And the measuring and control circuits work with tubes.
So does the concrete itself, if its information about the concrete you're trying to record. You don't need access to the inside of a concrete pillar to obtain information about the condition of a concrete pillar from the outside. What data are you hoping to record? Moisture level over time or something like that?
That's advice does not help me... i don't think it could be unfeasable to store digital informations for so long time.... The contrary we need and we are implementing methods to storage data for thausen of years
That's correct but we want surround the device in black epoxy resin in order to improve the reliability. I just miss the right components (non commercial). I belive that is not unfeasable build circuits that stays for centuries withouth fails
i would have access to these technologies
That's is a constraint of the experiment I do not want use external devices. The research it could be useful in the future to build more durable circuits and waste less electric parts that sometimes are unfeasable like satellites in orbits
Instead that replace the part we just want have more durable electronics
There are someone that have practical solutions or even some contacts hints that can help us to acheive our goal? I don't need advices liike "use paper"
Do you think humans be around in 200 years, and have the technology and interest to read your data?
Why should they care?
The research it could be useful in the future to build more durable circuits and waste less electric parts
If so, publish your research, and if anyone else is interested, they can store the publication for 200 years.
Write the data on a gramophone record, this is a proven technique.
to your question, I don't think the ESP32 is the answer to your need, I am not aware of technical solutions with a 200 year lifespan (and I doubt you'll find computers able to connect to BT in 200 years... we both won't be there to see it).
and with the current construction speed and rebar corrosion can your concrete actually last 200 years?
I'm sure I'm not the only one curious about what you are trying to measure (and also curious to know the context for 100 or 200 years)
Just make sure it does not use carbon for the nice black color.
@ Hardware: When you can feed a cable to wherever your device should be, then it can be replaced at any time - just do a bit research on how that's done in industry. And there's no need for wireless - that won't work for long anyway.
Your BT connection will break sooner than later, expect 10 years to be the limit. The data sink will not be accessible/usable in 20 years - and most likely be forgottten by that time anaway. Use paper stripes to print/punch live data as in the old days, these work even today (I have some programs laying around on cardboard from the early 70s). Printouts last long, my oldest listing is still readable (inkjet, early 90s). Can't say that for the other magnetic tapes or optical media: either the reader died long ago or the media degrated. Materials age, no matter what you do. When you got you stuff powerd up better don't turn it off ever again, it will most likely not come back to live. Spare electronics usually die the moment you need them most (that's why one spare is never enough).
So the purpose of the concrete is just to encase the experiment?
What data, exactly do you need? Any arbitrary data? If it's merely to see whether an Arduino will still work after 100 years of use, say, then your recording device will operate outside of the device in concrete, which you seem to have in your plan at least in theory.
In that case, take your pick - there's no evidence to show that any one microcontroller chip will be beneficial over any other on that time scale.
Then, just run the built in blink sketch (don't reprogram the Arduino in any way, if it doesn't work in 200 years then someone might go "Oh, well, he uploaded a sketch and maybe if he hadn't it would still be working.") , wire out pin 13 and ground (or the default LED pin for board) to someplace else with a note attached that reads "Do not touch until 2223. Measure the DC voltage here. It should read 5 volts."
It wasn't that long ago in academia we were storing everything on zip drives so we could ensure that we'd have it for the long haul. We all see how that worked out.
1 Like
You must go a lot further in protecting the device, as concrete will have the same moisture content as the average humidity of the surrounding air. ANY metal in your device will oxidize in the presence of water vapor. Best to store your device in a sealed nitrogen atmosphere.
get your software and hardware working.
Have it run unattended for six months, them you might start thinking about encapsulation and permanence.
I guarantee you’ll make changes leading to further six month test cycles.
You’ve got plenty of time to work through it.
2 Likes
9-level MYLAR punch tape.
In the 21st century, use of punched tape would be very rare, possibly in obsolete military systems or by some hobbyists
Obsolete? HaahahhHAHhahah!
2 Likes
If you can keep your solenoids and motors running for 100 years without any lube or maintenance - let me know,
How long are your blank tape spools ?
Instead of use some sort of irony you read carefully my post from the begin I specified that we need a programmable memory that can store executable code. I'm aware that the executable code (that is a type of data) disapear from the flash leaving the device unusable.
I want higlight that the goal is STORE EXECUTABLE CODE for long time and this implies to store data for long time bacause executable code is a subtype of data.
and how do you plan to re-flash that firmware ? if the executable code is gone, chances are that the bootloader is also gone or damaged...
to your question, all EEPROMs (Flash ROM), and EPROMs chips have a finite data retention time. Typically listed for 10-15 years. A device using that technology for firmware storage will just stop working when it is old enough even if all other circuits are still good. In reality, if kept at normal temperature and out of UV and other environmental issues, it can be much longer but no one will guarantee that at consumer price.
Mask ROMs (the type used in old calculators) do not have such a problem because the firmware is etched into the chip at production time.
So may be you need a custom design knowing how to read its firmware from Mask ROMs...
but there are so many other failure reasons over 200 years that are likely to arise and that you need to solve for - think solder joints crack, capacitors fail due to unstable power, corrosion breaks, shorts traces... Not to mention that if your code does run all the time, the SRAM will also degrade.
I'm curious (honest question) to know what you will be measuring and if you have a 200 year research grant to know that people will work about this over time.
If you think that all the suggestions freely given here are just ironic, then you should find another forum where the responses suit you better than insulting the members in this way.
to be fair the discussion around punched tape , Golden Records is not matching the original request where @gchini is willing to
➜ it's something live, not hardcoded once and forgotten (like a time capsule thingy).
so if you have a strong belief on this with consume products, then enlighten us.
I agree with that. What's the real goal? The idea of embedding only the sensor has not been commented upon. That seemed a smart idea