THP Semifinal Video
Below is the transcript for this video... Sorry it runs a little fast, but there's a ton of information!! =)
ramanPi started as a project to fill a need I had for another project I am still working on. I had to obtain the bond angle and raman shifts for samples I was producing. That called for a raman spectrometer and access to one is financially prohibitive as even used units run tens of thousands of dollars. I set out to build a raman spectrometer with no knowledge on the subject whatsoever.
Since I started the design has gone through four phases.. I started with the misconception I was going to be able to obtain a raman signal with a more rudimentary setup..I then moved to the thought of using the raspberryPi camera module and a couple edge filters. That notion might be fine with some extra time..I progressed into a series of lens mounts on stands in an optical bench type configuration.Some feedback from the hackaday crowd, and a couple late nights, and it evolved into what it is today..
ramanPi is fully open source, there is absolutely no secret sauce. Everything is available to everyone. I also used several open source libraries for the various sensors and other devices the system is composed of. There are also many open source libraries used in both the raspberryPi software as well as the client software.
I had no idea I would take the design so far when I first thought about this project. And I look at what I have now, and wow.. I think the 3D printed parts came out just amazing. I am very happy with the spectrometer design. The fact that I was able to write the .SCAD file to automatically change dimensions, shape, and angle for whichever optics you decide to use is way more than I ever hoped possible!
ramanPi is centered around a raspberryPi, which is connected to the internet via a WiFi adapter. It communicates with the client software as well as multiple online spectral databases to determine a match for the spectra of the compound under test. The client software and the raspberryPi exchange data which is displayed on the users workstation, phone or tablet. The client software also gives the user the ability to store that data in the cloud, share it with friends or colleagues and make graphs or presentations with it.! The client software can use whatever internet connection the device has. Without an internet connection, ramanPi wouldn’t do much beyond display spectra from a sample on a local screen. It would not be able to identify what compound you are testing.
There are a number of other uses for ramanPi as a whole as well as many components within the device. The entire system has many applications such as mineral identification, cancer research, biological studies, and a ton of others.. You can extend that by adding future expansions I have planned, such as fiber optics adapters, side scatter detection, and a few others! The spectrometer portion can be used completely independently of the whole system. Without modification,it can be connected to a PC through USB without the rest of the system and work as a fully functional spectrometer! To top that, you can also add future adapters to the spectrometer for many additional features such as fiber optics.
The design can be reproduced very easily. For high volume production you could use injection molding to make the 3D parts. 3D printing for lower quantities would be a totally acceptable option. The optics are right off the shelf and can be purchased individually or in bulk from companies like Edmund Optics. The electronics are very easy to source and they will be integrated down to two boards for the whole system very soon. The CCD is inexpensive as well and is the same version commercial systems use. It is a complex device, but there is really not much to it in this sense.
I made every engineering choice with the idea of reproducibility in mind. I kept the design as simple as possible while keeping the integrity of a raman spectrometer with a respectable spectral resolution. I also made sure the design was physically sturdy and would withstand the test of time.
I kept the parts modular, simple and in a format that is fairly commonly used in the industry. The individual parts are easy to work with and the overall cost is unbelievably low in comparison to the alternative.
The optical assembly and 3D parts are the showcase of this system in respect to ingenuity and innovation. This is a one of a kind system and the only DIY raman spectrometer design available, let alone open source. I have received a lot of positive feedback from many people who have tried to create a raman system with no success.. I hope to not let them down!
The modularity, ability to modify and tailor the parts to suit your needs, the use of two inexpensive edge filters as opposed to one expensive notch filter all are testaments to the innovation. The spectrometer has been called a work of art. There’s really nothing else like it. Not even commercial systems have this many features!
I made sure to design every part of this device so my own mother could use it without wondering what means what. The interface is very simple to understand and very easy to get up and running spectra right away! You don’t even need to interpret a squiggly line and figure out what compound it is, the system will verbally speak it to you when it has identified it! The client software should be available on a PC to start, then eventually phones, tablets and even chromebooks!
The main electronics consists of a control board, the power control board, an interface board, the imaging board, and the raspberryPi. The control, interface, and imaging boards are based on ST Micro STM32 Nucleo F401RE microcontrollers. The control board handles the nuts and bolts of the system, moving motors and reading sensors and so on. The interface board controls the display, the touch panel, audio and the RGB LEDs. The imaging board manages the CCD and reading spectra from the spectrometer. All the boards are connected to the raspberryPi via USB and their functions are dependent upon one another. The raspberryPi communicates to the internet, databases, and the client(s).
Currently, the control board functionally monitors the laser and cuvette temperatures via two DS18B20 sensors, controls the L298 HBridge which drives the two peltier devices. One peltier cools the ccd and the other both cools and heats the cuvette to maintain its temperature using a PID controller. It also monitors the current draw for the two peltiers. In addition, it reads the cuvette ends stops, controls the laser shutter servo, and both the TTL line and power relay for the laser. Priorities have kept the laser good TEMT6000 sensor, the BMP180, and the laser color sensor on the back burner.
The power control board currently performs all of its planned duties which include taking in power from the mini itx power supply, supplying power to all of the other boards, housing the L298 HBridge, the current sensor and controls the system power with a nice stainless steel button.
The interface board also currently performs all of its planned duties controlling the TFT LCD display, and taking input from the touch sensitive keypad..it also communicates that data to the raspberryPi when it has any user input. In addition it controls the arduino pro mini that drives the RGB LED ring.
The imaging board is another success story with it’s duties of controlling the CCD detector array and obtaining spectra which it then sends to the raspberryPi or PC via usb. In the future, it will also monitor the UV index for detection of noise or fluorescence. It will also monitor the temperature of the CCD to compensate for noise and move the CCD tiny fractions of a millimeter from side to side to increase resolution.
The optics section is a great success. All of the plastic parts fit perfectly, the unit is nice and solid.. The laser shines, the beam bounces from the laser and hits the splitter..half the beam goes to the objective lens and the other half goes to the filters..the selector wheel spins and chooses the filter...the beam goes through and into the spectrometer which images the spectra… I couldn’t be happier, it just needs a little alignment and calibration!
At this time, I have focused mostly on the design and construction of the electronics and optics. The software that currently exists is mostly for testing, I will be devoting as much time to the software as I have the hardware in the very near future as the hardware is getting very close to being complete!
Overall, this has been a terrific learning experience. Hackaday has been just great in its efforts and encouraging participation.. I would have been building this whether the contest existed or not, but being a part of it has made the experience far richer than I had imagined it could be. Thanks to everyone who has given me so much positive feedback the past couple months!!