Even though I’ve advertised this site as an ‘online engineering notebook’, there hasn’t been much engineering or even design included so far. I started this site after most of the design work for the teardrop was already done, so it’s mostly been a build log. I will try to go back and fill in some of the design work in the future (I have to clean up the Fusion 360 models of the teardrop so I can publish them), but it’s kind of like writing the documentation for a completed software project – I’m way more interested in doing the next thing than in documenting the last one. There still are some steampunk theme items I want to add to the teardrop – this post and the following ‘Steampunk features’ posts will document the design, engineering, and build of these items.
In the last post I talked about not liking gratuitous gears – if they are there I want them to do something, even if the result is kind of pointless (like a mechanical governor spinning away, governing nothing, but looking functional…). The first ‘steampunk feature’ item is a display panel for current position (latitude, longitude, elevation) and total miles the trailer has traveled.
Of course, there are easier ways to do get your current position – take out your phone and use the built in GPS for coordinates or map location.
An interesting (to me anyway) aside to the problem of determining position is the development of accurate, seaworthy clocks for determining longitude – read ‘Longitude: The True Story of a Lone Genius Who Solved the Greatest Scientific Problem of His Time’ by Dava Sovel for more information. Imagine going to sea in a wooden ship, powered by the wind, with only a rough idea how far east or west you were. And your maps were drawn by someone with probably less of an idea where they were than 
you have! Getting a reliable, accurate, sea clock was a huge advantage for the first country to get it.
I’m thinking something more like what you’d see in the Nautilus, a brass panel with physical numbers, like this…
Captain Nemo would probably have had a crew member taking a reading with a sextant, doing the calculations by hand, and manually updating the display (or calling it up to another crew member using a voice tube). I’ll use something a bit more modern to drive values – a GPS unit attached to an Arduino or some other microcontroller.
First, I need to know the display parameters. I doubt we’ll ever drive the teardrop out of North America, so the limits on lat/long are about 15 to 95 North (Central America to just short of the Arctic Circle) and 50 to 180 West (Newfoundland to Western Alaska). Now I need to figure out how many digits I need.
So 2 integer digits N/S, 3 E/W. Precision? Captain Nemo would have used degrees/minutes/seconds; modern displays would probably use decimal degrees. According to Google, 1 degree of latitude is around 69 miles (another fact I didn’t know – a nautical mile is a minute of latitude, around 1.5 miles). Therefore for longitude in DMS
- 3 integer degree digits – 69 miles/degree
- 2 integer minutes digits (00-59) – 69/60 = 1.5 miles/minute
- 2 integer seconds digits (00-59) – 1.5/60 = 0.02 miles per second.
- 1 decimal place accuracy for seconds – 0.1 second is 0.02/10 = 0.002 miles miles or around 100 ft, around 15 ft – close enough for GPS.
So for 8 digits for DMS display. Distance per degree latitude varies, but for consistency I’d use the same precision. I guess we’re not really saving any digits…
Decimal display – the same 3 digits for degrees.
- 0.1000 degree is 6.9 miles
- 0.0100 is 0.69 miles
- 0.0010 is .069 miles (360 ft)
- 0.0001 is .0069 miles (36 ft) – close enough.
7 digits for decimal display.
Distance traveled? We can remember how many times we go over 9999.9 miles.
5 digits.
Elevation? 0 to 10,000 feet should do it, no reason for a fraction since the GPS isn’t that precise.
5 digits. That gets us into the mid-stratosphere (99999 ft is almost 19 miles). I think we’re good.
Maximum of 26 Total digits for lat/long (8 each), elevation (5), and miles traveled (5).
Would Captain Nemo have been using meters? Probably; 20,000 Leagues Under the Sea was written after the French Revolution and the introduction of the metric system. But ‘league? (actually it’s a real thing – 3 nautical miles, which we now know is 1 minute of latitude). I’ll stick with miles for now – it can always be changed in software, right?
To get the look I’m after, I’ve narrowed the display options down to:
Mechanical Counter – like an old car odometer. They can run both ways, that’s just disabled for old cars for obvious reasons…
Advantages:
- Most retro looking, this could have been done in the Victorian era. OK, not
this one, but imagine it in copper/brass, with an old style typeface… - Reduced digits for DMS version – the minutes and seconds can each be on a single digit. Not every number, but enough to get the idea across, like 0-5 numbers with tick marks between
Disadvantages
- Lots of gears – but I have access to 3D printers
- Mechanical reset to 0 – the microprocessor could get out of sync with the actual position of the digits when the power is turned off, etc.
Nixie Tubes – 1950’s era electronic displays. A bit beyond the steampunk era, but still has the look of old tech.
Advantages:
- I’ve always liked this look.
- No moving parts
Disadvantages
- Requires 2 digits for minutes/seconds
- High voltage – these things require 70 VDC to glow. Stepping our 12V up to that isn’t impossible, but I’d rather not deal with it unless I have to.
- Cost – the Russians aren’t making these tubes any more, so what we have is what we have. Prices are going up.
- They’re old – last ones were from the late 60s or early 70s. They work by heating up a thin wire. They’re glass tubes. Reliability is a factor.
‘Lixie’ Tubes – similar to Nixie tubes, but using edge lit plastic etched with the value to display.
Advantages:
- No moving parts
- LED powered, so 5VDC is good
- Close to the Nixie look
- Durable
- I can make the digits on the laser engraver
Disadvantages
- Requires 2 digits for minutes/seconds
- Seems to require at least 2 LEDs/digit – that’s 52 LEDs.
- Circuit complexity. Even using Neopixels (3 color LED with built in drivers and latched addressing) there are a lot of wires and connections. 52 LEDs would require just over 200 solder joints for the LED array.
Nixie is out – really don’t want to deal with the voltage requirements and fragility.Maybe I’ll build a Nixie clock someday.
I did prototype a 1 LED/digit Lixie digit – it worked OK, but I think the digits need to be a little bigger and have 2 LEDs to really light up well enough to be seen in daylight.
I also built a test 2 digit mechanical counter. It looks pretty good, minimal soldering,
low power requirements – I think this is where I’ll start off anyway, pending running into something ugly when I start the real design work.
Costs?
Arduino and GPS module are constant for all versions – $20 for Arduino, $50 for a GPS module. Add some for wires, circuit board (I’ll cut my own from copper clad sheet), maybe a power converter. I’ll know more once I start designing the hardware.
Cost per digit for the Lixie version is in the $0.50 to $0.75 range (LEDs and acrylic).
For the mechanical counter, cost per digit is around $4 if I get them printed at The Village Workshop. I might be able to squeeze a little more material out to save some cost, but not much – I already just print the gears, the middle part under the number strip is a wood dowel. I might be able to print them for free at work – we run a lot of parts and these would fit in with other stuff being printed. I’ll also need 4 small stepper motors and drivers for each display to power the rotation.
Next post (unless I get back to the ‘finishing off the trailer’ post) will be designing the control hardware and software and getting the mechanical design finalized.
