For the past few years (well, almost), I've been using a handmade gondola that was modeled after Sandy's design:

It has worked very well, and for normal purposes it's still my go-to gondola.
More recently, however, I've been experimenting with drawing on surfaces other than paper. In many cases, the gondola's "stabilizing plate" -- the disc that slides on the drawing board -- would smear the ink on non-porous surfaces. This even happened during some of my more elaborate vector drawings on paper, where disordered line segments often had the gondola passing over previously drawn areas dozens of times.

So, I decided I needed a plate-free gondola for those occasions. Such gondolas are not new. Most hanging-V bot hackers first try something like the classic binder-clip gondola. The first window and wall drawing bots I ever saw didn't have a plate. However, I've never seen a plate-free design that could drawn as accurately as the classic Noble arrangement.

After some experimenting, I can honestly say that I have designed a plate-free gondola that, uh, still doesn't work as well as the plated design. But it works nearly as well, especially within a certain drawing window.
Here's what I've been experimenting with:


(clicking on the images above will take you to a flickr group with more photos)

Here's a test drawing, done with a pen I knew was on its last legs, so sorry about the fade out:

The top of the drawing doesn't look much different than those I've done with my old gondola. The corners are tight, and the up and down strokes, while not perfectly spaced, are pretty good. (click to enlarge)

As we get lower, the corners and spacing begin to get a little funky, but not too shabby.

The bottom of the drawing, where things get worse, will have to appear in another post...

So, as long as I stick with smallish drawings done near the top of the board, the plate-free gondola is an adequate replacement.

I had to add spacers to my motor mounts (which are different from what you get in Sandy's kits), so that the beaded cable would hang perfectly parallel to the board. Otherwise, the pen tip (being the only thing touching the drawing surface) became a pivot point for much wobbling as the gondola got pulled to and fro. The position of the fender washers I use as weights can be adjusted so I can get the mass where I want it. The hanger for the servo can be adjusted as well.

I should've added thumbscrews to my original gondola a long time ago. Not having to reach for a screwdriver to get my pen out is so nice.

By thee eck Kong, great work - thank you very much for sharing it too. I have a couple of questions:
I think this places an emphasis on getting the balance of the pen spot-on - so that the pen balances already on it's own, at the correct distance from the drawing surface? I know that wasn't really a question, but I affixed the question mark so I could get a confirmation.

If it _is_ the case, then I guess this only really works on absolutely vertical surfaces - because the cord will always dangle vertically, and on a sloped board, it will exert more and more of a force the further the pen moves down the board. That is, at the bottom, the pen will be pulled a lot harder towards the surface than at the top. Why the pen tends to wobble around its tip so much at the bottom - there's more friction there.

The shapes look very much like the loopy patterns I used to get when the gondola was young and didn't have stabilisers, albeit yours appear much later in the drawing.

I'm oscillating wildly between _feeling_ that the y-structures must help a lot, and _thinking_ that they can't make much of a difference. Do you think that a balanced cord connection point might have been enough on it's own?

Very interesting! I'm off to tinker!

sn

Have people tried adding an extra cord from the bottom that is perhaps spring loaded (like one of the spring-wound dog leashes) or weighted. The purpose would be to both reduce oscillations, but also to provide additional pressure for the pen to hold against the drawing surface.

If it were weighted cord, it may be viable to prevent it from introducing its own oscillations by having the cord rub against a smooth surface. For example, if the polargraph were mounted to a whiteboard, and tilted slightly back, the marker-holder tray along the bottom might serve the purpose.

Note that I have no experience with this set-up. I just mention it as an untested idea.

-Joe

Hi Sephen -- yeah, it was my intention to share the files at some point. I'm trying to finish up a couple freelance jobs that have been keeping me from having any free time (I really shouldn't even be typing this...back to work, slave!). Then I'll have time to clean up my backlog of unpublished designs. I hope.