Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

James Evens

Member
  • Content Count

    2,493
  • Joined

Awards

This user doesn't have any awards

About James Evens

  • Title
    Veteran

Profile Information

  • Location
    Europe

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. More practice required and parallel rail/guide but it can be done.  This was a quick first try (freehanded ).

    One of the problems with wood is that it contracts over time and with that aiming for 50 µm accuracy (both sides together 100 µm) is adequate.

    image.thumb.png.cac915fd35e909ae2790e8487dfb1755.png

     

    Noticed a "critical oversight" the slots/cutout only works in the longitudinal direction. On the transverse direction it is to brittle meaning it will break and it must be the last step (not the first like initially planned).

    1. This was spruce wood. Choosing wood which is stronger in the transverse direction will help.

    2. Defects can be hidden below the surface.

    3. Smaller router. Using the smaller 3mm bit might reduce the force on the wood.

    4. Thicker frame. For the longitudinal side 5mm is a good value. For the transverse direction it needs to be a lot more.

    5. Fancy support structure for the wood and carefully handling it until fully assembled (as a alternative, not feasible).

  2. Currently the latest mac address video isn't promoted on the sidebar.
  3. Apple: Everybody can join and get's the normal rate but they also have contracts with large customers which are confidential. Floatplane (as far as I know): You request to join and then get the same conditions as anybody else.
  4. Depends. If it is still a paste don't bother with it and just reinstall the heatsink. If it is "dried" it needs to be replaced. As you propably have no idea if the current thermal paste is one of those which survive years without drying up it is safer to have some new laying around.
  5. Normally this kind of databank would be associated with advertisement/campaigning agency's.
  6. Cherry MX red are the worst: extremely scratchy. Everything else is personal preference: no tactilely, no feedback.

     

    Sadly I completely harvested the keyboard PCB before checking the oscilloscope export/files. 

    No hardware/keyboard PCB. No repeating the measurement's. 😕

    Spoiler

    image.png.24294795f7320a62899be0dd425b0c31.png

    image.png.c5986364141c12e0bee12f942aecbd72.png

     

    The only data I have is one cycle without proper probing. Doing the math on this data there would be a 45 mA spike on the Roccat ryos mk pro backlight (all of the good/proper measurement's are lost :/).  The entire LED drive/circuity seems wired from the µC driving the matrix at 80 Hz to how it works down to the current spikes and reverse currents when the LEDs are disabled.

     

    Graph1.thumb.png.e488054e5c3d12b276ba8b717da04be2.png

    note: actual voltage is x10 higher.

     

  7. They are all WS2812B. As long as the connector fits it works with auro, fusion, etc.
  8. a.) internal sample b.) from another water bottle. Could be a camelbak.
    1.   Show previous replies  1 more
    2. James Evens
    3. Murasaki

      Murasaki

      11 minutes ago, James Evens said:

      For a single day. Didn't went well.

      Why am i not surprised LOL

    4. lewdicrous

      lewdicrous

      1 hour ago, Murasaki said:

      wtf there was a kinguin one?

      Here's their only interaction on the forum since joining.

       

  9. Copyright is way to complicated. In this case it probably is a violation but in other cases you are totally fine to reverse engineer parts of a product to make your own. In this case the probably just fear the competition/customer using the opensource alternative instead of paying for the same thing.
×