Thursday, April 3, 2014

The design challenges and failures of the Sit/Stand/Walk/Flowerpot Desk

This is Part 3 in a series about my Sit/Stand/Walk/Flowerpot desk.  See Part 1 for why I built it and Part 2 for how.

Reality Check

In this picture of the desk, I removed all of the cables and put some space around the desk so you can see the design clearly.  (I also used a good photographer.)

Here's what it actually looks like in normal use.

I use it every day.  Sometimes in the sitting mode, usually in the standing mode, occasionally in the walking mode.  The treadmill is a little noisy (not treadmill whooshing noises, more intermittent clattering) and the controls are unpleasant to use.  The flowerpot mode doesn't really work because the monitors don't go down far enough to get out of sight and because the muehlenbeckia is surviving but not thriving.  I was sitting too much so I got rid of the chair and in day-to-day use it's now a stand/walk desk, probably 80/20.

Original Goals

So how did it go, as judged by my original goals?
  • I can build it using only hand tools appropriate for an apartment balcony. YES.
  • It costs no more than a commercial equivalent. YES.
  • It supports my pair of 30" monitors (~20 pounds each, total width of bases around 35 inches). YES.
  • It has sitting, standing, and treadmill walking modes. YES.
  • It accommodates my chair (an Aeron). YES.
  • When sitting, I can put the keyboard in my lap. YES.
  • It can change between sitting, standing, and walking modes with less effort than moving 20 pound monitors up and down, and without any unplugging and replugging. YES.
  • It must be pretty enough to be in the living room, and ideally the large monitors can be put out of sight. YES.
All goals achieved.  But I wouldn't build it the same way a second time.  It took a month and far more manual labor than I expected, and it cost as much as a commercial equivalent; it didn't save me any money.  The main shortcomings of the desk as built include:
  • It took more time and money to build than I was happy with
  • It's a little bit wobbly
  • It has a big footprint
  • It's not super-easy to change modes (i.e., move the monitor, keyboard, and mouse up and down and switch between chair and treadmill)

Time to assemble

Three main factors make this take a month to build.  First, I used a lot of trial and error to finish the design.  This even included changing the lengths of the vertical pipes, which meant buying new pipes.  With a proven design, this wouldn't be a factor.

Second, sanding all of the black paint off of all of the pipes and then masking them, priming them, and painting them white took a week or so.  And was even less fun than it sounds.  One obvious step would be to buy galvanized, but it would still have to be painted.  And the front pipes, that the monitor and arms slide up and down, has to be bare metal.  I left it sanded but not polished, and it works fine, but it would probably work better, and it would certainly look much better, if it were polished.

Third, the wooden bar that holds up the monitors required a bunch of tricky hand-sawing, drilling, and screwing to get the two wedges per monitor mount needed to get the monitors level.  More expensive monitor arms could help, but even many $200 arms can't handle a 30° tilt.  Another option would be to make the front bars vertical instead of tilted back.  The reason they are tilted back is to give the whole frame a triangular shape, like a swingset, in order to improve rigidity.  Since the counterweight hangs straight down, the front pipes have to angle to keep the monitor bar from interfering with the weight bar.  Another possibility would be to make the front pipes vertical, so the monitor goes straight up and down, and to tie the weight bar to the rear (angled) pipes in the same way that the monitor bar is tied to the front pipes.  This could also reduce the total depth of the desk.

The A-frame shape works well in one dimension; the frame is rock-solid in a forward-backward direction.  However, swingset feet can be bolted to the ground, whereas this desk rests on felt pads on a wooden floor in an apartment; between that and the square shape it has from the front, it has some jiggle side to side.  The diagonal brace in the back helps but only up to a point.  So, since the A-frame isn't perfect, maybe abandoning it for a space frame would make sense. This would complicate the design, though, since many more bars and angled Klamps would be needed for rigidity.  Also, two sets of pulleys would be required to transmit the weight from the front vertical plane to the parallel rear vertical plane.

Difficulty changing modes

I modified the design of the monitor bar many times, both before and after building the desk.  Originally I thought the bar might roll up and down pipes on little wheels like skateboard wheels.  When I was ordering the parts I decided to try this design, with the Kee Klamp sliding up and down the pipe and the wooden bar connected to the Klamp.  The nice thing about this design is that the monitor bar can't fall off the pipes completely and break your feet.

But it works terribly as a slide.  The metal Klamp jams into the unfinished steel pipe at the tiniest impulse, and there's enough slack (a few millimeters) between the inside of the Klamp and the outside of the pipe that lifting one side up a few inches jams the other side.  Worse, the weight of the monitors is enough to twist the wooden bar enough to put twist the Klamp enough to lock it in.  What finally worked was to line the inside of each Klamp with a strip of teflon sheeting 0.020" thick.  This prevents outright metal-on-metal locking, and I suspect the slipperiness is secondary.  With a bit of effort to lift the monitors, the Klamps then slide up and down quite easily—thanks to the counterweight—and there is very little locking when one side is higher than the other.  And, there is enough tension in the system normally that the bar is quite stable and doesn't slide up and down with a simple touch.

The U-bar visible in the picture is a sort of added bit of security that doesn't normally do anything and is probably useless.

The keyboard and monitor are on their own arms, and have to be unbolted, slid, and bolted back each time I change modes.  This is not a big deal for the wireless mouse, but my keyboard doesn't come in a wireless model, and I have to move it carefully or the cable pulls it off the shelf.  I considered attaching the mouse and keyboard trays directly to the monitor board so they would all move together, but this design is simpler and cheaper than articulated arms.  Also, I put the keyboard in my lap when I set and so I need to be able to move that arm completely out of the way.

Large footprint

It's over three feet wide and six feet deep and seven feet high, and since the monitors stick out almost a foot on either side, it's even wider.  With a pure box design instead of the A-frame, it could be a lot shallower.  Also, there's nowhere to put the computer CPU, so that's another two feet square of footprint.  I used a tall Elfa shelf thing on wheels, but it would be nice to suspend that all within the desk.


Even though the keyboard and mouse are not directly attached to the monitor board, typing action can still cause the monitors (which are hanging at the ends of swing arms) to jiggle a little bit.  The whole frame is rock-solid in a forward-backward motion but has a bit of sway when pushed side to side.  The treadmill doesn't touch the desk (and the apartment building has very solid cement floors under the fake wood flooring) and walking on the treadmill does not affect the monitors at all.


It seems fairly safe?  The pulley system is supporting a bit less than a hundred pounds; that weight goes to the top Klamps, which sit directly on more Klamps, which sit on the tops of the pipes so they can't slide down.  Therefore the weakest point is the connection of the cross-pipe in the A shape, which is held in tension by the bolts in the Klamps.  Those connections haven't given any sign of slipping.  If the pulley wires somehow failed, most likely at the wire clamps, the weights would fall a few feet; the monitors would slide down but should be stopped by the other Klamps bolted to the front pipes.


If I were to do it again, I would experiment with a shallower, box-shaped version instead of the A-frame.  I might put the mouse and keyboard on a separate stand to minimize vibration, but probably not.  And I would buy galvanized pipes instead of black, I would still go through the hassle of hand-painting them, and I might polish the front pipes.  I would consider spending more on a nicer treadmill so that I would want to use it more.

(All posts about the Sit/Stand/Walk/Flowerpot Desk Index

No comments :

Post a Comment