About a month ago I purchased a Digi-Comp I v2.0 from Minds-On Toys (see http://www.mindsontoys.com/kits.htm for information). This “toy” is a mechanical computer kit that, once built, can execute simple programs. I haven’t experimented much with the programming aspect yet, but have completed the assembly of my Digi-Comp I. This posting attempts to describe my process, mistakes and hints so that others might build and use their Digi-Comp I v2.0 with a few less problems.

January 28, 2006 update: This entry was originally posted on Jan. 16, 2006. Since that time, I was made aware by the manufacturer of the Digi-Comp that my Digi-Comp’s base was incorrectly oriented (blush). I’ve since rebuilt the kit, updated the photos here, and added additional photos and notes. The manufacturer has said that the manual will be updated to draw attention to how the base should be orientated.

Except for step 0, all steps described correspond to the steps in the Digi-Comp I User Guide’s assembly instructions. The following pictures can be clicked on to get larger size images from Flickr.

Step 0 – I unpacked the kit and quickly read page 3.

Packaged kit:

Before you unpack the kit, I suggest that you gather the required tools and materials: scissors, straight-edge, tape (which I ended up not using—more about this later), white glue (such as Elmer’s), and a lead pencil. When you unpack the kit, watch for small pieces of punched-out laminated binder’s board. When I opened my kit, one piece of the kit had detached itself from the binder’s board and fell onto the floor. You wouldn’t want to loose a spacer.

User Guide:

Read page 3, paying attention to the tips at the bottom of the page.

Step 1 – Created the frame supports and spacers.

When punching out the pieces from the binder’s board, make sure that you push out from the reverse side of the board and use the razor blade to cut any difficult to remove sections.

Step 2 – I followed the instructions in step 2 for the frame supports…

Well, I didn’t pay attention to the tips on page 3 and tore part of one of the frames. I ended up glueing some of the torn laminated paper back onto the board. From this point on, I removed the pieces from the reverse side of the board (not the side of the board with the cuts).

Step 3 – Pushing the push pin through the 12 holes wasn’t too bad, but my thumb got sore pushing the wire nail through those holes.

Step 4 – Using the template to poke the holes was straight forward. Cutting the grooves into the foamboard took some patience and a steady hand. I cut the grooves a bit small (which is better than too large) and ended up having to remove a bit of foam in step 5.

Notice that the gray side is up.

Step 5 – This step required me to remove some foam from the base plate grooves to allow the frame supports to be pushed completely down.

Step 6 Inserted the slider plate.

Step 7 – Inserted the rod support plate.

A bit tricky. I removed frame supports 2 and 4 to allow insertion of the plate. After the plates were in the slots, I reseated the frame supports.

Step 8 – Inserted the spacer stacks between the rod support plate and the slider plate.

A pencil helped wedge the stacks into place.

Step 9 – Inserted the lower and upper clock plates.

Do not do this:

It turns out that this one plate is oriented different from all of the others. The other plates are oriented “right side up”. This plate should be assembled with the reverse side up, like so:

Step 10 – Bent the crank wire. Luckily it bent correctly and I didn’t need to use pliers to reform it.

Step 11 – Added the plasting tubing to the crank wire.

Step 12 – Marked the position of the crank assembly on the base plate.

Step 13 – Mounted the crank assembly to the base plate.

Tested the crank’s range of motion by pushing the crank back (as pictured) and pulling forward.

Step 14 – Attached wire rods to front of Digi-Comp.

Step 15 – Attached rubber bands to rods and tabs.

Step 16 – Placed sliders on the slider plate.

Step 17 – Attached wire rods to rear of Digi-Comp.

Step 18 – Cleared 90 slits from flip-flops.

Lots of razor blade action. Make sure that you clear each slit now—it will be difficult to clear once the Digi-Comp is assembled.

Step 19 – Trimed the stop pieces.

Step 20 – Attached the stop pieces to the flip-flops.

I went on to steps 21, 22, and 25 while these dried (I let them dry a full hour to be sure that they were dry).

Step 21 – Numeral inserts.

Folded and then cut apart.

Step 22 – Cut plastic tubes. It was somewhat difficult getting straight cuts and cutting off the correct amount to leave no overlap.

Step 23 – Attached the numeral inserts to the flip flops.

Flip-flop A was a bit large for the slot in the frame, so I used the emery board to reduce the size of the flip-flop.

Getting the numeral inserts connected to the flip-flops was a bit of a pain. No need for the plastic tube to perfectly fit (tight) the numeral insert, since the numeral insert papers will be glued to the flip-flops in the next step.

Step 24 – Inserted the flip-flops and glued the numeral inserts.

Front:

Top:

This is it for the assembly of the Digi-Comp. All that is needed is the programming and the media to communicate the programs (input and output tubes).

Step 25 – Created the input and output tubes.

This was one of the more difficult steps for me. I created the Tube Guide, as described in the instructions, but it came apart after I inserted the second tube. I ended up transferring the tube length cutting guides to another surface (stack of paper plates) and then placed each straw on the guides and pressed down to flatten the straw to cut it with the razor blade. This turned out to form nearly perfect (square-ended) tubes.

January 28, 2006 update: When discussing this entry with the manufacturer of the kit, I learned that the method I used may weaken the tubes and that I probably rolled my Tube Guide’s strip a bit too tight. So today I tried again and was somewhat successful using the Tube Guide as described.

Attempting the recommended method, with a less tightly rolled Tube Guide:

Success. The tubes stand:

I still prefer my alternate method. They are easier to cut and the resulting tubes seem to work well. Also, if they wear out, I’ll just buy and create more tubes. But be warned: not following the manufacturers recommended assembly instructions could result in reduced performance and may invalidate any warranty. ;-)

Testing – I immediately went into experiment 1 in the Lab to see if the Digi-Comp worked. I only had one real issue: one of the four phases of the clock cycle didn’t work because of friction. I ended up solving the problem by using the emery board to smooth the notches to the right of the input rods of the lower clock plate, then added graphite (pencil lead) to these notches, and bent the input rods to bow very slightly to the right. Finally, when I push the crank back (to run phase), I hold the machine with my left hand and use my right hand to push the lower clock plate to the left. I also watch the input rods, and if the lower clock plate binds, I jiggle the input rods to determine which rod and/or clock plate notch is causing the problem.

Conclusion – The assembly process took quite a bit longer than I expected (maybe 6 to 8 hours total, while watching movies). The assembly instructions are very good. There were a few steps where I needed to either read ahead or really think about the design to complete the step correctly. Some may have difficulties building the Digi-Comp I or end up with a machine which needs tweaking to operate correctly. The Digi-Comp I v2.0 is made from laminated binder’s board, so I expect that it won’t last for too many clock cycles, but only time will tell. Since the materials can break, I usually feel compelled to watch the lower clock plate during each run phase to ensure that it does not bind, taking away from my ability to learn how the machine works. The price is great—much better than purchasing an original Digi-Comp I on eBay. The Digi-Comp seems pretty cool after performing just one experiment. I plan to complete the Lab in the near future.