- Risks
- FPGA availability
- FPGA chips were difficult to find, so we plan to switch to using a separate FPGA dev board or a microcontroller
- Dev board may have latency issues because high-speed communication will need to be sent back and forth over wires to a separate PCB
- Microcontroller speed/data transmission also has a latency issue, because it’s slower than FPGA
- MCU speed: with the MSP430 that we’ve used before, 25MHz is the fastest option.
- We could also use 2 microcontrollers to mitigate the speed issue, and use 1 per laser
- Plan: test out both the microcontroller and FPGA dev board in a separate setting to choose which option
- Component availability between early stage of design and component ordering
- Some components are low stock. We will order components once we are confident we will be using rather than ordering once our pcb design is released
- Lasers may not meet speed requirements
- If not all components that we selected work, we will swap to a single laser setup using the one that works. We are confident that green will work because it has specs for modulation frequency, but the IR laser does not. We plan on testing this next week.
- FPGA availability
- Design Changes
- Change to FPGA
- As described above, we were unable to find a FPGA to put on our custom PCB, so we will need to either use a FPGA dev board or MSP430 microcontroller.
- Not using sinusoidal modulation
- We will instead send a square signal over the laser. This is because fiber optic laser drivers do this as well, so there is no commercially available IC to modulate a sinusoid with enough current to power our laser.
- Not use fiber optic laser driver
- After creating a partial schematic and selecting support components for fiber optic drivers, it became clear that our application has a large number of changes required to manufacturer suggestions in order to work. Our project will be easier and more likely to succeed if we just create the driver circuits discretely rather than using a modified fiber optic driver circuit.
- Switch to 2 lasers: IR and green
- We found that blue lasers are very expensive and there is only one with a low enough power rating for us, and it is most likely too slow. As a result, we switched to 2 lasers. Changing red to IR makes the gap in wavelength larger, so we can use commercially available photodiodes that only receive light at certain wavelengths to filter out the other color.
- Considering adding a higher voltage power supply
- The green laser has a forward voltage over 5V. We are concerned about the effect that a boost converter will have on signal integrity, and do not consider our power supply to be the important part of our system.
- Change to FPGA
- Our project includes considerations for safety and security
- Safety: The lasers must be of low enough power to not injure the eyes of users and bystanders
- Security: Our use case concerns itself with the secure transfer of private information. This method should ensure the private and secure transmission of data with minimal chance for remote interception.
- Schedule Changes
- Because we are unsure if we’re using FPGA now, Anju will test out and set up environments for microcontrollers and FPGA to test out the parts for next week.
- Photos
- Starting to set up FGPA to test out compilation
- Starting to set up FGPA to test out compilation
Anju Ito, Roger Lacson, and KJ Newman