Brainstorming an Open Bicycle Computer

davedarkodavedarko wrote 07/07/2014 at 12:59 • 1 min read • Like

So I took my bicycle out to my parents, about an hour ride and I noticed that it's quite embarrassing as a tinkerer not having my own arduino based bycicle computer. I've looked around the web and could not find something satisfying to start with, only a youtube video 

The software seems to be a good starting point, but it's not complete and the hardware lacks at least a case and a pcb. So right now i'm collecting things one could do with an arduino based bycicle computer. Which information is relevant for me, what should it at least do, what could be nice to have etc.



Kenji Larsen wrote 11/07/2014 at 01:11 point
You might consider an accelerometer, where you discard data obtained below a certain wheel speed (to eliminate non-riding times), to map against GPS, roughness of road. It's essentially a texture map - 1mm characteristic distance is a "smooth ride", 10mm is potentially "bumpy", 100mm is wheel damaging.... as you ride you can fill in an XY map with a d^2Z/dt^2 response surface. Then later you can optimize for smoothest ride, or shortest ride, with bumpiness within certain limits, as you like. At minimum it might save you some future rims!

  Are you sure? yes | no

davedarko wrote 11/07/2014 at 09:17 point
Sounds like a good feature, but I still have to build a working prototype. I got distracted by other projects and sometimes they have to rest for a while.

  Are you sure? yes | no

Kenji Larsen wrote 11/07/2014 at 11:12 point
I know this so well myself.... Maybe we could put our heads together on a time machine. Not to travel in time - one that just manufactures more!

  Are you sure? yes | no

davedarko wrote 11/07/2014 at 11:53 point
that may end up in just another project that rather steals time :D Let us make some money instead, so we can tell our AI's to build our ideas fast and add a little hot rod red to it. We need AI think tanks and money. And a robot farm.

  Are you sure? yes | no

Kenji Larsen wrote 11/08/2014 at 04:30 point
I like your style! And robot farms. And we can just use the normal time-compression methods (coffee)...

  Are you sure? yes | no

sq6gtt wrote 11/02/2014 at 11:22 point
It would be nice to have calculated current torque based on cadence and suggest gear change.

  Are you sure? yes | no

davedarko wrote 11/05/2014 at 20:39 point
I guess speed vs. Cadence could get you something to work with?

  Are you sure? yes | no

mr.jb wrote 07/17/2014 at 11:51 point
check this out

consider attiny861 ... in case you need differential gain ( I love this chip )
My next choice would be an xmega

charging LiFePO4, I'm sure you could use the same method I use with a smaller step-down and a reverse blocking diode ( and move the sense resistor to the other side of the diode )

Consider Nokia 5110 ..draws almost no current ( less than 1 mA )

GPS suggestion.....could be supplied with the buck boost I use ...;-)

project box suggestion....use a clear case lid like in my project but replace the box ( using a 3d printer + opscad )


  Are you sure? yes | no

davedarko wrote 07/18/2014 at 07:11 point
thanks a lot for your input and willingness to share :)

where is the difference between the attiny861 and the atmega328 used in the arduino? I like the concept of getting anyone started with a small setup of things already in the hands of a tinkerer so using the atmega328 on the selfdesigned board makes sense to me. But nothing is prohibiting me to use a "multicore" system, one for charging and one for controlling.

I guess I would have to use a rectifier and some magic capacitors first and feed that to a step-down thingy? I'm not quite sure how many cells I will/want to use and what voltage I want to have since 3.3v or 5v are an option for an atmega setup and batteries.

The display choice comes down to the nokia display with a backlight function or an OLED display. Since this is open and I want to play around with the options I will look into both but the OLED looks quite cool and sometimes you have to go with cool.

GPS was not in my previous considerations but this little module looks quite promising. My hacking pocket money is killed for this month though.

The project box will be an open-source 3D printable case, which anyone can replace with a box of his choice (maybe a big box of tictacs...) should be water resistant at least. I'm going for a stationary setup without a display for the bike, so it will still log all the data in case you forgot your display/button thingy. This will be in a pipe (good place for batteries) with terminal screws integrated, but is still a raw concept.

  Are you sure? yes | no

Stefan-Xp wrote 07/14/2014 at 21:27 point
You are right, Its embarrassing to have no arduino device at the bike :-)
But it would be challenging to create a device at reasonable costs (cheaper than a standard tacho).
But i also like the idea of creating things by your own ;-)
I thought about a bike POV device which tells the actual speed :-D but not ready for use at this time.
Have you thought about energy harvesting of kinetic energy?

Best regards, gruß Stefan

  Are you sure? yes | no

davedarko wrote 07/15/2014 at 00:25 point
Yeah, I bought a 5 Euro bicycle computer this week, it was really a cheap thing but its comparable to an arduino with one sensor and a display. Add two sensors to an arduino and you get a computer with cadence, the frequency of your pedals - now equivalent to a price range of 50 Euros. I could not find one that stores the data of your rides in the 5min I looked, but I guess that's up to the 300Euro GPS enabled devices. But what about lights and distance sensors? I've seen singular projects for those tasks but I want something more integrated.

Your POV could benefit from this project, I guess :)

using a decent dynamo is planned, I want to buy a hub dynamo (nabendynamo) and integrate it. Also some batteries that are capable to withhold the german temperatures better then the Berliner S-Bahn.

  Are you sure? yes | no

davedarko wrote 07/10/2014 at 10:36 point
for monitoring the power of a lipo cell for protection of undervoltage etc.

  Are you sure? yes | no

Ken Meyer wrote 07/07/2014 at 16:34 point
GPS absolutely drains the battery, but the utility is massive! I'd still trust a speedometer on the wheel far more than GPS, but tracking cadence and gear on different slopes would be difficult without GPS to later show you what slope you were on!

Connecting to a phone would be the ultimate step. Then you wouldn't have to rely on the phone's accelerometer (awkwardly strapped to one leg) to get cadence, and you could always have accurate speed information even when GPS cuts out in a tunnel or under heavy trees.

I wonder what it would take to get a good measurement of wind speed and direction, probably at the front of the bike. As long as we're adding sensors, might as well throw a full weather station on there!

It'd just be interesting to graph stuff like heart rate and cadence against slope and headwind.

I haven't raced for over a decade, but I'd still love to have my Croll fitted with all sorts of electronics!

  Are you sure? yes | no

davedarko wrote 07/08/2014 at 19:13 point
Measuring wind speeds on moving elements sounds tricky, since you have to calvulate the cycling speed out and stuff. Lets leave that open for the refinements and add ons. Cadence on the other hand seems doable. Would it be possible to use an accelerometer for the slopes? Well i ordered an iic display to start with the speed and distance anyways :) thanks for the input and discussion!

  Are you sure? yes | no

Ken Meyer wrote 07/08/2014 at 23:00 point
I'm not looking for a actual wind speed, I actually think it would be more valuable to have a rough measure of drag from airspeed

  Are you sure? yes | no

davedarko wrote 07/10/2014 at 10:43 point
Yesterday I had wind from the front and thought that it would really be cool to see in the stats how much wind there was on that day. I have no intent to ride races against anyone else but me but I'm also interested in how much the wind slows me down. I want to log all the data and be able to make some fancy graphs with it, all time related, stored on an sd card.

  Are you sure? yes | no

Ken Meyer wrote 07/07/2014 at 14:57 point
My cell phone does a remarkable job of tracking speed via GPS, but for anybody who's serious about racing, cadence (RPM of pedals) and gear shifts are pretty important.

I'd add a gear-shift sensor and a GPS unit and log everything for later analysis to an SD card.

  Are you sure? yes | no

davedarko wrote 07/07/2014 at 15:30 point
Using the cell phones GPS may works great and I guess there are good apps but it also drowns the battery a lot. For an approximate speed and distance calculation the old school way should do fine. I may upgrade it to be connectable to a cell phone anyways. Think of it as a board computer, I don't need to know where I'm going, but how ;)

  Are you sure? yes | no