The Fusion Silvertouch Power Mach4 Turbo Customplus Platinum Team 7000 Mega™
Cartridges
Holochip Fabricators
Near Eye Display
Primary Researcher
Unassigned
Traitorous Rebels
Misc.: Paper Introduction, Nature Article Specifications, Pulsed laser plasma, Big intimidator focal point images, old Princess Leia Team page
PWM Intensity curve linearization
Increased PWM frequency on Arduino
CO2 Galvo scanner documentation (A.K.A. Golden Plates)
Big intimidator focal point images
Trapping factors
Plasma display
See Erich's page for a more detailed daily log.
15 March 2016
Got a picture of a color circle!!!
9 March 2016
While waiting for a new 1W violet laser, I decided to see what I could do with my 100mW diode. I put it in where the violet laser was before, with the UV filter etc. I moved the 2nd lens as close as possible to the galvos in order to get the focal point a little farther out. I was able to hold a few particles for roughly 5 seconds:
6 March 2016
PWM control for the red/green/blue lasers on the galvo scanner is now working. There are only certain pins on the arduino that are PWM capable, RGB control is now on pins 4-6.
To test how smoothly the colors vary, I wrote some code to trace out a circle while sinusoidally varying the intensities of each color, out of phase with each other by 120 degrees. As you can see, the transitions are pretty smooth, and there is a good range of colors. The lasers are only going from 0 to 80 (out of 255), because they seem to reach their max brightness well before 255. This picture is just the lasers hitting a piece of paper, no particle is being held.
8 March 2016
Raster.
Non-shaded butterfly
24 February 2016
Raster.
Detailing the 'Y'
12 January 2016
Raster.
Images Worth Oogling
Leia:
BYU Y
7 January 2016
Raster.
Before
After
21 October 2015
First Attempt at BUtterfly in Hand
Second Attempt at Butterfly in Hand
Best Butterfly in Hand
Butterfly Perched on Finger
15 October 2015
Erich Nygaard
After cleaning the lenses and mirrors (especially the galvo mirrors) I was able to get particles to be held pretty well, similar to what I have seen before with this system. It may be holding better because the beam profile is different. This is a picture of what the dot looks like:
It is brighter on one side than the other, probably because it is going through the first lens off center. This also has the effect of bouncing reflections from the lens farther out of the optical path.
The following pictures were taken from behind looking down, in line with the direction of the light exiting. It was a bit tricky minimizing the splash from this angle.
Not cropped: You can get an idea of how much we have to block out in this picture.
We are going to take pictures from the other side so we can put a hand under it.
I also increased the drawing speed up to 0.1 ms delays between adjacent points (that may be lower than the actual delay) and took some video. I can't upload the file, but it is in the group folder. It held a particle at least once at that speed (at about 4:28 in the video).
With higher speeds, the right side of the butterfly is squashed (down to a flat line, in fact) due to the x-axis galvo being slow over part of the range. Assuming it is the galvo and not the controller, switching out that galvo should fix this problem. I have pictures of this as well, but can't upload them now.
Picture of the exterior of the box:
31 August 2015
Erich Nygaard
Results of overnight galvanometer test from 8/28/15 - 8/29/15.
Laser at 7W, moving particle back and forth 10x, cycling between 10 speeds (ranging from .11 m/s to 7.8 m/s). Checking for particle (after 4 second delay) after each pickup and each back and forth test.
We need to improve the pickup method so we can get more pickups each night (more than 100-200).
19 August 2015 Erich Nygaard
Arduino mega digital outputs are at 5V. This means you could damage the raspberry pi by connecting an arduino output to a Rpi input (which is 3.3V). You can make a simple circuit like this for logic level conversion: http://hobbytronics.co.uk/schottky-logic-level-conversion.
I connected the arduino to the Rpi with a circuit as described in that link (except with a .6v diode instead of .3v - should work). The galvo test code we were running today is still not correctly waiting for the “done” signal from the arduino. The next step would probably be to double check the pin connections and possibly try a different input pin on the Rpi.
Jacob Van Wagoner
(Update: The below is obsolete, the analysis done was done using the occlusion code found on the hard drive Under the occlusion folder)
Testing occlusion by brightness analysis:
11 August 2015 Joel
Nuff said. Its half a butterfly.
Jacob Van Wagoner
The original stand 2.0 photos had an opening in the middle so it was decided to cut out some paper to make a black covering for the hole to make it more professional looking, but when putting the paper in some of the mirrors got bumped so you can't see the particles in all of the mirrors. So the bottom left four are the best examples of occlusion.
5 August 2015 Jacob Van Wagoner
In order to prove occlusion we suspended a particle with the laser then using an array of eight mirrors reflecting the particle from different angles we took a picture of all the mirrors in a single shot. In the first, rough version of the mirror stand (using clay) the mirrors were not all perfectly aligned, as seen in the following photo, but we are making a 3D printed stand that will allow for adjustments.
The above photo was taken while the laser was holding multiple particles so that it would be highly visible from all the angles. The following photo however was taken so that we could see the varying brightness of different angles:
We can see in this photo that the second mirror from the left reflects the particle and is very bright, whereas the photo to it's right is dimmer and the one to the right of that is dimmer still. Because the different angles have different levels of brightness we can conclude that occlusion is a factor in this method of volumetric display.
17 July 2015 Joel
Here are the first pictures from the new intimidator. For the most part everything is smooth but we have a very large difference between the pixel resolution for the Z axis and the y axis. In the picture shown below of the partial square there are 10 steps in the y for every 600 steps in the Z which is what shows as choppy steps along an otherwise fairly smooth line.
You can visibly see each of the small steps it takes. We are running in the smallest micro steps possible for the Y motor. So this may cause future problems with drawing.
The following picture is without the Y steps on the bottom. This is just the Z moving which may show part of the problem from before. The particle dances around the line but doesn't draw straight. I think this comes from the lens moving a little as it travels up and down. I may be able to fix it by securing the motor and shafts more to the chassis.
The final problem I had was getting the particle to hold for the entire drawing. This was the most complete square that I drew today. Though there were many particles at the beginning and it has the same problem as mentioned above with the wavy line.
We will need to get additional lenses if we are to match the setup on the first intimidator with optimal holding time.
15 July 2015 Here is a better picture of the gray scale levels:
10 July 2015
Jacob Van Wagoner
Here we go, the gray scale photo. This was taken by using pulse width modulation. the red laser had 12 V on it and the trigger was hooked up to a PWM pin on the arduino. When we first started using PWM we saw that the change was not linear so we found the equation x = 1/(1+EXP1)*255
(http://electronics.stackexchange.com/questions/1983/correcting-for-non-linear-brightness-in-leds-when-using-pwm)
nonlinearpwm.pdf
where x is the desired PWM and y is a number from 0 to 255 which represented the PWM rate. It was found that there were seven levels of brightness (one of them being off) so using that we were able to chose a number from each level and that is what is shown in the picture below.
29 June 2015
Jacob Van Wagoner
Here is the successful image of a red square with a square taken out of the middle:
Also made a block of solid black liquor in order to test to see if it melts as it is picked up by the laser. The laser (4 watts) still picked up even though it was a solid block.
Last week I tried the Poisson spot again and took pictures along the length of the beam in four cases: 1) far from the focal point and before the focal point (far left), 2) near the focal point and before it (center left), 3) near the focal point and after it (center right), and 4) far from the focal point and after it (far right).
26 June 2015
Jacob Van Wagoner
Last night we ran 1500 tests to see how the laser at 5 watts picked up. 475 picked up meaning it had a 31% pick up rate.
Also added a line to the code that keeps track of which tests hold and which do not. In this test we see a correlation that shows that over time less particles picked up:
Using this excel file: 15625pickupexcel.xlsx
25 June 2015
Jacob Van Wagoner
Last night we ran 2000 tests to see how the laser at 4.5 watts picked up. 680 picked up meaning it had a 34% pick up rate.
Also added a line to the code that keeps track of which tests hold and which do not. In this test we see a correlation that shows that over time less particles picked up:
Using this excel file: 2015625pickupexcel.xlsx
Also I added a section to the Google sheet, it is the tested pick up rate from 2 watts to 4.5 (as of now) and can be updated by replacing the “-” with the resulting percentage.
Poisson spot -
While taking pictures of the focal point the three most important (in my opinion) pictures are the following:
From left to right they are just before the focal point of the laser going through just the glass, obstructed by a small dot, and obstructed by a large dot. But as we can see there is no change, they all look identical. Potential flaws could be that the pictures were taken too far from the focal point so the Poisson spot was not visible or that I may have set it up poorly. Tomorrow I plan to (1) have some else set it up and (2) take photos all along the beam for all three cases, not just one like I did tonight.
24 June 2015
Jacob Van Wagoner
Last night we ran 1500 tests to see how the laser at 2 watts picked up. 125 picked up meaning it had a 8% pick up rate.
Also added a line to the code that keeps track of which tests hold and which do not. In this test we see a correlation that shows that over time less particles picked up:
Using this excel file: 2015624pickupexcel.xlsx
23 June 2015
Jacob Van Wagoner
Last night we ran 1500 tests to see how the laser at 2.5 watts picked up. 329 picked up meaning it had a 21% pick up rate.
Also added a line to the code that keeps track of which tests hold and which do not. In this test we do not see a correlation of if the test picks up over time:
Using this excel file: 2362015pickupexcel.xlsx
Colors that we are currently able to use:
The timed raster was trying to make a hollow square (left) these are the pictures we got (right two):
22 June 2015
Jacob Van Wagoner
Data from the weekend's test
Parameters:
Lens 1: 50mm
Lens 2: 75mm
Laser Power: 3 Watts
Dip Time: n/a
Vibrator: none
Filter: Aperture half closed/half open
Days since particles ground: 0
Total Tests = 67
Average Holding Time only for Tests where particle was picked up: 3980 seconds
Longest Hold = 61971 seconds
Number of Tests where NO particles were trapped: 9 (13% of tests)
Average Holding Time for ALL Tests = 3446 seconds
Test Summary and calculations:
21_june_2015_test_email.pdf
21june2015data.xlsx
Also here are pictures that were taken with the timed raster. The movement in the y direction (vertical) requires at most 380 ms to travel the distance so by setting the allotted time to 330 ms we can control exactly how long it will take to move the distance. In the code we can see that there is intended to have a single stripe horizontally down the center but at first we got Image 1 (left). In order to fix that we added 2 ms to the off time of the laser thus producing Image 2 (center). Here we can see the end of one full rastering (right end) and the beginning of the next rastering (left end) and how one is in the center but the other is on the top and bottom, this was solved by ending the full rastering with a system delay thus giving Image 3 (right). (Sorry for the poor quality of photos, the particles didn't want to hold long enough to get a really good photo, not sure why.)
19 June 2015
Jacob Van Wagoner
Last night we ran 1500 tests to see how the laser at 3 watts picked up. 314 picked up meaning it had a 20% pick up rate.
Also added a line to the code that keeps track of which tests hold and which do not. In this test we are able to see that overtime there are less particles that get picked up:
Using this excel file: 1962015pickupexcel.xlsx
Note: I wanted to see what the probability of success over time looked like, so I made a graph from this data. It is based on averages over 100 consecutive trials. As you can see, the pickup rate begins high and quickly drops to about 40%, then later to less than 10%. We could do similar graphs for the other tests to see if any varied from this pattern.
18 June 2015
Jacob Van Wagoner
Last night we ran 1500 tests to see how the laser at 3.5 watts picked up. 1002 picked up meaning it had a 66% pick up rate.
Jeremy Goodsell
Summary of a few articles and their findings.
https://www.osapublishing.org/ol/fulltext.cfm?uri=ol-19-22-1807&id=12676
cw Nd:YAG laser at a 1.064-μm wavelength gold (2 to 5 μm. in diameter), silver (2 or 3 μm), and bronze (2 to 15 μm). The gold and bronze particles were almost spherical. The silver particles were small and irregularly shaped. Because bronze particles tend to adhere to the surface of the sample cell, probably because of electrostatic forces, they were suspended in ethanol. Other particles were suspended in water.
https://www.physics.sfsu.edu/~laser/pdf/Zack_GaussianTrap_OE12.pdf
532 nm Verdi Laser found that absorbing particles such as carbon and silicon particles could be readily trapped by a focused Gaussian beam without the need of any special beam engineering. Importantly, we found that spherical particles cannot be stably trapped as compared to nonspherical particles. Once a particle is trapped, its position can be altered by changing either the laser power or the focusing condition of the beam. absorbing silicon particles of 5-20 µm sizes are used. The laser output power is about 1 Watt, and its beam diameter at the focal point is about 60 µm. A series of experiments shows that the silicon particles can be trapped easily, but the trapping is not very stable and is susceptible to ambient perturbations. It seems that the particles trapped after the focal point are more stable than those trapped before the focal poi
https://www.osapublishing.org/vjbo/fulltext.cfm?uri=oe-23-3-3630&id=311735
UV (at 351 nm) and the deep UV (at 244 nm), photophoretic trapping of a wide range of bio-aerosol particles is possible
spherical abberration trap graphite particles: 5 microns, 25-30mW carbon nanoparticles: 1-5mW could easily withstand air currents estimated at 10 − 50 cm/s depending on the focusing geometry and the type and size of the particles. For a fixed trapping power a tighter focusing provides a more stable trap because in this case the intensity minima are surrounded by regions with higher light intensity
http://iopscience.iop.org/2040-8986/14/5/055302/article
2 counterpropagating 532 nm Verdi Lasers at 1W, each with attenuator trapped graphite coated glass shells between 50 and 100 microns -particle transport speed greatest with horizontal polarization -particle trap stability greatest with vertical polarization
17 June 2015
Jacob Van Wagoner
Last night we ran 1000 tests to see how the laser at 4 watts picked up. 654 picked up meaning it had a 65.4% pick up rate.
15 June 2015
Jacob Van Wagoner
Data from the weekend's hold
Parameters:
Lens 1: 50mm
Lens 2: 75mm
Laser Power: 3.5 Watts
Dip Time: n/a
Vibrator: none
Filter: Aperture half closed/half open
Days since particles ground: 0
Total Tests = 5
Average Holding Time only for Tests where particle was picked up: 27001 seconds
Longest Hold = 54161 seconds
Number of Tests where NO particles were trapped: 0 (0% of tests)
Average Holding Time for ALL Tests = 27001 seconds
Test Summary:
13_june_2015_test_email.pdf
12 June 2015
Jacob Van Wagoner
Data from last night's hold
Parameters:
Lens 1: 50mm
Lens 2: 75mm
Laser Power: 4 Watts
Dip Time: n/a
Vibrator: none
Filter: Aperture half closed/half open
Days since particles ground: 0
Total Tests = 125
Average Holding Time only for Tests where particle was picked up: 488 seconds
Longest Hold = 5743 seconds
Number of Tests where NO particles were trapped: 22 (18% of tests)
Average Holding Time for ALL Tests = 457 seconds
Test Summary:
10_june_2015_test_email.pdf
9 June 2015
MatLab Approximated Fourier Transforms for Several of the Masks
5 June 2015
This was an important part of our research today.\
*like*
4 June 2015
Jacob Van Wagoner
Set up the R(G)B laser so that both the red and the blue can be used while rastering to color the image. The picture below is of the green laser rastering and of the blue and red laser alternating on and off.
The main coloring problem that we are having right now is that the red and blue do not fill the entire aperture so we are going to move it to the side and send them through a lens that will expand them to fill the whole aperture and thereby covering everywhere the green laser could be.
2 June 2015
Jacob Van Wagoner
Using the Raspberry Pi and the microscope camera we took the following pictures of the focal point with the filters printed on transparencies.
On the left is the picture of the mask and the right is the outcome:
No Filter:
Archimedes Spiral Single:
Archimedes Spiral Double:
Archimedes Spiral Reverse:
Neiod Spiral Single:
Neiod Spiral Double:
Neiod Spiral Reverse:
Log Spiral Reverse:
Inverse Root Mask:
Sum of Sines Gradient:
First Order Spiral:
Second Order Spiral:
Third Order Spiral:
Fourth Order Spiral:
Phase shift(???) Spiral:
1 June 2015
Jacob Van Wagoner
Testing using mirrors to attenuate laser while taking photos of the focal point. It was found that using optical density filters works much better, as shown below. The picture on the left is using a mirror. We see in that picture that there is a double image and that the image is not as clear as the image we see on the right, which was the same picture only using optical density filter instead of the mirrors.
29 May 2015
Jacob Van Wagoner
Last night we were able to get the green laser to raster and the blue light to flash to make images, below is one of the photos taken.
28 May 2015
Jacob Van Wagoner
We were able to take pictures of what the focal point looks like using the microscope webcam and dropping the power for the laser way down (by using an attenuater and optical density filters). These will be included with the report of the previous night's test results.
Data from last night's hold
Parameters:
Lens 1: 50mm
Lens 2: 75mm
Laser Power: 4 Watts
Dip Time: Every 3 minutes
Vibrator: Pulse for 1 second twice at 1.2 volts before pick up
Filter: Aperture half closed/half open
Days since particles ground: 0
(Particles should be dried daily due to the swamp coolers causing extra humidity)
Focal Point:
Total Tests = 200
Average Holding Time only for Tests where particle was picked up: 113 seconds
Median Holding Time only for Tests where particle was picked up: 183 seconds
Longest Hold = 184 seconds
Number of Particles Trapped for Complete 3 minutes: 7 (4% of tests)
Number of Tests where NO particles were trapped: 187 (93% of tests)
Average Holding Time for ALL Tests = 7 seconds
Median Holding Time for ALL Tests = 0 seconds
Number of Particles which fell toward the source: 4 (2% of tests)
Number of Particles which fell away from the source: 0 (0% of tests)
Number of Particles which fell due to an Anomaly: 2 (1% of tests)
Below is the Excel file used to find the above info. Histogram and plot are both in the Excel File.
Joel Rasmussen
This is the set up for the new intimidator which I am building. The three drivers are controlled by the Arduino Mega and can run simultaneously. The motors that are currently attached are run at an approximate voltage of 12V. The drivers have a full, half, quarter, eighth, and sixteenth step capability. In order for any given motor to run smoothly it must be driven by the correct voltage and have secure connections to the driver pins.
27 May 2015
Jacob Van Wagoner
Data from last night's hold
Parameters:
Lens 1: 50mm
Lens 2: 75mm
Laser Power: 4 Watts
Dip Time: Every 3 minutes
Vibrator: Pulse for 1 second twice at 1.2 volts before pick up
Filter: Aperture half closed/half open
Days since particles ground: 0 (However the particles somehow got wet so they all clumped and were moist
Total Tests = >200
Average Holding Time only for Tests where particle was picked up: 23 seconds
Median Holding Time only for Tests where particle was picked up: 24.5 seconds
Longest Hold = 40 seconds
Number of Particles Trapped for Complete 3 minutes: 0 (0% of tests)
Number of Tests where NO particles were trapped: >196 (>98% of tests)
Average Holding Time for ALL Tests = 0.5 seconds
Median Holding Time for ALL Tests = 0 seconds
Number of Particles which fell toward the source: 2 (<1% of tests)
Number of Particles which fell away from the source: 2 (<1% of tests)
Number of Particles which fell due to an Anomaly: 0 (0% of tests)
Below is the Excel file used to find the above info. Histogram and plot are both in the Excel File.
26 May 2015
Jeremy Goodsell
Masks for 3D surface spirals compressed onto density plot.
23 May 2015
Jacob Van Wagoner
Data from Thursday night's hold
Parameters:
Lens 1: 50mm
Lens 2: 125mm
Laser Power: 4 Watts
Dip Time: Every 3 minutes
Vibrator: Pulse for 1 second twice at 1.2 volts before pick up
Filter: Aperture half closed/half open
Days since particles ground: 0
Total Tests = 185
Average Holding Time only for Tests where particle was picked up: 54 seconds
Median Holding Time only for Tests where particle was picked up: 3 seconds
Longest Hold = 184 seconds
Number of Particles Trapped for Complete 3 minutes: 20 (11% of tests)
Number of Tests where NO particles were trapped: 73 (39% of tests)
Average Holding Time for ALL Tests = 32 seconds
Median Holding Time for ALL Tests = 3 seconds
Number of Particles which fell toward the source: 81 (44% of tests)
Number of Particles which fell away from the source: 8 (4% of tests)
Number of Particles which fell due to an Anomaly: 3 (2% of tests)
Below is the Excel file used to find the above info. Histogram and plot are both in the Excel File.
22 May 2015
Jacob Van Wagoner
Using the 405nm laser (UV) I was able to hold black liquor and Rhodamine G. The below picture is of thee black liquor, as the Rhodamine G looked the same (it was not fluorescing). The particle is the white dot above the nebulizer, it was captured using the fork dropping through the beam.
21 May 2015
Joel
Stepper Motor Specs
Attenuator Motor
200 Full Steps in 360 Degrees
3200 1/16 size steps in 360 Degrees
Run with BigEasyDriver at 20V
Make sure connections are sound or the motor will jump.
Jacob Van Wagoner
Data from last night's hold
Parameters:
Lens 1: 50mm
Lens 2: 125mm
Laser Power: 4 Watts
Dip Time: Every 3 minutes
Vibrator: Pulse for 1 second twice at 1.2 volts before pick up
Filter: Neiod Spiral Single
Days since particles ground: 1
Total Tests = 82
Average Holding Time only for Tests where particle was picked up: 30 seconds
Median Holding Time only for Tests where particle was picked up: 9 seconds
Longest Hold = 182 seconds
Number of Particles Trapped for Complete 3 minutes: 1 (1% of tests)
Number of Tests where NO particles were trapped: 56 (68% of tests)
Average Holding Time for ALL Tests = 9 seconds
Median Holding Time for ALL Tests = 0 seconds
Number of Particles which fell toward the source: 19 (23% of tests)
Number of Particles which fell away from the source: 3 (4% of tests)
Number of Particles which fell due to an Anomaly: 3 (4% of tests)
Below is a Plot of the holding time for the tests that held a particle. Note that each test is spaced by three minutes.
Below is a Histogram showing how many particles stayed trapped for a range of times.
Below is the Matlab code and excel file used to find the above info.
The poor performance of the filter may be due to the transparency being dirty. Another anomaly of note is that in these videos the particle “slipped” before falling out more so than in the other tests.
20 May 2015
Jacob Van Wagoner
Data from last night's hold
Parameters:
Lens 1: 50mm
Lens 2: 125mm
Laser Power: 4 Watts
Dip Time: Every 3 minutes
Vibrator: Pulse for 1 second twice at 1.2 volts before pick up
Filter: None
Note: I ground the black liquor so it was a fine powder again.
Total Tests = 85
Average Holding Time only for Tests where particle was picked up: 62 seconds
Median Holding Time only for Tests where particle was picked up: 22 seconds
Longest Hold = 184 seconds
Number of Particles Trapped for Complete 3 minutes: 12 (14% of tests)
Number of Tests where NO particles were trapped: 16 (19% of tests)
Average Holding Time for ALL Tests = 50 seconds
Median Holding Time for ALL Tests = 11 seconds
Number of Particles which fell toward the source: 44 (52% of tests)
Number of Particles which fell away from the source: 2 (2% of tests)
Number of Particles which fell due to an Anomaly: 11 (13% of tests)
Below is a Plot of the holding time for the tests that held a particle. Note that each test is spaced by three minutes.
Below is a Histogram showing how many particles stayed trapped for a range of times.
Below is the Matlab code used to find the above.
19 May 2015
Jacob Van Wagoner
Data from last night's hold
Parameters:
Lens 1: 50mm
Lens 2: 125mm
Laser Power: 5 Watts
Dip Time: Every 3 minutes
Vibrator: Constant at 1.2 volts
Filter: None
Total Tests = 82
Average Holding Time only for Tests where particle was picked up: 42 seconds
Median Holding Time only for Tests where particle was picked up: 5 seconds
Longest Hold = 180 seconds
Number of Particles Trapped for Complete 3 minutes: 3 (4% of tests)
Number of Tests where NO particles were trapped: 53 (65% of tests)
Average Holding Time for ALL Tests = 15 seconds
Median Holding Time for ALL Tests = 0 seconds
Number of Particles which fell toward the source: 20 (24% of tests)
Number of Particles which fell away from the source: 7 (9% of tests)
Number of Particles which fell due to an Anomaly: 4 (5% of tests)
Below is a Plot of the holding time for the tests that held a particle. Note that each test is spaced by three minutes.
Below is a Histogram showing how many particles stayed trapped for a range of times.
Below is the Matlab code used to find the above.
Jacob and Joel Most Recent File for Blue Rastering Method - Documents»Arduino»BlueRasterY2_0
15 May 2015
Jeremy Goodsell
Data from last night's hold
Parameters:
Lens 1: 50mm
Lens 2: 125mm
Laser Power: 4 Watts
Dip Time: Every 5 minutes
Vibrator: Constant at 2 volts
Filter: None
Total Tests=52
Average Holding Time only for Tests where particle was picked up: 72 seconds
Median Holding Time only for Tests where particle was picked up: 17 seconds
Longest Hold=5 minutes
Number of Particles Trapped for Complete 5 minutes: 4 (8% of tests)
Number of Tests where NO particles were trapped: 12 (23% of tests)
Average Holding Time for ALL Tests=54 seconds
Median Holding Time for ALL Tests=8 seconds
Number of Particles which fell toward the source: 25 (48% of tests)
Number of Particles which fell away from the source: 8 (15% of tests)
Number of Particles which fell due to an Anomaly: 3 (6% of tests)
Below is a Plot of the holding time for each test. The bottom axis is the test number. Note that each test is spaced by five minutes. The vertical axis is the maximum time at least one particle was held during that test.
Below is a Histogram showing how many particles stayed trapped for a range of times.
14 May 2015
Jeremy Goodsell
Data from last night's hold
Parameters:
Lens 1: 50mm
Lens 2: 125mm
Laser Power: 4 Watts
Dip Time: Every 5 minutes
Vibrator: Constant at 1.2 volts
Filter: None
Total Tests=51
Average Holding Time only for Tests where particle was picked up: 103 seconds
Median Holding Time only for Tests where particle was picked up: 46.5 seconds
Longest Hold=5 minutes
Number of Particles Trapped for Complete 5 minutes: 7
Number of Tests where NO particles were trapped: 15
Average Holding Time for ALL Tests=73 seconds
Median Holding Time for ALL Tests=21 seconds
Number of Particles which fell toward the source: 23
Number of Particles which fell away from the source: 4
Number of Particles which fell due to an Anomaly: 2
Below is a Plot of the holding time for each test. The bottom axis is the test number. Note that each test is spaced by five minutes. The vertical axis is the maximum time at least one particle was held during that test.
Below is a Histogram showing how many particles stayed trapped for a range of times.
13 May 2015
Joel and Jacob
Achieved juxtaposition and simultaneous functionality of arduino control and raspberry pi systems. Raspberry pi will send a command to the arduino to pick up a particle and hold it in a specific position while the raspberry pi takes pictures. Once the particle has fallen from view the pi will send a command to retrieve another particle. All parts of the system work except for the correct detection of a picture. The control picture which is supposed to be complete blackness, is brighter than the pictures when a particle is being held. This will give a false positive and when there is not a particle and a vice versa.
The rest of the system performs very well and the two boards communicate through a new cable which we soldered and hung as well. We also have additional wires prepared if there are any other signals that need to be sent between the boards/vibrators etc.
12 May 2015
Jacob Van Wagoner
Because the z axis on the current motors have more accurate abilities, and Princess Leia is projected by R2D2 as such, we decided to use a vertical raster.
The image is very white because there was no filter, will use the makeshift filter for future shots until we get an actual green filter. As we can see the lines are very close and will contribute to a potentially higher quality photo. The only potential problem I see is that when turning the blue laser on and off it moves the particles being held by the green laser. In the picture below you can see where turning on the blue laser knocked the particle out of the hold of the green laser (lasers are coming from the right going to the left).
Used the makeshift filter to take a picture of the Y. We see that the idea of holding with the green laser and coloring with the blue laser is possible and can look really good.
Jeremy Goodsell
The following are several mathematica renderings of common spirals. For each, parametric plots using just one spiral, 2 spirals shifted by a phase, and two spirals going in reverse are displayed. Testing results for effect on holding time of a particle are still pending.
11 May 2015
Jacob Van Wagoner
We want to rastor a particle with the green laser and flash a blue laser to make an image. The blue laser was sent along the same path as the green laser as shown in the schematic:
currently the blue and green lasers are not perfectly aligned, as seen below:
But this can can be fixed either by more accurately aligning the two or by passing the beams through an aperture of sorts. Below we see two particles, the one to the left is reflecting the blue light. The reason the right particle is not reflecting the blue is shown in the picture above (the beams are not perfectly aligned).
Not rastered.
The green laser was at 1.5 watts and the blue laser was running at 12 V. Pictures taken were taken with the cannon camera through a set up using a dichroic mirror and a silver mirror to reflect the blue and filter out the green as best as possible.
Makeshift filter:
Joel Rasmussen
Developed code to automate the vibrator on the particle tray. This will vibrate once for each image that is being drawn. At the beginning of the loop before the particle is retrieved from the tray it will vibrate for one second, this will allow the particles to be pick up more easily.
Details: AO analog in put pin is attached to 3.3 V pin on arduino and multiplied by .38 to achieve the desired 1.2 volts for the vibrator to run. This is given as an output voltage on pin 13. and transmitted through the orange wire to the vibrator.
If needed we can adapt the code easily to vibrate for longer or shorter amounts of time. Code is implemented only in the SquareY drawing right now. Documents» Arduino» SquareY_YZTrace
Kevin Costner
Code that I have been working on. It's written in python for a raspberry pi.
8 May 2015 Joel and Jacob
Achieved automation of particle pick up using a stand to which the laser drops to pick up a particle. A buzzer is attached to the stand to move particles into desired location to ensure a particle is picked up each time. We also flipped the Y drawn previously right side up. We will code the arduino to run the buzzer each time that we go to retrieve a new particle, it is important not to have it vibrating while trying to retrieve a particle though because this will often knock the particle our before it is held.
We determined the proper positioning of the axis to ensure that we do not turn the laser toward where is can be harmful. We had previously used the opposite axis.
Jeremy and Jacob
We also ran more tests on the butterfly, iron man, and “Y” points to see if they improved at all. Only the “Y” turned out well.
Below are what the patterns for the iron man repulsor and butterfly should look like ideally.
7 May 2015 Jacob and Jeremy
We created an array to draw the particle back and forth in the yz plane so that we can later use this array and choose certain points to illuminate with a blue laser as an alternate way of drawing images.
This first image used a delay time of .5 seconds between points which was not long enough to for the z-axis to finish running from point to point.
This second image we corrected to use a delay time of 1 second allowing the z-axis to complete its full extension and retraction on each sweep.
We also created our first attempt at a butterfly. We discovered that y-axis of the image may not have been centered around the correct coordinate which may have led to the disfiguration visible in the image below.
6 May 2015
Jacob Van Wagoner
Adjusting camera:
We were able to make the code to draw a “Y” and drew it.
Document»Arduino»SquareY_YZTrace
There are some things that need to be looked at and understood in the code. Specifically why the “Y” needs to change from point 12 to 13 (first and second picture below), the “Y” is upside down (all pictures), the stem of the “Y” needed to be extended because of scaling reasons (second picture below).
Joel Rasmussen
We have additional work to do when it comes to moving in two axis at the same time since they move at different speeds. The z-axis moves much faster than the y-axis does which distorts images when tracing a diagonal line. This will probably need to be corrected using independent loops which are calibrated to run simultaneously and arrive at their destination points at the same time. The square y is our simple solution which we turned to when a diagonal line wasn't yet feasible. The diagonal line code which still has issues is found at =Documents»Arduino»Y_YZTrace
Jeremy Van Wagoner
Reran the correction code to a straight line.
Documents»Arduino»LineCorrection
Below is an image of an uncorrected line:
The below picture suggests the change in the z-axis is differently scaled. (1 cm in the y-axis is roughly 13 points whereas in the z-axis 1 cm is roughly 143 points.)
As such the correction code needs to be scaled accordingly. Below is an image where the corrections we scaled 1:10.
Also started to look into the Fine tilt/Fine pan vs. Tilt/pan. When the pins are switched in the code there is no response. Need to look into this further for smoother movement.
In order to make a 3D printed case (R2D2) for the intimidator the measurements were taken, shown below:
Jeremy Goodsell
This is the mask which can be printed on a transparency to then place before the lens to create a deep potential well at the focal point.
5 May 2015
Jeremy Goodsell
Below is the link to the google doc pdf containing a Mathematica PDF which shows a function which, when applied spatially over a mask, should create a deep potential well as shown in the document.
Jacob Van Wagoner
Tested straight line code to see if the correction code helped. There was not an obvious difference between the two. We can however see the three points the laser stopped at. Picture 1 is without the correcting code:
Picture 2 is with the correcting code:
This helped us realize that we should have been decrementing our z-axis when we were actually incrementing it.
4 May 2015
Joel and Jacob
Camera was positioned allow the picture to be taken without too much reflection from other materials. This will permit us to take pictures of the images being drawn without interference from the light reflecting off the mirror.The camera mount was secured to the table in a fixed position. We will need to adjust the position to make sure we are able to take the pictures we need.
We are using the following Arduino files to trace out a square with specific points. Upon completion of the algorithm mentioned above we will be able to produce a square with rectified lines.
Documents»Arduino»Square_Trace_With_Z_Correction
The other code we have used is found at the following locations:
Documents»Arduino»_3DSquare
Documents»Arduino»SimpleXYSquare
One of the difficulties we have found was getting the camera to the right exposure setting and length of time needed to capture the traced image. With increased exposure time the images became more difficult to see due to the light pollution of the image. We also noted that if more than one particle is trapped with the beam it will affect the quality of the image. Below is shown what happened when multiple particles were trapped using the _3DSquare Code.
The following picture is what the SimpleXYSquare produced. As can be seen there is little curvature in the lines that make up the square.
Specifications of this Capture:
Exposure time: 13 seconds
1st Lens: 50 mm focal length
2nd Lens: 125 mm focal length
Laser Power: 4 W
Camera Position: Opposite end, middle of box, right hand side(Facing from R2D2).
Enclosed area to limit air movement.
We also tried to trap a particle without the enclosure but found that this was more difficult for it to hold for a long enough period of time.
We found that the particle was not only able to travel a significant distance despite the odd configuration of the lenses but we were also able to sustain the particle for a maximum time of 6 minutes.
We noticed in addition to a shape being drawn that there were different intensities of light based on the position from which the particle was viewed. This demonstrates, in part, ambient occlusion.
Below is the current setup of the Intimidator. The laser comes from what is depicted as the left and passes through Lens 1 (1st Lens above) then reflects off the lower mirror (at bottom of chassis) and then passes through Lens 2 (2nd Lens above) and bounces off the upper mirror then finishes focusing after reflecting off the mirror.
1 May 2015
Jeremy Goodsell
Below is the link to the google doc pdf containing Mathematica code for how to determine truly straight lines in the xy plane or how to determine straight lines in either the xz or yz planes.
30 April 2015
Jacob Van Wagoner
Encasing for Intimidator (20x12x22 inches) was constructed and holes for viewing, power cords and laser entry were cut into the box. Next step is to cut view hole larger and cover with the absorption sheets and to cut a hole for the camera to take long exposure photos.
Due to lack of sliding mirror we are currently placing a mirror in front of the other mirror being used by the testing/automated part of the Leia group to redirect the laser to the Intimidator. Because the Intimidator is lower than the laser (can't raise the Intimidator) the first mirror lowers the beam and the second mirror levels it out. Then the laser passes through Lens 1 (f=50mm), enters the Intimidator and then passes through Lens 2 (f=~125mm). With this set up we were able to hold a particle with the verdi laser (set to 4 Watts) and draw a cube.
29 Apr 2015
Joel Rasmussen
Optimized and debugged cube code. Created functional setup to levitate particles. Max Time: 3 seconds. Lenses being used are the following: Lens 1: Lens 2: Purchased Material for enclosure and worked towards understanding of milestones.
Goals: Increase hold time by optimizing setup and lens choices. Construct a complete enclosure, acquire necessary materials for automation.
