Skip to content

Zwiftifying My DeskCycle¤

The Device¤

My best buy in this century:

Bought it since my knee begins to hurt when I don't move it, sitting in front of the monitor all day long - and was willing to pay the price to fix that, i.e. for sure less focus for work when I cycle...

But then I got it and I realized:

That thing does not at all pull focus - but to the contrary.

Heart rate slightly up, body not in this mode of total lethargy, with only the brain working, sooner or later backfiring...

Deskcycles in School

My sister, who is a teacher, told me that the positive effect on brain activity is well understood and that they introduced them within their school: In the last row already 5 of these, where each pupil may go anytime, during lessons(!) and start cycling away tensions or tiredness.

First Mechanical Mods¤

Appearance, Height¤

Ripping off the plastic insult around it, plus adjusting the height (like to have me legs more horizontal when cycling - not so much need for click pedals then, with the inactive foot's weight not on the pedal):

Desktop Shifting, While Cycling¤

Next step was to reposition the shifter, which is unreachable from the desk down there and wire it up to the desk - attaching an old Shimano Ultegra Shifter:

Material:

  • Old shifting unit for the rear derailleur (any amount of gears is fine)
  • Long gear or break cable
  • Luster terminal (from normal household inventory) to hold the cable - the white thing in the picture (you need to have open end to get the cable through a normal bicycle shifter - the original cable did have 2 closed ends, not feasible here)

Shifting Gears - Basic Principle

The cable, when shortened by the shifter, pulls the magnet in the picture back from the big rotating metal wheel, reducing the power required to turn it.

The spring is actually not necessary, since the magnet is attracted by the wheel.

Ok, can shift like a pro now, from the desk, w/o interrupting cycling. Nais! 😎

The Problem¤

I want to know what I achieved!

With History and stats.

Nice to have - comparison with colleagues?

Zwift, basically...

Zwift¤

Do they offer desk cycle support?

No:

Their default: Mount normal bikes on a stand, either 'direct drive' style, where the rear wheel is removed and the chain attached to a measurement unit - or a roll trainer, which measures trivially the speed of the wheel.

https://www.saris.com/post/blog-smart-trainer-calibration-hot-and-cold

Indoor cycles, e.g. Schwinn

https://www.indoorcycling.org/schwinn-mpower-echelon-ii-konsole-mit-watt-sensor.html

"Comment: Guessing Watts by distance of magnet..."

Ob drahtlos oder verkabelt, wäre an dieser Stelle m.E. egal, denn die Konsole wird ja an/mit einem stationären Gerät betrieben. Dennoch finde das Echelon II maximal mittelmässig. Die Nachvollziehbarkeit der Watt-Anzeige ist nicht gegeben. Hier wird anhand der Einstellung der Magnetbremse die Leistung eher geschätzt als real gemessen. Die Watt-Leistung kann nicht an das Körpergewicht angepasst werden, somit ist die berechnete Geschwindigkeit i.d.R. deutlich zu hoch. Vermutlich wird die Geschwindigkeit in der Ebene berechnet und die eingestellte Bremswirkung bezieht sich ehe auf den Gang als auf die Steigung der Strecke. Die Daten können ausschließlich als csv-Daten auf einen USB-Stick gespeichert werden. Um diese Daten z.B. nach Strava zu übernehmen, müssen diese in tcx umgewandelt werden. Die Konsole ist besser als Nichts, aber für den Preis erheblich zu teuer. Einer einseitigen Watt-Mess-Kurbel, gekoppelt an eine andere Konsole, würde ich deutlich datentechnisch mehr vertrauen. Zudem schade, dass Echelon keine Bluetooth Schnittstelle besitzt. Ich denke, dies ist ein Auslaufprodukt, für dass es keinen - besseren - Nachfolger geben wird.

The Vision: Ideal Solution¤

  • We want to keep manual shifting. Can't see the benefit of shifting (only) through an app, when I can just grab the shifter and click.

  • Computer knows cadence (RPM) AND current resistance - i.e. can calculate power.

  • Computer can ALSO control resistance - but automatically.

Use case:

  • Imagine watching a scenic ride YT vid on mpv, where you can control the playback speed.
  • Now get the gradient data from some cycle routes database, add that to the vid - and the computer could increase the resistance when going uphill automatically - plus adapt the playback speed according to the cadence AND the current gear

Means, we need to, via Software:

  • A) Know the manually selected gear
  • B) Control the resistance, additionally to the selected gear

Calculated speed would then depend on cadence and gear but on total power, which has to overcome the computer generated resistance as well...

Out of the Box Sensor: RPM only¤

They deliver a normal magnetic tacho, to measure RPMs, i.e. cadence - and only that:

Apart from the problem that it cannot keep history: The problem is that those RPMs have not much to do with the power needed, to produce them.

That power depends on the "gear" I select via the selection wheel. Which changes the distance of a magnet to the rotating wheel. There is no mechanical Translation of RPMs into higher speeds of any wheel inside:

Even the market leader can NOT do it better, also sensing only RPMs - so you would have to go to their website(!) after each session, or use their app, entering the selection wheel position, to finally get to your power spent and calories burned:

Reminder: We are in 2022....

Because it's expensive to measure power. Measuring power could easily add over $200 to the cost of the bike.

Yes, power sensors cost well over 200 (see below), if not over 1000 if you want 2 pedals.

But those are made for real cycles, i.e. with tons of constraints, which do not hold for desk cycles:

  • ultralight
  • ultra low battery
  • cannot simply use gear info plus rpm (uphill, downhill out there)
  • wireless, apps, and and and...

So no - it cannot be that measuring power without those restrictions must be that expensive!

Ok - how to do it.

A first attempt with our voltmeter, simply measuring the volts on the wheel1 against the magnet did show that the voltage is simply proportional to the RPMs - but not the power I need to get to those RPMS.

We have to dig a bit deeper.

Eddy Current Brake¤

German: wirbelstrombremse

https://studyflix.de/elektrotechnik/wirbelstrombremse-1644

    b^2 * B^2 * v
F = -------------
         R

With:

  • F: Force which acts on the metal wheel, braking it
  • b: Area of the wheel through which the magnetic field goes
  • B: Field strength of magnetic field
  • v: Speed of the wheel
  • R: Electrical resistance of wheel (Ohm)

Facts:

  • R is constant.
  • v we know from the RPM sensor

B and prbly also b depend on the distance of the magnet, i.e. the current gear.

We see three possibilities:

A) Buy a power meter

B) Know current gear selected - and find the transformation to B and b through measurements

C) Keep b constant by NOT changing B mechanically - but use an electro magnet.

D) Generate the Watts not into heat but into (measurable) electricity

We discuss separately...

  1. "wheel" we denote the heavy metal wheel where the braking force is acting upon, through the magnet. ↩

Back to top
Copyright © Gunther Klessinger
Made with Material for MkDocs