S3/BlocklyProp - best quality traces

John Kauffman

Our middle school has an offer to display S3 traces in our public library. I want the students to re-run them to be as polished as possible. We use about 60 cm square sheets of poster paper. Below is what we do for best results. Any more suggestions?

Run at slow speed:

reduces over-shoot and squiggle coming out of turns.
however, leaves larger "bleed dot" when pivot turning

Get paper relaxed and as flat as possible:

use very flat table surface (laboratory bench)
carefully clean bench and paper of any small debris
relax paper with slightly damp cloth, stretch and tape to table then quick dry with fan.

Check wheels / tires:

tires cleaned witch damp cloth and dried by fan.
carefullly clean any small debris that will intro error to encoder setting of circumference.
tires well seated
tail ball clean and free-running
no hair wrapped around axles

Marker:

Keep one set of markers just for high-quality traces
Check that tip is sharp (not worn to be rounded, no extra fuzz)
ink is flowing (not slightly dry)
Shim around marker so is tightly seated in port:

eliminates some wiggle during the turns
reduces wiggle from re-seat as come out of turn

Size and position

double check scaling value in code so result fits on paper with margins correct for intended mat & frame
use preliminary run to hang on wall, back up and judge if want changes before final
establish start position and angle (before every trial run we make soft pencil marks at 3 contact points so can replicate)
If desired, scale so leave space on paper to tape on print of block code, print of text code and/or student name

At run:

Insert marker at last minute to reduce bleed spot
clear plenty of space beyond paper for wheels going off paper during turns

Code:

Include at start a pause and count-down beeps so know exactly when motion will start.

ToDo:

Experiment with papers of different surface finish and thickness
Experiment to see if there is a straight-line speed which is optimal: probably slow, maybe not slowest
Experiment to see if there is a turning speed...
Try thin-point markers instead of thick sharpies (but may not be dark enough to see the trace at a viewing distance)
Try lifting pen by hand during the pivot turns (we have not had success with sevor-lifter. A student might look at improvement using 3D printed parts)
Somehow use Foam-Core (and maybe vacuum press) to keep paper as flat as possible.

Any other suggestions?

Thanks, John

Phil Pilgrim (PhiPi)

Poster board instead of paper maybe? Different background colors would add variety to the exhibit.

Test the drawings with dry-erase markers on a portable whiteboard before committing them to a permanent surface. (You can wrap the dry-erase markers with masking tape to fit the S3's pen holder more tightly.)

-Phil

John Kauffman

Phil: Good thought on poster board - the art faculty are just down the hall so should be able to get some test sheets.

Whit

John - Scribble with light... See this thread! https://forums.parallax.com/discussion/160957/nikos-fibonacci-spiral-challenge/p4

and this... http://forums.parallax.com/discussion/163960/the-s2-led-dot-matrix-printer-thread

John Kauffman

Six of eight students are now done their work and it is displayed in the lobby of our public library.
https://parallax.com/news/2017-05-22/s3-robots-drawing-their-way-elementary-school-programs-these-educators-are#comment-form

A few more tips we learned:

During development we saved $ on paper by students using rolls of cheap art paper. The same sheets can be used multiple times by using different pen colors. There was a suggestion to run trials with dry-erase markers on white board but we were unable to find a board of right size for our needs. We ended up drawing a lot of soft pencil lines on the paper for guides, margins, etc. That would be harder on a white board.

For a finer line we bought a set of fine-point sharpies. The barrel is narrower so we made a collar from cardboard. Wraps of masking tape would have worked, but we are using a dozen different colors of pens. (One student has designed a plastic collar. CAD drawing is done; first 3D print is supposed to be delivered to me by Friday.)

One of the most important practices we learned is to mark on the paper the location of each wheel at the start of each run. Without this info it is very difficult to exactly replicate runs.

Oddly, we found that sometimes running at slower speed made a squiggle line that did not show up at 100% speed. Perhaps pen is reacting to fine variations in paper at slow speed but at higher speed has momentum to jump over them. For high velocity of stops we got better results by placing a 175 gm weight on aft deck of the S3. This prevented S3's momentum in a quick stop from lifting up the back wheel. Next time we need to use ramping.

Our final version went on Bristol Drawing paper #300. Looks great but $25 for 20 sheets. Many of our pieces are displayed flat in low vitrines. The vertical ones we glued to cardboard backing using a very light coat of spray-on glue. Use only slightest amount of spray - one pass about 12 inches away from paper; do spraying outside. The high quality of the paper made better results than if we had run directly on the cardboard surface. But that is an aesthetic judgement.

It was helpful to have large paper cutter, T square (36 in), xacto knifes and Stanley knives. Heavy cardboard backing was much cheaper than foamcore and worked fine. Next time we will try to enlist high school art students that are learning to cut mats.

We did screen shots of the block code and pasted in Word for some cropping. (I guess other app would have been better, but we were comfortable with Word) Before crop & size, measure space for label on the piece. Before screen shot move the block of code to far top left corner if you plan to include block palette in print-out. Longer programs were harder to fit on one screen shot so we removed the start-up beeps and waits. We did keep comment block with name and student initials. In Word, above the screen shot, we added a title with name of the work and student name. Check with parents about student names: some wanted no name, some full name and several wanted variations with initials. Don't trust what the student's say.

The scaling and positioning were good for students to use algebra that they are learning this year. We worked hard to be mathematically predictive rather than running many empirical tests. Some students got it, others had trouble. Next time I'll ask a math teacher to come by for a guest lecture to talk them through scaling.

In the vitrines we placed actual S3s and a BeoBot. To have a sharpie in place would mean a splotch on the case floor. So we pulled out the tip of a sharpie and hammered on a pen cap to seal the hole. Next time I'll try sealing the hole with pinch from hot pliers.

Before the run, masking tape all edges of paper because wheels of S3 will probably go outside of paper and can trip as come back on. Tape is easily removed from the high-quality Bristol paper.

Because it is easy to do a second run, we could make a copy to hang on the lab wall as well as give the students the copy from the exhibit. It never hurts to have families see how their Parents Committee money is being used.

Feel free to message me if you have other questions.