Saturday, February 8, 2014

NIR Stimulator Build


Most people are familiar with the terms hard science and soft science. A hard science is one which is subject to rigorous quantitative statistical analysis whereas a soft science more often deals with qualitative intangibles. This “soft to hard” spectrum wasn't developed as a criticism. In fact, the developer of the idea, Auguste Comte is considered the founder of sociology which in terms of scientific hardness is found just above “paranormal investigation” and just below talc. Another rather maligned science when viewed in terms of hardness is psychology. Psychology often deals with totally qualitative factors such as mood and mind; however, this isn't to say that soft sciences don't contribute to humanity. In fact, many of the social ills we experience are due to policies ignoring the findings of sociology, psychology, and criminology. Furthermore, the hardness of a science varies depending on the particular area of the umbrella under which one stands. Sensory psychology studies phenomena which can be entirely expressed quantitatively and thus stands as hard a science as biology.
In the 1930's, research by Wolfgang Metzger confirmed that hallucinations and electroencephalogram changes consistently occur when a person is subjected to a featureless field of vision. This study explained and confirmed the hallucinations reported by Arctic explorers, trapped miners, and those placed in sensory deprivation. The use of an EEG provides reliable quantitative data which confirms the study as being worthy of the label hard. To produce a featureless field, Metzger used a Ganzfeld (German for “complete field”) stimulator, which provides a uniform and featureless field of color. This design has been further developed into multi-modal devices such as the “mind machine” which uses a flickering Ganzfeld along with binaural beats as a means to produce a drug free altered state of consciousness.
Ganzfeld stimulators are also still used to study the electrical activity of the eye. In electroretinography, the response of the retina is measured while being subjected to a stimulus supplied by a Ganzfeld stimulator. Unfortunately, for our purposes a normal Ganzfeld stimulator simply isn't adequate. The electrical response of retinal cells to the stimulation of light are nearly uniform regardless of the wavelength used. In order for our ERG data to be meaningful, and demonstrate that the retina has developed sensitivity to previously undetectable near infrared light, a new type of Ganzfeld stimulator had to be invented and built. I present to you, the Gibson Rift.
OK, so forgive my momentary lapse into grandiosity. The reality is that the design is rather simple. It isn't much more than a few infrared LEDs blinking away inside crudely made fiberglass goggles which function to block out ambient light. The device wasn't actually named by any member of team Science for the Masses either. It was deemed the Gibson Rift by prominent biohackers Amal Graafstra and Rich Lee at a transhumanist conference and the name just stuck. Going into this project, I had nearly no knowledge regarding fiberglass fabrication or electronics and hit a number of roadblocks. My hope is that this post helps others to avoid such pitfalls should they want to repeat our research.
Construction of the Frame
CONCORD DIGITAL STILL CAMERAThe frame was initially modeled in floral foam. The advantage of floral foam is that it's easy to work with and it doesn't break down from the solvents in fiberglass resin. The foam pieces can be glued together using a hotglue gun. As long as the person for whom the device is being constructed is present, it's easy to cut and glue according to the size of that person's head. It was rather more laborious to construct the frames from measurements, but if need be the ratio of head width to length tends towards 2:3. So if a person's head is measured from front to back as being 9 inches, the width tends to be around 6 inches. This approximation allows a person to extrapolate size from circumference. Don't put too much effort into shaping the foam as features will be lost underneath the fiberglass and you'll just need to shape them again later.
After roughing out the foam, apply a layer of resin followed with a layer of fiberglass cloth and another layer of resin. Fiberglass supplies can be purchased from nearly any hardware store. I advise purchasing an extra container of hardener and mixing the resin “hot” meaning with a larger proportion of hardener. It's much easier to sand off a mistake that hardened too soon than it is to fix the wet droopy mess that results from too "cold" of a resin. Fiberglass resin hardens much faster at higher temperatures and under UV light, so this isn't a rainy winter day project. After the initial coat hardens, grind and cut out features such as the anterior space which holds the electronics, the eyecup which holds the IR filter lens and the IR LEDs, and the mounting bolt for the straps.
CONCORD DIGITAL STILL CAMERAWhen people fix automotive projects with fiberglass, Bondo is a BIG time saver. Because our project is so small though, I found that an airdryable lightweight modeling medium works much better. Don't worry about small mistakes such as areas with minimal resin, cracks, and places where you sanded the material too thin. Fill in mistakes and shape out features using this airdry clay and then apply another layer of fiberglass and resin. Be far more cautious when applying this layer though, as it will be the final outer layer. Using long strips of cloth rather than many smaller bits is also crucial as this is what will provide a durable final product.
CONCORD DIGITAL STILL CAMERATo mount straps onto the frame, I used carriage bolts which, again, can be purchased from nearly any hardware store. Carriage bolt have a long area sans threads, so once installed and glued into the frame, you're left with a nice smooth surface for the straps to rest on. I first drilled appropriately sized holes into the frame on the left and right posterior regions and through the portions of the frame which extend out of the superior surface. I used an epoxy glue to secure the bolts and then used a grinder to remove the portions of the bolts which extended beyond the desired size of the frame.
CONCORD DIGITAL STILL CAMERAFor the anterior box, I chose an L arranged conduit body. This is intended to be used for industrial and home wiring as an area in which one can splice wires. It allows for easy access and installation of our electronic guts. Prior to attaching this piece, I drilled appropriately sized holes along the superior surface so that I could later install the switches that control the IR LEDs.I hot glued the conduit body to the anterior of the device with one of the tubes passing through the portion that will serve as our right eye piece. I mounted a project box at the end of the other tube and drilled a hole so that wires could be passed through.
After attaching these peripherals, fill in any gaps between that part and the body with more airdry clay, and apply your final resin coat. If you want it to look pretty, feel free to spray on some color and a layer of clear gloss coat. The last peripheral portion to be attached is the IR filter lens, but this shouldn't be done until the electronic portion has been installed as it will prevent you from being able to arrange the LEDs.
Circuit and Switches
FVE3Z48FCTI5PC4.LARGETo drive our LEDs, I found a number of different tutorials using an 8 pin 555 IC. There are a vast number of “instructables” on the subject and damn near everyone is different. This image shows the circuit I settled on. This post isn't intended to teach people to solder or work with electronics, but I'll tell you lessons that I learned: Solder isn't meant to physically hold things together. Secure your wires and pins mechanically first. Not a single one of those conductive pastes or pens is worth a damn, so don't waste your money. And finally, it's totally worth your while to start with a breadboard and make a mock up first. I promise it will save you from later frustrations.
CONCORD DIGITAL STILL CAMERAInstead of directly mounting the LEDs to the board or soldering leads running off the board, just use terminal blocks. Make a second simple board on which to mount the LEDs and then add the wires between later. I know everything I'm saying is remedial to those who are into circuit building, but no one ever gave me this advice, and I earned the knowledge through hours of frustration. And finally, because of the size constraints of the boxes on the front of the device, you'll need to make sure your circuit board is going to fit prior to actually assembling it. A very strange thing to me, is that the sizes of project enclosures available at Radioshack, don't match any of the blank circuit boards. I ground the circuit boards down to the appropriate size prior to assembly and it worked fine.
This 8 pin 555 IC circuit results in LEDs which flash at a set rate. The reason that you want the LEDs to flash rather than simply glow, is that we are looking for an A/B signal response in our ERG data. If we find A/B fluctuations which match the flash rate of the NIR LEDs, this confirms that our retinal cells have gained sensitivity.
After assembly, simply mount your driver circuit in the project box, the switches into the holes along the conduit box, and the LED board should fit right into the tubing anterior to the eye piece.
CONCORD DIGITAL STILL CAMERAThe LEDs used in our project have peak emissions at 850nm, 950nm, 1070nm, and 1200nm. To test the LEDs, you unfortunately won't be able to use your bare eyes, except for the 850nm. Although 850nm is beyond the range of normal human vision, LEDs emit light within a narrow wavelength range. While the lowest wavelength LED may peak at 850nm, the lower range of its emission pattern does extend down into the visual range. This actually works to our benefit, in terms of testing, in that the final LED to be activated serves as a control from which we can determine the expected retinal response. If and when our retinal cells gain sensitivity into the ranges of the other NIR LEDs, they will elicit a similar ERG signal. In order to test the 950nm, 1070nm, and 1200nm LEDs one can use a digital camera. The majority of cheap digital camera can pick up wavelengths into the NIR. Smartphone cameras will also suffice. After attaching your electronic guts, hot glue the IR filter lens in front of  the LEDs immediately before where your eye will rest.
Foam Padding and Straps
Strangely enough, this was one of the more difficult issues to solve. A compressible foam layer that blocks light is critical. I tried everything from carving silicone to felt covered foam, but nothing seemed adequate until I started
With this crude fiberglass, I will SEAR MY EYEBALLS WITH LIGHT IN THE NAME OF SCIENCE!
reading blogs by people who are into cosplay. People build these huge elaborate suits ranging from knight's armor to giant Teenage Mutant Ninja Turtle heads. Furthermore, they build these things to sustain some serious beating albeit with foam ninja swords. The best solution I found was to cut and glue together pieces of EVA foam to build up the eye pad shape. Then, alternate between spraying on layers of 3M adhesive and Plasti Dip. It really does take a rather surprising number of coats. The first three coats end up being absorbed by the foam and it isn't until coat 5 or 6 that the eye piece really starts to look promising. By coat ten or so, you end up with a pliable soft rubbery eye piece that suits our purposes for blocking out ambient light with some degree of comfort. Hot glue this eyepiece into the frame and your device is nearly ready.
After the previous steps, sewing the straps together will be easy as pie. I used 1'' black nylon straps and arranged them with one that goes over the head and attaches to the strap that wraps around the sides of the head. I used two buckles to make adjustment easy. Once attached, your “Gibson Rift” is complete.
Overall, construction of our stimulators was an educational experience for me. Because we knew going into this process, that there was bound to be a considerable amount of waste from failed prototypes, we chose to fund construction primarily with our own funds rather than the donated funds. The one exception to this was the 1070nm and 1200nm NIR infrared LEDs which at 26$ a piece were far costlier than originally expected. These devices have been tested and function as expected and will fulfill the needs of our project. Furthermore, the experience has lead to skills in both electronics and fiberglass fabrication which will likely be of great use in our future Science for the Masses endeavors.

1 comment: