Electromechanical valving prototype

Simplest possible thing you can imagine: five solenoid valves, each feeding a separate pipe, actuated in sequence.

The effect was about what I expected: not very smooth.  If I put another five of these in a row, it might look better, but in general this is what I'm going to get.  If I want to do this for real, two things to improve:

1) I want to make these all highly linear, in a single tube format.  That might mean devising a way to create end-to-end pipes or tubes, with internal threading and seals between them, and then using copper tube to run out of each unit.  Or, encasing the whole thing in a larger tube.

2) Diffusing the flame a bit to improve the visual effect.  I can't fit the "rail" I made over this unit (the pipes are a bit skewed because, with the caps, they don't fit properly in a single line.  But putting the whole thing in a larger tube might do it.

I'm not thrilled with the effect, but not sure how to improve it other than using MOAR SOLENOIDS.  And the amount of labor to get just one of these working right is quite large.  I still haven't figured out a way to multiplex with fewer valves.

Another rotating mechanical valve prototype kind of sucked, but I have an idea for improving it:

This is a 1/2" pipe inside a much larger diameter pipe.  The larger diameter pipe has slots cut at different offset angles. The difference between this design and previous similar designs is the gap between the inner and outer is intentionally large so the flame is always lit.  The outer simply "gates" it.  Two issues came up here:

1) This video makes it look like it works better than it does.  The wind is a huge factor.  I saw this on the playa last year as well with my smaller flame art.

2) The gating doesn't work that well, I saw similar "leakage" to what I saw with previous rotating prototypes.  My thought here is to add yet a third layer, on the outside, which has a series of holes along the top.  This increases the complexity a lot, again, but I do like the way this one gives a smooth translation effect.  I think it's worth trying.  

My original estimation was that electromechanical valving would be more expensive but easier; it turns out that doing it right isn't really easy.  The amount of plumbing required is significant (see above.)  So it's not clear that, even with a three pipe design, that this is much more complicated.  Especially if no precision machining is required, which it isn't.  Work-holding on a 4-foot pipe is a problem that I found (I created the slots with a 3/8" end mill, and the pipe would slip sometimes.)  But not insurmountable.