Some problems that need working out

[sfield]

1. When installing the software on Windows Vista, one of the components was missing.
By Googling for the name of it, I was able to find it at SparkFun and install it.
The software then worked. Unfortunately, I did not record the name of the missing
driver.
2. If the spring is not made to be very tight, the egg (we actually used a light bulb since
they are more easily erased and re-used) wanders in the rubber cup as it rotates. This
of course makes the distance to the pen vary, which makes the pen either draw when it is
not supposed to, or not draw when it is supposed to. Tightening the spring solved this
problem, but it should be mentioned in the instructions.
3. Even after fixing that problem, the very small difference between up and down on the pen
makes it impossible to get a drawing that does not have missing parts or lines drawn
when the pen was in the UP position. Nothing we tried (eggs, ping-pong balls, light bulbs)
could be centered so accurately that there was not a millimeter of difference between
the surface and the pen at all times when the pen was up.

Because the difference between pen-up and pen-down is only about a millimeter, we
always get lines drawn when they should not be, or missing lines.

Three solutions come to mind:
A. Give the solenoid more range of motion. A millimeter is not enough.
B. Use a lighter pen. The solenoid looks like it is trying to lift the pen, but does not lift
it as far as it wants to. This may be because the rubber band and zip-tie spring is
not strong enough to actually lift the pen. I tried to pull the rubber band so it would
be tighter, but I had no luck. I will get a new zip-tie and rubber band and try to simply
replace the original with a tighter arrangement.
C. Make the spring stronger, or eliminate it by reversing the solenoid. If the solenoid
lifted the pen (instead of pulling back against the spring to allow the pen to fall) then
the spring would not be as critical (since gravity would assist it or make the spring
unnecessary).
D. I think there is a more robust solution that has other benefits.
Consider replacing the current pen and solenoid arrangement with a foot (like a sewing
machine has) that rests on the egg at all times. Then have the pen and solenoid move
up and down relative to the foot, instead of relative to the center of the sphere. This
would always guarantee that the pen was either on the surface or above the surface
(pen-down or pen-up), and would allow the device to draw on less spherical surfaces,
such as apples, cylinders, hour-glasses, etc.

[EmbeddedMan]

You are absolutely right about the pen height delta not being enough. Bruce is the king of getting the solenoids to work well, so I hope he chimes in here with better advice. But I know that the rubber band at the bottom of the solenoid will get loose over time. You need to grab one end of it, and pull it tighter across the bottom of the solenoid. The zip tie will hold it in place, but you want it as tight as you can get it without it coming off. This tightening will give you the most distance up/down. Also it is critical that the solenoid be adjusted properly in its little bracket. You can slide the solenoid up and down. What you want to do is adjust things so that the solenoid hits its internal hard-stop when it is energized (down), and that when in this position, the weight of the pen arm and pen hold the pen on the egg.

The solenoid is definitely the part of the design that needs the most work. We are currently experimenting with other ways of creating the up/down pen motion, with the goal of making this as simple and reliable as the rest of the egg-bot kit.

Thanks for all of the great advice - we really appreciate it!

*Brian

[jschrempp]

Me too. Just posting so I can sign up for notifications.

[jschrempp]

Hi Guys,

I have sent emails and been watching this forum for help on getting the solenoid to work correctly. Can you shed some light on your current thinking? It would make me feel better if you would post "still working on it" once a week. Otherwise I start to feel like there is no solution coming and I should give up and return the kit for a refund.

Jim

[bruce]

Hi Jim,

Very sorry for the delay in addressing the problems you've outlined. Getting detailed feedback like you're providing is crucial to us-- thanks for taking the time to describe the problems, and for offering thoughts on improving them. It's clear that the biggest problem with our beta design is the pen up/dn. It was the best I could do in the time I had to get the kit ready for the Maker Faire. I had actually planned not to have pen up/dn, but I was seduced by getting the rubber-band solution to work. We know it isn't the right way to go; it was just what we had at the time. And we agreed that it was better than no pen up/dn at all. The way to get it to work is to pay very close attention to the "pen ud/dn window." This can be envisioned as a spherical shell , with "holes" at the "poles." The shell's thickness is the distance between up and dn. The pole holes are the areas the pen cannot reach because of the egg-holding cups. The two key goals are:

1) Make sure the distance between pu and pd is >1/8" (1/4" should be possible).
2) Make sure your egg's (ball's, bulb's,) surface is within that window.

Starting with 1) --
First position the solenoid so that the plunger holds the pen-arm (with pen in it) up-- meaning approximately horizontal, and that the pen+arm are being held up by the plunger and not the side of the solenoid itself. You should be able to push the pen down and see the plunger move down about 1/8". Slide the solenoid up or dn in its holder as needed. If it appears that no matter where you position the solenoid, the pen+arm are too heavy for it to hold up, then the rubber band needs to be tightened. Do this by grasping the longer end and tugging it, while pinching the other end to make sure you don't pull it completely off. This can be tricky, and if it comes off just put the band over the solenoid, and slip the wire tie over it. It takes a little patience, but you can adjust the band's tension quite readily. You should be able to achieve enough tension so that the plunger will hold the weight of the pen+arm without sinking very far. Once you have that done, test pen dn (energize the solenoid). You should see the pen move dn 1/8-1/4". If it doesn't (but you can push it down that far) the band is too tight. Pull the part of the band inside the wire tie to move a little more rubber to the inside, producing less tension. Yes this is cumbersome and it is hardly a great solution-- but it does work. I have been able to get it to work every time.

Once you have your pen moving up and dn well, focus on 2). The multiple holes in the pen arm allow you to use different diameter eggs (up to 4"). Select the hole so that the pen+arm are close to horizontal when the pen is up. If you are using a spherical "egg," then the pen-motor is set all the way up on its slots (this places the pen-motor axis at the height of the sphere's center and is the default). If you are using a real egg, or other "squashed sphere" (football shape), then you will need to slide the pen-motor down to get the egg's surface to stay within the pen up/dn window. Of course, if you lower the motor, you may have to increase the height of the pen-arm by changing the hole that mates with the pen-motor shaft (depends on egg diameter). Yes, again, this is cumbersome. But once you have things set right, you can plot on the same shape/size egg again and again, without needing adjustments.

I appreciate your ideas as well-- I've considered having a "foot" touching the egg's surface (a "feeler" which would allow following irregular surfaces) but unless you are using ink which dries extremely fast, it will smudge previous lines. Brian is working on testing a new approach (using a small hobby servo) for the pen up/dn mechanism. This should make things much better. In the mean time, I hope these pointers help-- please let us know.

Bruce