︎

   Projects   Info   CV

DIY Wireless Optogenetics


This accompanies and supports wireless optogenetic implant fabrication as detailed in Montgomery KL, Yeh AJ, Ho JS, et al. Wirelessly powered, fully internal optogenetics for brain, spinal and peripheral circuits in mice. Nature Methods. 2015;12(10):969-974. doi:10.1038/nmeth.3536.

Originally published in November of 2015 by the lab of Ada Poon, Integrated Biomedical Systems Group at Stanford University. Last updated in June of 2016 by Derek Crowe (derek.crowe@urmc.rochester.edu) while working in the Minsoo Kim lab at the University of Rochester. A copy of the original DIY document published by the Poon group can be found here.


Figure 1: System Overview



MATERIALS


A complete parts list with source links and price estimates has been compiled for your use. Feel free to comment on it with suggestions for better sources, cheaper prices, questions, or changes. I suggest you create a copy on your local account and share it with your purchaser. Communicate the status of parts via the color key. The Poon Group resource page with CAD and Gerber files and their assembly videos can be found here. (If for whatever reason that page is dead, I uploaded a local copy of those files on 6/26/18. The assembly videos are on YouTube.)



ASSEMBLY INSTRUCTIONS AND NOTES



Signal Generation

  • The signal is generated, amplified, divided, phase shifted, and delivered to the cavity. The general layout is outlined above.
  • You’ll need to run the software through Windows.
  • As of May 2016 the Windfreak signal generator seemed to be less stable than the Vaunix solution. For the price, it was good enough for our application. If you have more stringent power delivery requirements, you’ll want to invest in the Vaunix.
  • The cables I listed are short and stiff. They are also the cheapest. If you want a bit more flexibility, get thinner or longer wires.

Power Amplification

  • Tutorial for soldering Tamiya connectors
  • Heat shrink is more elegant than electrical tape for sealing off the connections to the power amp.
  • The power amp will get warm pretty quickly. Make sure you’re cooling it down with a fan if you’re running it for a significantly-lengthy experiment.

Cavity

  • Speak with someone at the Poon lab before you send the linked files for fabrication. They may have an updated design that’s cheaper to manufacture and easier to tune.
  • You may get push-back from the folks fabricating the cavity about the top and middle interfering. The files the lab provided are accurate; proceed without editing them. The top is made to have a slight bevel that will touch the middle and create a strong connection when tightened. Do not remove the top once you’ve tightened it down the first time. Access the cavity from the bottom if you really need to get in.

Tuning

  • If you have access to a HF network analyzer (any EE dept. should probably have one) you don’t really need the power sensor. They both accomplish the same thing.
  • If you have access to a network analyzer AND you know how to use it, you win. In short, you need to optimize S-Parameters at your particular frequency of interest. If you’re replicating exactly this situation, that’s 1.497 GHz.

Implant

  • The Poon Group’s soldering video is a great place to start.
  • A circuit diagram and component layout schematic is included in the figures below.
  • Have the manufacturer add a solder mask on your PCB boards. This will save a step.
  • Use as little solder paste as possible. You will be surprised how far a few beads go. The solder paste on the parts list is specific. We’ve tried others and they are drier and less capable of holding the components in place before soldering. Store it in the fridge and heat it up in a water bath before you use it.
  • Place the components on the pcb from left to right. I find the middle pad the most difficult to access when you do go to solder (if you know how to edit .gbr files, you could add more pad to the top for solder tip) but it’s not too bad once you get the hang of it.
  • Solder all the components EXCEPT the 10nF loading cap, make sure they are stable, THEN place and solder it.
  • The coil wire ends need to be stripped with a knife (or a 34 AWG wire stripper) and pre-soldered before attempting to solder them.
  • Once the coil is attached, check to see that the circuit is powered with an antenna, which you can construct with the end launch solder attachment. Solder wires to the middle and the side. If you have a network analyzer, you can just clip off the wires until their lengths optimize the signal at your frequency. If you’re using 1.497 GHz, make them 4.7cm long. See the image of the antenna, below.
  • Flip the implant over and solder wires to the pads. You’ll attach the leads of an oscilloscope or voltmeter to these and power the antenna. If you get a change in voltage, you win. Take care to also watch the change in voltage that happens to the system when the wires are NOT attached to the leads. The antenna will induce power in the wires without the circuit.
  • Once you have confirmed that you have a working circuit, fill the coil with enough epoxy to coat the board and the components and wait 24hrs.
  • After 24hrs, pre-solder the LED leads and place them as shown on the other side of the implant.
  • Put it on the cavity and power it up, if it turns on, you win. Keep in mind that this implant is optimized to function in tissue. You may need to hold it in your fingers to couple enough energy. You may also need to try to move it around. The orientation of the coil and its location above the cavity are all variables in its ability to couple energy.
  • I made about 7 of these before I got one to work.

Surgery

  • Follow the protocol for your particular situation
  • TIPS: only clear fascia where the implant will be placed, use lint rollers to pick up shaved hair, silk sutures. 


Figure 2: Implant Layout and Circuit Diagram



Figure 3: Antenna for checking implant circuit