Flywheel BPM Timing Test - mmxt-flywheeltest

1.0 Test Design

1.0.5 Safety Concerns
A Flywheel stores brutal amounts of force. When things go wrong around a fast spinning heavy flywheel, serious injuries can be inflicted. If you are uncertain about how to mitigate these safety risks, do not perform this test.

1.1 Research Question: 
Primary question: How consistent RPMs (Revolutions Per Minute) can we achieve using a hand-crank and a flywheel?
Extra question, if applicable: How consistent RPMs can we achieve using a treadle pedal in stead of the crank?
1.2 Suggested Test Setup: 
  1. Use a stable Bench as a base.

  2. 2 Axles, mounted on separate brackets, allowing for easy distance adjustment/belt tensioning. 

  3. Flywheel, Large and heavy, Thick sandwiched MDF can be a cheap option. Adding scrap metal pieces could increase weight and balance. 

  4. Flywheel Pulley.

  5. Crank Pulley and Belt. I recommend using variants of Flat belts since it is easier to make cheap DIY pulleys than toothed timing belts. Feel free to use any belts and pulleys you have access to.

  6. Beater, some form of stick sticking out, mounted in the same direction as the crank handle.

  7. Target, the beater is hitting the target, creating a clicking sound on each revolution. The beater moves faster the further out from the center of the axle it reaches. Faster beater results in a more exact and louder clicking sound.

  8. Crank, it is essential to design this with safety in mind, a large solid circle inside the crank will protect your hand the best.

  9. Download a free digital metronome, or run a click from your music software

  10. Record the sound of the beater and the metronome. Try different tempos. Crank the prototype and try to match the beater click with the metronome click. 

  11. More straightforward but much more Dangerous Setup. Using only one side of the bench and putting all components on that side is much simpler to build but less safe. You will expose yourself to strong, spinning, hazardous objects, and the power from this Flywheel will have serious potential for causing extreme injuries. DISCLAIMER: Wintergatan is not responsible for any damages caused during any testing related to the Marble Machine X.

  12. EXTRA POINTS: Add a treadle pedal to the setup and test with the pedal instead of the crank.

2.0 Test Variants and Variables

2.1 Definition of "Test Variant"
Perform the test in different Test Variants, Example:

The test rig is completed and ready to go
Film and Record the first five minutes of you attempting to synk the rig with the metronome, without changing anything in the test rig setup.
The data from this first 5 minute attempt will be registered under "Test Variant #1"

We want to see the beginnig learning curve in the test - the first tries will be difficult and imperfect and that is exactly what we want to see - how fast and easy/hard it is for the tester to play in sync with the metronome.

Change one of the test variables on the test rig, and record another 5 minute attempt of syncing the clicks. The data from these 30 marble drop will be registered under "Test Variant #2" etc.

Change only one variable at the time, so we know what caused any differences.

2.2 - General Goal of this test regarding Moment Of Inertia

The general goal with this test is to make a flywheel with much more moment of inertia compared to the current Flywheel On the MMX.

Current flywheel data on the MMX:
Weight: 23kg
Pulley ratio: 5/1 (Flywheel turns 5 revolutions when crank turns one.)

My recommendation would be to start the testing range on at least 2X moment of inertia and then potentially increase from there.

Moment of inertia can be increased by either increasing speed of the flywheel (pulley ration) or weight of the flywheel. Experiment with different settings so we can learn what combo gives the best timing synk. Feel Free to experiment with any values that you think gives us the best result.

2.3 - Test Variables
    - Flywheel diameter (Thesis Range: 40cm-80cm diameter)
    - Flywheel weight (Thesis Range: perhaps 20-60 kg (?)
    - Crank pulley diameter (Thesis Range: 6/1-15/1 ratio, Crank pulley is the larger pulley)
    - Fly pulley diameter (Thesis Range: 1/6-1/15 ratio, Flywheel pulley is the smaller pulley)
    - Crank diameter (distance of crank from the center of axis, Thesis range: 8-14 cm)
    - Beater diameter, Thesis range: None
    - Treadle pedal travel distance (distance between highest to lowest position) Thesis Range: 8-15cm
    - Metronome values, use these exact tempos: 70 BPM, 100 BPM, 130 BPM, 180 BPM, 210 BPM
    2.4 Do not optimize for fast Start/stop:
    The Flywheel will be slow to start and stop, by design. Do not optimize for fast start/stop. Optimize for stable RPM while the Flywheel is running at the correct test speed.
    2.5 Audio Recording:
    Best: Record the metronome on one audio channel, record the beater tick on a separate audio channel.
    Great: Record the metronome on the left channel of a stereo track, record the beater tick on the right track of the same stereo track.
    Good: Position metronome close to beater, record sound of both straight into the video camera. No audio equipment is needed.

    3.0 - Data Collection

    3.1 - Register your Discord Username as a Data Contributor
    - Join The Wintergatan Discord Server
    - Assign yourself the "Tester" Role by Clicking the test-tube emoji here

    3.2 - Data Collection Form
    For every unique test variant fill out the Data Collection form here.
    Under "What test are you submitting data for" Choose: MMXt-Flywheel Test
    If you have performed 8 unique test variants, fill out the form eight times.

    3.3 - Video documentation
    Phone camera is OK
    Submit two video clips for each variant:

    Clip 1: Handheld camera, showing the setup up close with clear details, keep it short, under 1 minute.
    Clip 2: Camera on stand, showing the first five minutes of trying to sync the rig with the metronome

    NOTE: The videos you are submitting will be publicly available, do not share personal info or include people in the footage who do not want to be seen.

    4.0 General Info

    4.1 - Volunteering Collaborator agreement
    Read the Volunteering Collaborator Agreement here

    4.2 - Reimbursements
    We are thankful if you manage to build this test setup from free scraps and materials you already possess. But if you want to perform the test and are missing some vital parts, please reach out to to discuss reimbursements for expenses. To be sure of reimbursement you must obtain written approval from Joanna before making any purchases.

    4.3 - Discord discussion Channel
    Discuss and collaborate with the Community in the dedicated mmxt-flywheeltest Channel

    4.4 - Youtube Videos & Social Media
    Feel free to share any part of your process publicly in any way you like
    Use Hashtag #mmxtflywheeltest for all your online posts regarding these tests to make them searchable

    4.5 - Project Summarization
    Once you have submitted your test data - you are done!
    A data summarization team on discord will compile the meta-analysis
    The final Results will be published here on, on discord and on Wintergatans Youtube Channel

    4.6 - Credits
    We will publish an official database with credits and promotion links to anyone submitting valuable test data and have been involved in making Marble Machine Xt happen.

    4.7 - VIP World Tour Concert Ticket Reward
    As a symbolic gesture we will offer Two Free VIP World Tour Tickets to anyone submitting accepted valuable data. The VIP Part includes visiting the Marble Machine Xt up close, meet Martin and receive a personal thanks. This would happen before or after the concert, so that you and a friend can enjoy the resulting Machine you have been part of creating.

    But - there are some serious caviats that needs to be made here:

    - The World Tour might never happen, we will try but we can not know for sure.
    - The Definition of "accepted valuable data" to be eligible for the ticket reward will be a subjective decision taken by Team Wintergatan. We are sure that any serious test will be eligible but we need a limit to filter non-serious data entries.
    - There might be logistic or security hindrances on some concerts or festivals that makes arranging the meetup impossible, we will try to plan the tour so that the meetups are possible or arrange alternate meetups or acces to other concerts when necessary.
    - we are currently in the beginning phases of designing a database to track these rewards, its not built yet, it might take a while before you can check your eligibility status.

    - It feels kind of silly to give out a reward that might never exist, but when i think about it more it does make sense. We are all believing in this Machine and we are all invested in one way or another. If we can pull this off i would like everyone who has been involved in the Marble Machine X to visit the World Tour as part of the team who made it happen.

    Thank you to everyone who will be involved in this, Lets try to make it happen!