Stimulator design. a. A fully assembled stimulator. b. Rendering of the custom-printed circuit board which accommodates the microcontroller, the LED driver and up to 24 LED channels. c. Schematics illustrating the circuit that controls the LED output. The blanking input can be inverted by a switch before reaching the output enable pin on the LED driver (electronically switching off the LEDs) and sending the signal to the micro-controller. A second switch controls the blanking signal voltage as it needs to be adapted depending on the logic of the microcontroller used (3.3 V for ESP32, 5 V for Arduino). The microcontroller controls the LED driver through an SPI connection and sends a trigger signal output to an external DAQ-system. KiCAD schematic are available on the GitHub repository. d. Illustration of the raster scan method described. The “blanking signal” is synchronous with the scanning logic, enabling the LEDs during the scanning mirrors retrace (black) and shutting them off during the acquisition (red), therefore providing temporal separation between stimulation and detection (schematic in (d) inspired from [6]).

The blanking signal controls LED illumination. a. Oscilloscope reading of the blanking signal (blue) efficiently switching off an LED (yellow). The blanking is operated here without noticeable delay. b. as a., shown for a 1 ms scanning cycle, the two possible configurations for the blanking signal input, with a LOW (top) and a HIGH (bottom) blanking signal input for an inverted and original signal input, respectively.

LED electrical power over duty cycle (PWM). a. Power recording of a 4 LED system using the TLC5947 (solid lines) and their expected brightness if directly controlled by a microcontroller (dashed lines). All LEDs have been set up to the same power (40 nW), with equal maximal intensity values in the Arduino code (c.f. 6.3). b. as a. but with LEDs set up at different maximal intensities in the Arduino code. Here the linearity of the LED intensity output remains constant.

Potentiometer Mount PCB. Wiring example of the LED channel 1 to its trimmer potentiometer. Note that LED polarity as indicated on the stimulator PCB must be respected.

Stimulator box. a. Rendering of the stimulator box 3D files set here by default for 4 LED channels and 4 proxy LEDs. b. Rendering of the fully mounted stimulator with all PCBs and components tightly fitting their respective space.

3D-printed Illumination systems. a. SCAD files for adapting 5 mm LEDs and dichroic mirrors to standard 30 mm optomechanical system. b. Rendering of the LED illumination system for the visual experiment. c. For optogenetics experiments, we designed a mounting platform that holds four 5 mm LEDs and can fit a RC-40HP chamber (SmartEphys, Warner Instrument). d. Same as c. but designed to fit a small petri dish (ø 35 mm) lid.

Stimulus example. LED sequence (On/Off steps of light over three loops) described above, along with trigger recording.

Zebrafish retina experiment. a. Overview of the setup described for the visual stimulation experiment performed on the tetrachromatic zebrafish. b. Drawing of the larval zebrafish retina highlighting the IPL. c. 2 photon scan field of the IPL with regions-of-interest marked by red circles. The 64x32 pixel image was obtained by at 1 ms scan rate. d. Ca²⁺ traces (mean traces in black, n = 5 trials in grey) in response to consecutive red, green, blue and UV On/Off flashes. e. Trigger timing recorded by the DAQ highlighting its accuracy over time with a precision of 0.1 µs. t(n + 1) = t(n) + T, where “t” is the recorded trigger time and “T” the trigger period.

Drosophila optogenetics experiment. a. Schematic of a fruit fly first instar larval head expressing the red-shifted channel rhodopsin CsChrimson in olfactory sensory neurons and GCaMP6s in pan-neuronally. b. Rendering of the experimental setup: The mounting chamber (Fig. 6d) is placed in a 3D-printed holder (c.f. 3), screwed onto a rigid stand (ThorLabs). c-d. Standard deviation projections of 2 photon scan fields of the larval brain with antennal lobes marked by red circles (left) and Ca²⁺ traces in response to red flashes (right). c. Stimulation duration = 0.5 s, inter-stimulus interval = 3 s, image dimensions = 256 × 230, scan rate (lines) = 1,081 Hz, frame rate = 4.7 Hz. d. Stimulation duration = 0.5 s, inter-stimulus interval = 10 s, image dimensions = 256 × 170, scan rate (lines) = 1,077 Hz, frame rate = 6.34 Hz. Middle panel is a heatmap of pixel intensities showing high GCaMP6 fluorescence in the antennal lobe following optogenetic stimulation; obtained by subtracting a pre-stimulus from a during-stimulus image (median filter, kernel size = 2).