Voyager Space Probe distance counter

[Senormodeler]

Hi,
I have an idea for a project that I'd like to get help with, but I'm not an electronics builder of any kind. This forum seems to be filled with very smart, very creative people, so I thought I'd bring my idea here and ask for guidance.

I'm building a model of the Voyager Space probe (http://www.cybermodeler.com/hobby/kits/has/kit_has_54002.shtml) and I thought it would be cool if the model sat on a digital counter that tells you the current distance that Voyager is from Earth. So it would be a counter that is constantly running. I did a quick search online and I didn't find anything like this already.

I think the constraints of the project are:

1.) what kind of digital display is it? LED, digital, etc. I kind of like the idea of old school red digital numbers. Kind of like this: http://www.shopvoxx.com/product/rca...qfHSYL4T5NFpIlDzZRVAoNpR5ikv2DrIqpThXSFPw_wcB

But i'm not sure how easy or hard that is.

2.) The bigger issue is probably how you make it possible for the number to get larger and larger without running out of space. Obviously, Voyager is far away and getting farther. So how do you convey that through the digital counter without it being limited by how far it gets. Meaning, how do you make it so that this thing can keep running as long as it's powered and not have a cap.

3.) Also, I would want to be able to plug in a starting number that approximates the current distance at the time that the counter is activated so that it is somewhat accurate (giver or take a few hundred miles).

4.) Should it be AC or DC powered? If it's AC, then I don't have to worry about changing batteries, which of course would mean that the counter would be interrupted, and I'd have to reset distance and make sure it's accurate again. Sounds silly, but it would be great if I didn't have to do this.

So, I have no idea how to build this, and really don't have the skills. I'm good with models, and building things in general, but ideally I'd find someone here that is interested in collaborating with me on this. I'd of course pay a nominal fee for parts and labor. This is a gift for my wife, who's a bit of a sci-fi nerd like me.

If interested, please let me know your thoughts or ideas!

Thanks.
J

 

[KrisBlueNZ]

Hello Señor and welcome to Electronics Point

Your project is fairly straightforward. It's actually just a kind of digital clock, but with a much wider display and some extra calculations. You will need:

• A microcontroller of some kind. Nothing powerful or fancy. This will contain the program that reads the time, calculates the distance, and drives the display. You can use a single-chip microcontroller such as a Microchip PIC (see http://www.cursos.ucv.cl/eie48700/referencias/Intro_PIC_C1.pdf / http://www.engineersgarage.com/articles/pic-microcontroller-tutorial / http://www.microchip.com) or an Atmel AVR (http://www.atmel.com/avr). Alternatively you can use a fully made-up board with a development system such as Arduino (http://www.arduino.cc/).
• A real-time clock. This is a small device that keeps track of time. It uses a watch crystal and a lithium coin cell so it continues to update while the circuit has no power. A common choice is the DS1307 which is available as a chip (http://www.digikey.com/product-detail/en/DS1307+/DS1307+-ND/956883 (through-hole version - easy to work with) and http://www.digikey.com/product-detail/en/DS1307Z+T&R/DS1307Z+T&RCT-ND/3647910 (surface-mount version - smaller)) or on a little circuit board (http://www.adafruit.com/products/264). There's an article about using the DS1307 here: http://www.best-microcontroller-projects.com/real-time-clock-ic.html. Some of the PIC and AVR microcontrollers can actually implement real-time clocks without any external ICs (just the crystal and coin cell).
• A display. Seven-segment LED displays (like you describe) are the easiest to drive. You need to decide how many digits you need (which depends on what you're going to display - metres, kilometres, miles, etc - and how long you want it to run for before the display overflows), and other things like colour (personally I prefer green to red, although red is much brighter), height (they're available up to five inches tall, though ones taller than one inch are harder to drive), and brightness (efficiency). They're also available in single digits, pairs, and groups of 3 and 4, but single and paired have the most options. There are literally hundreds of options available. Here's a selection of suitable units HERE, sorted with the brightest ones first.

You will also need some circuitry to drive the display. A display with lots of digits needs to be multiplexed (see https://en.wikipedia.org/wiki/Multiplexed_display / http://maxembedded.com/2013/01/30/seven-segment-multiplexing / http://www3.nd.edu/~lemmon/courses/ee224/web-manual/web-manual/lab10/node4.html / http://www.instructables.com/id/Multiplexing-7-Segment-displays-with-Arduino-and-S/) and this requires more current than a microcontroller can provide. Details depend on how many digits you have and how many I/O pins your microcontroller has.

Have a think about it. There are people here who could help but I doubt anyone will do your hardware and firmware development for a "nominal fee". But if you get involved with micros you could open up a whole new range of opportunities for your modelling work!

 

[kpatz]

For #2, provide enough digits to cover the distance for the foreseeable future. You're going to need to know the current distance and speed to calculate the distances, so do the math and figure out how far it would be in say, 20 years and provide enough digits for that.

If that's more digits than you want, you can always use multiples of a unit, like thousands of miles or kilometers. When it overflows, go to 10,000 mile/kilometer increments, etc.

For #3, it might be easier to program a starting date/time and known initial distance, and then use a real-time clock to keep the current date and time, and calculate the distance based on the time difference between the starting point and the current time. That way, if the power is interrupted, the counter can resume at the correct count when power is restored, especially if there's a battery backup on the RTC.

#4: I'd power it with AC and use a battery-backed real-time clock. In conjunction with #3, as long as the time is correct the distance will be accurate, provided your initial distance and speed constants are accurate. You'll need to provide some way of setting the clock initially. A couple of buttons and appropriate firmware code would do the trick. Perhaps offer an option to display current date/time too by pressing a button in case you need to adjust the clock. Although quartz crystal clocks are fairly accurate, they do drift over time.

You could even go as far as to make the clock self-setting by using a GPS module to pull atomic-accurate time from a GPS satellite.

If you're going to have more than say 8 digits it's going to be hard to multiplex them and keep them bright enough. You may need to multiplex them in groups of 8, or use constant-current serial LED driver ICs, then they can be driven in parallel with a minimum of fuss. A typical IC can drive 16 LEDs (2 digits + decimal point), so you'd need to chain a bunch of them. If you don't need the decimal points you could drive 11 digits with 5 ICs or 18 digits with 8 ICs.

Another benefit of using these ICs is they are sink drivers so you can use a higher supply voltage to drive larger common-anode displays, while still using 5V to power the logic.

 

[Senormodeler]

Wow, Kris! Great ideas, Kpatz, as well!

I found this NASA site with a real-time odometer on it that tells how far it is in 2 different units of measurement -- AU and KM. I think I prefer Kilometers.
http://voyager.jpl.nasa.gov/where/

I did the math, kpatz, and in 37 years, voyager has travelled 19 billion km. So if I were to limit the counter to 99 billion km, then this device would be useful for another 180 or so years, approximately. But if we add one more digit then it won't overflow for another 2000 years!

I REALLY like the idea that this could be a family heirloom that gets passed on and on and that would eventually need the to roll over to 100 billion km. This probably won't happen, but it's fun to think about.

So that means that I need 12 digits total in the display.

And I also love your idea that it would be tied to the accuracy of the time clock. Wikipedia tells us that Voyager 1 is traveling at about 17 km/s (though I'd want to find out the EXACT number so that the counter doesn't roll into error over time), so given the right formula and programming, it seems like it could be a function of the current time.

Kris, I think given the number of digits (12) I'd want the display to be about an inch or less tall. That hopefully would keep the full width of the display to about 8 inch, give or take?

So here's what I'm thinking... I have way too many hobbies to add microprocessing to the list. I need help with this. If either of you (or anyone else reading this) is enthusiastic and interested in this project, send me a private message and tell me, based on further conversations, about how much you think this would cost, and how much you would charge for your time. I realize a "nominal fee" isn't fair. I'd like to pay what's fair. We can also jump on skype and discuss details in order to get closer to a bid.

I'm also open to the idea of a multi-part collaboration. Not sure if that's an efficient way to go or not, but if multiple people want to be in charge of different components (i.e. someone build the case, someone else does the circuitry, etc), that could also be cool. Again, I'm fine doing this or not doing this. It's not a priority.

Anyway, let me know what you guys think.

 

[KrisBlueNZ]

Re the family heirloom idea... it's a lovely thought but not practical. Nowadays, when electronic products are often obsoleted within five years, time scales over 20 years are seldom considered. Program retention in PICs is unspecified for cheap variants, 40 years for mid-range variants, and 100 years for the high-end 8-bit devices. But I doubt that the displays would last more than 20 years. How are you going to set the time? Will USB still be around in 20 years? Pushbuttons can be replaced, but an LCD won't last that long.

Regarding timekeeping, yes I agree with kpatz about using GPS. That will probably still be here in 20 years and it avoids the inevitable drift of a real-time clock. You could have an RTC with a UI to set it, and/or use mains frequency for timing (would need to be set after a power failure) as a backup. What if the Voyager's speed or course changes? You might need to change the conversion. That would require some kind of UI - even a very simple one - to change the base time and the scaling factor. Those would be stored in non-volatile EEPROM memory.

Another approach would be to try to get Voyager's distance from NASA directly via the internet, using a cellular modem or through a data connection such as ADSL or cable. But this would incur ongoing costs from your cellular provider or ISP, and would become useless if the network changed or disappeared, or NASA changed their information service.

Re LED driving, kpatz's suggestion of serially connected driver ICs is probably best. With 12 digits you could use 4:1 multiplexing, but this would require at least 23 I/O pins on the micro, and lots of circuitry. Better to use ICs that are designed for the job. And it would avoid the multiplexing flicker, which you can see if you shake or move your head around quickly while looking at a multiplexed LED display (e.g. a digital alarm clock).

Regarding helping with the project, I can't commit to anything at the moment, but I definitely recommend kpatz based on the excellent advice he has given on this site. I'm sure he's competent to do both the hardware and firmware aspects.

Ideally I think it would be nice for the project to stay "out in the open", i.e. in the forums, but I realise that you may want to monetise the project in future, and if you're paying for the development, you obviously have the right to choose how closed the development will be.

 

[Senormodeler]

Great point, Kris, about the obsolescence of the components. Maybe the thing to do is keep the display to 11 digits and go for midrange variants. Also, I like the idea of keeping it fairly simple with the push buttons to set the time as opposed to GPS. The clock may drift over time, but could be easily corrected (which hopefully would correct the distance, in turn), I would think.

I think it's too much to make it internet capable. Especially in light of the obsolescence that you brought up. I guess I'll have to give up the idea that this thing will be passed on for generations, for 2000 years!

It's a fun idea though.

I love the idea of keeping this project out in the open on this forum. I don't plan on monetizing this as a product to sell. I expect that there will need to be direct communication to work out some of the details, but the documentation of the project on this forum is a great idea. I would also document the building of the model as well.

kpatz, are you interested in talking in more detail about this and possibly taking it on?

 

[Senormodeler]

Oh, and I would assume that Voyager won't change speed or course in the next 40 years. It would only do so if it encountered a gravitational field in that time, and there are no objects that close to us in its path.

 

[BobK]

Oh, and I would assume that Voyager won't change speed or course in the next 40 years. It would only do so if it encountered a gravitational field in that time, and there are no objects that close to us in its path.

Isn't it slowing in the sun's gravitational field? I expect this would be significant over 40 years.

Bob

 

[BobK]

I just went to the NASA page linked above. You need to read the full text under the odometer. The distance from earth actually decreases in some parts of the earth's orbit! It is not simply a clock, the earth, in it's orbit is either moving toward or away from voyager depending on the time of year.

Bob

 

[Senormodeler]

Hey Bob. Yeah I read that. The second odometer below that has the distance that Voyager is from the Sun. That's the number I would use since it's fairly consistent.

Also, I'm not sure if the gravitational field of the sun is significant, but it's definitely something I have to look into!

 

[BobK]

I just calculated it as 0.79 km/sec per year. So yes, it is significant, since the speed right now seems to be about 12km / sec. So 10 years from now, it should be going less than half as fast (if my calculations are correct). Of course the slowdown is slowing down as well as it gets farther from the sun.

Bob

 

[KrisBlueNZ]

There's probably no point going for 11 digits instead of 12. If you really only need 11, you could have one digit below the decimal point.

For something that's essentially just decorative, I don't think you would want to be setting the time too often. That's why I thought GPS synchronisation would be a good idea. It depends how important it is for it to be truly accurate. For example, if it was 10 km out, would you really care? Only you can answer that question. Also it might be workable for firmware to compensate for clock drift... as long as the clock drift was stable.

And in the light of Bob's very helpful comments, it's clear that the conversion between date/time and distance is more than just an offset and multiplication. You'll need to come up with a formula for the conversion. The only input data you have is the date and time.

 

[BobK]

I looked at the data on the NASA site, and it looks like my calculation is off by a factor of 100. In the last year, it has slowed from 17.039 km/sec to 17.032 km/sec, so it is slowing at a rate of 0.007 km / sec / year.

Still significant if you want 12 digit accuracy.

Bob

 

[Senormodeler]

Excellent point, Kris and Bob. Yes, there's a formula that I'll need to figure out, or research. It's probably out there somewhere. I live LA so maybe I could get in touch with JPL and buy somebody lunch. I wonder if the formula is what NASA uses on that link, or if they truly pull that data in real time from JPL.

I think the tolerances on this, for me, should be reasonable. If, for example, it's 10 km off over the next 1 to 5 years, that's fine. Given the vastness of space, we'd still be relatively close to the mark.

I like the idea of one decimal place. Makes it feel even more accurate.

 

[davenn]

Really .... what sort of number do you want to display ? ... do you really need the smallest digits ( the 10's and 100's of km's) spinning over at high speed ?

for a 6 digit counter you could display 10's of billions, millions and 100's of thousands

eg 19,155.5 billion (19 billion, 155.5 million)

just a thought

cheers
Dave

 

[KrisBlueNZ]

Really .... what sort of number do you want to display ? ... do you really need the smallest digits ( the 10's and 100's of km's) spinning over at high speed ?

I think that's the idea... to show that that thing is moving very fast, and to try to convey the magnitude of the distances involved. To show that, even at its huge speed, it has a long way to go.

It could well be that the bottom few digits are just a blur, and there may be no point in actually driving the right numbers - it might be simpler just to alternate through a short pattern that looks convincingly blurred.

 

[davenn]

I think the tolerances on this, for me, should be reasonable. If, for example, it's 10 km off over the next 1 to 5 years, that's fine. Given the vastness of space, we'd still be relatively close to the mark

As of 2013[update], the probe was moving with a relative velocity to the Sun of about 17 km/s.[12] With the velocity the probe is currently maintaining, Voyager 1 is traveling at about 325 million miles per year (520 million kilometres)

you don't even need to worry about 10's or 100's of km of accuracy, its travelling almost 20km every second

I just calculated it as 0.79 km/sec per year. So yes, it is significant, since the speed right now seems to be about 12km / sec. So 10 years from now, it should be going less than half as fast (if my calculations are correct). Of course the slowdown is slowing down as well as it gets farther from the sun.

Bob

you are forgetting that the further out you go, the less the solar gravitational effects will be ( inverse square law = every time the distance is doubled, the field strength is only 1/4 as strong)

EDIT: just read your last sentence properly, Bob, that may be what you were referring to ?
The gravitational effect isn't just reducing a bit, its reducing quite substantially

Dave

 

[Senormodeler]

What Kris said. To have the whole length of numbers there in kilometers, and the last few numbers cycling at good clip is the idea. I don't know if it would be such a blur though. If the probe is traveling at 17 km/s, the counter doesn't have to show all the numbers between 1 and 17. If the counter is updated every 1/4 second then it would cycle from 1 to 4 to 9 to 13 to 17. That would be plenty fast.

If power is an issue, then it could update every half second. If it's not an issue at all, then maybe it would be even more impressive if it updated more frequently, say 1/8 of a second. I don't know what kind of demands that places on the power source or the components.

 

[KrisBlueNZ]

I'd estimate that any digit that's incrementing faster than 20~30 times per second isn't going to be visually distinguishable from a random blur. That would only apply to the rightmost digit, the hundreds of metres digit (first digit right of the decimal point). That could be driven quickly with a simple predefined sequence designed to look like the blur of a number that's updating too fast to follow by eye.

The next digit, the kilometres digit, would need to be updated 17 times per second, which should be reasonable for a small PIC. It depends how complicated the formula is. One of the fun things for kpatz to figure out, if he gets involved

Personally I think it would look best like that, rather than just updating the whole display at a certain slow rate, like snapshots in time.

 

[Senormodeler]

I like it!

I sent an email to someone at cal tech in the hopes of tracking down a mathematical equation that will account exactly for the speed and deceleration of Voyager 1. We'll see if that bears fruit.