of the gases.

You don't need to get down to 100 C, just ~350 C would allow silicon-on-insulator electronics to function, and you don't have to do the entire temperature delta in one stage, but any portion of it done using the heat source directly as a power source would probably be a considerable efficiency gain over using electrical power.



(molten salt vs. sodium-sulfur: a temperature gap)

It'll have to be both.

The point is it's already one of them. There is no temperature gap. Modern molten salt batteries only operate at temperatures lower than that of Venus as a result of research with the intent of reducing the operating temperature. I wouldn't be surprised if the upper end of the operating range of commercial ZEBRA cells (which are formulated for prolonged operation) is limited by the packaging. The actual chemistry may be limited by the boiling point of sodium, 883 °C.



(stuff on various sorts of transistors)

I agree. One can construct a simple 8-bit chip much like the earliest CPU chips.

Fun fact: Embedded systems often use 8-bit and 16-bit chips despite some people believing them to be obsolete.

The 6502, for one example, only uses 3510 transistors. Comparison with more modern devices is troublesome because the lowest level modern logic design tools generally work with is gates, and there is no fixed relationship between gate and transistor count, but 8-bit AVRs and 32-bit Cortex M0s are both in the 12k gate range, which would still work out to tens of watts at the stated power consumption...not infeasible to handle.



However, a camera must be much like the Viking landers' cameras. Here is some stuff on the Viking ones:
Camera, Imager, Viking Mars Lander | National Air and Space Museum (http://airandspace.si.edu/collections/artifact.cfm?object=nasm_A19810661000)
NASA ADS: The Viking Mars lander camera (http://adsabs.harvard.edu/abs/1975SSI.....1..189H)
Those cameras had a more-or-less single-pixel photodiode detector, with light directed onto it by a rotating mirror that did the vertical scanning. This assembly rotated to do the horizontal scanning.

Furthermore, it would have to transmit as it goes, rather than store and transmit later.

I can't find details on the Venera landers' cameras, however.

We can do a fair bit better now with compressed sensing techniques that work with sparse subsets of samples (possibly but not necessarily single pixel samples) scattered across the field of view. In the case of a rover, you can also correlate a sparse set of samples with what you know about the surroundings from previous samples instead of having to stop and reconstruct a new snapshot completely from scratch. The tradeoff is that interpreting the data requires a great deal of processing power.

And I thought you wanted an autonomous rover...how would you achieve that without being able to store an image of the surroundings?

John Mendenhall
2015-May-25, 09:35 PM
How about airship floater missions first? Then you could work your way down through the atmosphere mission by mission, developing and testing the technology as you go.

IIRC, the russians had the devil of a time finding out that the atmosphere was simply burming up their probes on the way down.

cjameshuff
2015-May-25, 11:45 PM
How about airship floater missions first? Then you could work your way down through the atmosphere mission by mission, developing and testing the technology as you go.

IIRC, the russians had the devil of a time finding out that the atmosphere was simply burming up their probes on the way down.

Venera 1 and 2 were flyby probes that failed long before reaching Venus. Venera 3 failed during reentry. Venera 4 operated until it was crushed, and was initially claimed to have reached the surface, but Mariner 5 showed that the surface pressure was far too high. Venera 5 and 6 were successful atmospheric probes, Venera 7 was the first attempt at a lander and it was successful despite a problem with antenna alignment. No Venus lander has ever burned up or been crushed, and one of the Pioneer Venus Multiprobes survived impact and operated for over an hour on the surface despite being designed as an atmospheric probe...the other three all survived until impact.

The surface pressure and temperature of Venus are not difficult to survive, the challenge is constructing electronic devices to continue operating in those conditions for long periods of time. There is little point in a series of incrementally lower atmospheric probes, each giving little more data than the previous ones and no surface data. There's no gradual process of incrementally increasing operating temperatures to go through: silicon will simply not work at the surface temperatures, silicon carbide will, and if you have silicon carbide electronics, there's no need to mess around with atmospheric probes.

John Mendenhall
2015-May-26, 12:53 AM
Venera 1 and 2 were flyby probes that failed long before reaching Venus. Venera 3 failed during reentry. Venera 4 operated until it was crushed, and was initially claimed to have reached the surface, but Mariner 5 showed that the surface pressure was far too high. Venera 5 and 6 were successful atmospheric probes, Venera 7 was the first attempt at a lander and it was successful despite a problem with antenna alignment. No Venus lander has ever burned up or been crushed, and one of the Pioneer Venus Multiprobes survived impact and operated for over an hour on the surface despite being designed as an atmospheric probe...the other three all survived until impact.

The surface pressure and temperature of Venus are not difficult to survive, the challenge is constructing electronic devices to continue operating in those conditions for long periods of time. There is little point in a series of incrementally lower atmospheric probes, each giving little more data than the previous ones and no surface data. There's no gradual process of incrementally increasing operating temperatures to go through: silicon will simply not work at the surface temperatures, silicon carbide will, and if you have silicon carbide electronics, there's no need to mess around with atmospheric probes.

Hm, I should have said damaged by the atmosphere. Isn't sulfuric acid a component of the Venerian atmosphere?

cjameshuff
2015-May-26, 02:44 AM
Hm, I should have said damaged by the atmosphere. Isn't sulfuric acid a component of the Venerian atmosphere?

Some of the clouds are almost pure sulfuric acid...which in the absence of water, is mainly damaging to organic materials that it can pull water out of. I've never heard anything about it causing problems for any of the Venus probes, and it would be easy to select materials to avoid problems. It's not a concern for a surface rover, as sulfuric acid decomposes long before it reaches the surface.

lpetrich
2015-May-26, 03:20 AM
You're talking about putting all the hardware required for autonomous operation in the environment most hostile to the function of that equipment, vastly increasing the complexity of the software by requiring it to manage non-volatile storage and perform all the processing tasks required for autonomy, requiring volatile and non-volatile storage for all the extra code, increasing the number of high-temperature transistors with relatively short MTBFs by several orders of magnitude, and relying on some currently-unknown flash process that can function at the surface for which the probe would be one of the first applications, and you're calling a system with a minimalist communications/control system in ROM a failure waiting to happen?
The non-volatile storage would be something like Flash memory. A disk drive would be folly -- it has a lot of moving parts. I'm guessing that the nonvolatile memory would be attached to the computer's main bus and be addressed with

(memory address) = (initial memory address from the bus) + (internal memory address)


You don't need to get down to 100 C, just ~350 C would allow silicon-on-insulator electronics to function, and you don't have to do the entire temperature delta in one stage, but any portion of it done using the heat source directly as a power source would probably be a considerable efficiency gain over using electrical power.
350 C is an upper limit -- one will need a safety margin.

The 6502, for one example, only uses 3510 transistors. Comparison with more modern devices is troublesome because the lowest level modern logic design tools generally work with is gates, and there is no fixed relationship between gate and transistor count, but 8-bit AVRs and 32-bit Cortex M0s are both in the 12k gate range, which would still work out to tens of watts at the stated power consumption...not infeasible to handle.
Looks good. Should be enough for rather simple control tasks.

We can do a fair bit better now with compressed sensing techniques that work with sparse subsets of samples (possibly but not necessarily single pixel samples) scattered across the field of view. In the case of a rover, you can also correlate a sparse set of samples with what you know about the surroundings from previous samples instead of having to stop and reconstruct a new snapshot completely from scratch. The tradeoff is that interpreting the data requires a great deal of processing power.
And likely a lot of memory also. At least in the megabyte range.


And I thought you wanted an autonomous rover...how would you achieve that without being able to store an image of the surroundings?
I concede that it's a design compromise. If one has only a few kilobytes of memory, one can't do much with an image.

cjameshuff
2015-May-26, 12:35 PM
The non-volatile storage would be something like Flash memory. A disk drive would be folly -- it has a lot of moving parts. I'm guessing that the nonvolatile memory would be attached to the computer's main bus and be addressed with

(memory address) = (initial memory address from the bus) + (internal memory address)

I wasn't suggesting a disk drive. You can address it however you like, the issue is that flash memory requires special procedures to erase and write, long term use requires implementation of things like wear leveling and tracking of bad blocks, and safely replacing the programming loaded in it while not bricking the system is a non-trivial process. This all adds to the amount of code that has to be carried, increases the number of places to get something wrong, and increases the total number of transistors, increasing the odds of a critical one failing...and with a new high-temperature SiC technology, MTBF is probably going to be relatively short.

In addition, SiC flash memory is much further away than SiC integrated logic, and like any new flash technology is likely to have issues that only become apparent after a long period of usage, process control issues that only show up for particular batches or particular areas of a wafer, etc. A flash memory is a grid of MOSFET transistors with gate dielectrics carefully designed to reversibly acquire a floating charge with reasonable write voltages, and hold it against leakage for an extended amount of time. A ROM is just a sparsely populated grid of diodes or plain old switching transistors (or for one-time programmable ROMs, a fully populated grid, with some of them either burned open or fused short). You can build a ROM with any kind of transistor, flash will require well-developed MOSFETs.

A minimal system held in ROM that can perform basic rover operations and downlink additional code into RAM to operate instruments could be far smaller and far simpler. Simple enough to be reasonably assured to be free of severe defects, extremely predictable in its behavior, and computers operating on satellites or in the upper atmosphere would be much better equipped for actually controlling it.

John Mendenhall
2015-May-26, 01:01 PM
cjames, your post #17 is the best summary on flash nenory that I have ever read.

Thank you! I'm afraid I have been putting too much faith in FD's. I'm going to backup my backups again.

lpetrich
2015-May-26, 10:29 PM
Non-volatile random-access memory - Wikipedia (https://en.wikipedia.org/wiki/Non-volatile_random-access_memory) discusses various alternatives to flash memory. Nonvolatile memory would only be used for stuff that would not be changed very often, like control code.

cjameshuff
2015-May-26, 11:25 PM
Phase change memory is something I was thinking about. Operation of one type involves heating a chalcogenide alloy to 600 °C, and either rapidly quenching to a glassy high-resistance state or slowly cooling to a crystalline low-resistance state. It might be stable enough at 460 °C to be useful, you'd just need addressing/read-write logic. It might even use less power on Venus, though it'd write more slowly...the main question would probably be whether you could get it to cool fast enough to quench to a glassy state.

jumpjack
2015-May-27, 01:47 PM
the electronics in those spacecraft are carefully insulated from ambient conditions.
Insulation is not enough, it will eventually always "fail" allowing heat to pass; you need active thermal control. But producing heat is much easier than dissipating it.
In a cold environment, all you need is an energy source and to convert that energy into heat.
In a hot environment you have too much energy, and no place where to "throw" it (until somebody invents a "reverse Radioisotope Heat Generator", which would just stay there absorbing energy rather than emitting it :D )

CJSF
2015-May-27, 06:31 PM
Can I just say, here, that I am loving this thread. The inputs and thoughtfulness of requirements are great! I mean, we might never design THE Venus Rover, but it's fun to figure out what might be needed!

CJSF

ravens_cry
2015-Jun-01, 03:47 AM
Idea, off load as much of the computing onto an orbiter or even dirigible, making the rover basically an RC vehicle with the computer elsewhere. That way, as much of the more delicate components are in less extreme, or at least more easily manageable conditions.

Noclevername
2015-Jun-01, 04:25 AM
There's a previous thread on Venus rovers that might prove helpful: http://cosmoquest.org/forum/showthread.php?148337-Can-someone-swot-a-proposed-fictional-vehicle-a-Venus-Rover

publiusr
2015-Jun-06, 06:28 PM
I wonder what can be done with the drinking bird builds on Venus. A small bellows will allow one to rise from Venus--I wonder about Minto wheels as well....