100 Year Long Tone?


I'm an artist currently working on a piece that would require the generation of a 100 year long tone. To do this commercially on a sound generating program, seemed to have many problems. So I was looking to see if I could program Arduino to create 100 year long tone. I tried to do it with delay, but it wouldn't allow the number of milliseconds in 100 years, which is, correct me if I'm wrong, 3,155,760,000,000 Any help or suggestions would be very much appreciated. Thanks.


I don't get it on the "100 year long tone"? Is the tone sound for 100 year? or to sound the tone once in 100 year?

Break it down into shorter times., with loops if necessary. I'd love to help you with the testing, but I don't expect to be around

I dare say many of the dozen or so of us who have read this so far are thinking - only an Artist would be asking a question like this. XD

I am going to take a punt that you want this tone to last for precisely 100 years, and anticipate a party at the other end watching the clock and waiting for it to end. Aside from the obvious questions, such as whether there are going to be people around in that climatic wasteland, taking an interest in your particular artwork though it certainly could have some novelty - even then - as a time capsule, the next more obvious questions are - what happens if the power goes off, the batteries go flat or - a component fails? Reminds me of various "relics" that have no doubt, needed "touching up" from time to time. XD

A RTC (Real Time Clock) accurate over 100 years is not going to fit with your Arduino - presuming you could afford it. You could use GPS which is presently our most accessible, extremely accurate and reliable time reference but you know, I very much doubt that will still be present in its current or compatible form, in 100 years. I don't know what else will be, but history suggests that something different will be.

So I fear I am going to be at least as much of a "wet blanket" as everyone else.

The tone's duration would be 100 years long. I'm more focused about the potentially of the tone lasting 100 years rather than it definitely succeeding to last 100 years if you catch my drift. This project is an example of something that could be more permanent and reliable. I was looking to do it as whole as possible to get the sense of the duration and weight to the tone. In the back of my mind I half knew that I might have to resort to looping, in which case I would do so. Thanks for the help so far, I hope this answers some questions you might be wondering. :)

I think flash data retention is only just guaranteed to 100 years (don’t ask me how they tested that!), and even then probably in a temperature-controlled environment.

philip13: . I'm more focused about the potentially of the tone lasting 100 years rather than it definitely succeeding to last 100 years

In that case you don't have any timing problems, you just design it to run for ever, secure in the knowledge that it probably won't. All you have to do is come up with a suitable power supply and and a sound-proof enclosure. The rest is easy, although I imagine some multiple redundancy would be in order. You can then spend the rest of your pondering on the irony that, the closer you get to success, the less likely you are to be around to pick up the plaudits.

A mere century?

I don't want it to last forever though. I want the duration to be 100 years. It's the concept, and the implications of it that I'm after, rather than any reality. So as close as I can get to this the better. The piece itself will only be on for the duration of the exhibition. So its just the creation of the tone I'm looking for?

Regarding precision, we haven't heard the actual requirement. RTCs with temperature-compensated crystal oscillators are common, reasonably priced, and accurate to a few parts per million. That's a few hours over 100 years' time. Is that good enough? The accuracy requirement needs to be stated in numerical terms, not as "precisely 100 years", "as close as I can get" etc.

However, most RTCs have algorithms that are not advertised as being good for 100 years. For example, the datasheet for the Maxim Integrated DS3231 (Accuracy ±2ppm from 0°C to +40°C) says its leap year compensation is only valid until the year 2100. Still, it would be fairly straightforward to use it simply as a time base and the microcontroller (Arduino) could have an improved algorithm.

But by far the biggest challenge here is reliability of the electronics and of the power source. Designing for such an extended period is extremely rare. Read about the Long Now clock (AWOL's link) and some of the considerations for that design.

Here is an example of producing long tones, an organ performance scheduled to last 639 years. I'm not sure what steps have been taken to ensure reliability.

philip13: So its just the creation of the tone I'm looking for?

No, you need to come up with some serious-looking hardware, and I submit you also need to come up with a portfolio of plausible lies i.e. enough technobabble to ward off those who loudly think it won't last the distance.

Bear in mind that the transducer (converts the electrical signal into sound) will almost certainly have moving parts, that will wear and degrade. Maybe consider a plasma tweeter.

A lot more specifications need to be given. Off the top of my head I would ask questions like:

How accurate must this hundred year interval be, 10 milliseconds, one second, one hour?

Will the tone need to be audible to people the whole time. If amplified sound to speaker(s) is required that will probably set the minimum energy requirements for the project. The power source longevity and reliability is probably the first design task you have to solve, as if that fails before 100 years then nothing else in the project will matter.

That link that AWOL posted ( http://longnow.org/clock/ ) gives a good description of what is involved to keep a device functioning for very long periods of time.


philip13: The tone's duration would be 100 years long. I'm more focused about the potentially of the tone lasting 100 years rather than it definitely succeeding to last 100 years if you catch my drift.

OK it is simple. You have a continuous tone and you tell everyone that it will turn off in exactly 100 years. In other words you lie, the concept however is got across and nobody had to be nailed to anything.

I suggest that your goal should be to create a count-down display capable of showing a hundred years with resolution of (say) a tenth of a second, and make it count down in real time. In practice, only days/hours/months/seconds/tenths would actually need to be dynamic since it won't be left running for months/years. Perhaps rather than a continuous sound you could have your device put out a 1Hz beep and LED flash, which seems to be the standard "I'm an electronic countdown" display according to TV shows.

This part is certainly possible with Arduino, and all you'd need would be a suitable number of seven-segment displays.

If you aren't committed to using a microcontroller-based approach then it would be easy to display the countdown on the screen of any PC, laptop, tablet, smartphone etc.

Somebody's going to smash the speaker LONG before 100 years is up. XD

If you set it to a frequency of 1 / 3,155,760,000 Hz, all it has to do on the expo is toggle a pin once.

Years are defined by days.

Days are defined by daylight and nightdark.

A light detector with a VERY long time constant, coupled to static CMOS counters to count to only 36525 could probably run on capacitor power storage and a few redundant silicon solar cells. That's the time port.

The TONE? Frequency? Signal level? Acoustic output or other output??


Have a little something like this running, with the date displayed continuously.
Or run it backwards, "countdown to tone stopping: " and display the appropriate data.
And tone can just be a PWM output.

Or add some portion of this sketch in to change the tone every so often, or just play the notes in random chords…

piano13keyBurstFix.ino (11.2 KB)

I agree with Mike, the idea of actually getting this to work and then proving it is total BS, so just put something in a box and tell everyone it will work. The arty guys will believe you and the techos won't, so don't let anybody technical near the installation.

Job done and move on to something achievable.