Skip to the content
2 layers99.5 × 89.5 mmCustom ffc, the printed circuit board in 3D

Design and implementation of an automated high-content screening cell perfusion system, including associated control software

Size
99.5 × 89.5 mm
Copper layers
2
Thickness
1.6 mm
Parts
9
Nets
9
Pads
25
Tracks, vias
49, 0
Pours
21
Routed
Yes
KiCad
KiCad 10.0 (file 20260206)
Created
11 Dec 2025
Updated
5 Nov 2026

Overview

Custom ffc is an open-source printed circuit board by olxssa, from the olxssa/perfusion_automation repository on GitHub. It is a 2-layer board of 99.5 × 89.5 mm with 9 parts on 9 nets, routed with 49 tracks and 0 vias.

README

README.md

perfusion_automation

Design and implementation of an automated high-content screening cell perfusion system, including associated control software

Here, the design, implementation and benchmarking of a fully integrated, modular screening platform for automated, multiplexed bath perfusion experiments is documented. The system was developed to provide reproducible and sample-friendly solution exchange for optical characterization of biosensors.

The platform consists of three primary systems: a modular fluorescence microscope, an automated bath perfusion add-on, and an integrated hardware and software control framework.

Repository Content

FreeCAD Files

Using the software FreeCAD, the provided CAD files can be exported as .stl and 3D-printed using PLA, 15 % infill and 0.2 mm layer height. Alternatively, the files can be adapted to your liking.

FPC Design Files

In order to create a more elegant design for controlling the microvalves, a custom Flexible Printed Circuit (FPC) was designed using KiCAD in accordance with the information given by the company memetis. The microvalves are controlled via a constant-current source. The valves are all connected in series and can each be bypassed on the circuit board using a MOSFET, with a tap between the valve contacts. This allows individual valves to be removed from the current path and taken out of the control circuit.

Custom-designed flexible printed circuit (FPC) for ECU-I 8CH and eight microvalves positioned on two 4-valve adapters and one 2-valve adapter. (A) Top-view of perfusion arm where 8 individual cables were connected to the controller adapter. (B) The PCB layer of the custom-designed FPC in KiCAD. (C) Top-view of the perfusion arm with the FPC.

Firmware and MMConfig

Low-level hardware management is split between two microcontrollers and using the firmware rx_system.ino and tx_system.ino. High-level microscope hardware control is achieved via Pycro-Manager [1], which bridges the Python backend to a headless Java-based Micro-Manager [2] (v2.0.3) instance over port 4721. The Micro-Manager instance is created using the config file MMConfig.cfg.

Software

After the build, the setup can be controlled via a custom-developed web application called "Automated Bath Perfusion App". The web application serves as a single-page dashboard designed to organize all operational controls into three dedicated panels arranged from left to right:

  • Preview:
    • Preview window featuring modes live, BF snap and fluorescence snap
    • Choose ideal fluorescence exposure time for your experiment
    • Choose indeal BF light source brightness for the software autofocus to work with your samples
    • Manual stage movement
    • Choose ideal software autofocus parameters for your experiment
    • Test flow from reservoirs 1-5
  • Perfusion Experiment:
    • Set folder location to save all experiment files
    • Multi-well plate-type tool
    • Set imaging parameters: observation time after perfusion and imaging frequency
    • Name reservoirs accordingly to perfusate solutions
    • Set solution sequence and volumes to be perfused
    • Additional tools:
      • CenterFOV: By default, the well center coordinates are used as Field of View
      • AutoFOV: Automatically determine best Field of View for every well
      • ManuFOV: Manually determine best Field of View and focus position for every well
  • Preliminary Analysis:
    • Cell segmentation of first fluorescence image for every well
    • Dynamic classification of cells that are fluorescent
    • Background subtraction and computation of fluorescence change over time for individual cells for every well
    • Comparison of mean of fluorescence change over time for all wells

Data

The system's performance is validated by assessing pipetting procedures, demonstrating software autofocus stability and conducting automated multiplexed perfusion on cells expressing GFP and NEMO-based calcium sensors under agonist conditions.

Documentation of the Project

1. Building a Modular Fluorescence Microscope

Schematic optical light path of the modular fluorescence microscope.

2. Constructing an Automated Bath Perfusion System for Screening

Implementation of the automated bath perfusion system. (A) The complete automated screening perfusion system integrated into the modular fluorescence microscope. (B) The perfusion arm responsible for movement of tubes into single wells and for the perfusion and aspiration of the desired solution.

Fluidic Circuit

Wiring of the fluidic circuit. (A) Pressure distribution system. (B) Reservoirs with perfusion solutions. (C) Microvalve circuit and channel numbering. (D) Perfusion site with precise volume exchange.

Electrical Circuits

Wiring of the electrical circuits. (A) The RX system receives signals from the PC and controls the servo motors, the stepper motor drivers, relays (switchable power socket, 3/2-valve relays), and the Neopixel LED ring. (B) The TX system gathers information from the flow-rate sensor, the fluid level sensors and the stop button and sends the information to the PC.

3. Developing a Custom Web App as a User Interface using Flask

Web App Architecture. Multi-layered system showing the interaction between green software layers, blue hardware nodes, and the physical microscope components in grey.

Perfusion experiment workflow sequence. (A) Initialization and sensor validation. (B) Automated well traversal, solution perfusion, and imaging loops. (C) Decontamination protocol and park execution.

4. Calibration and Benchmarking

Flow-rate stability characterization. (A) Flow-rate robustness and transient oscillations observed during operational tuning. (B) Flow-rate decay resulting from increased system resistance upon integrating additional reservoirs. (C) Hydrostatic flow-rate decay as reservoir fluid levels deplete (tested across 100 mL, 50 mL, and 10 mL capacities). Modularity allows for interchangeable reservoir sizes to accommodate varying experimental volume requirements.

Software autofocus performance throughout a 6-well plate. For each well, except for the starting position A1, a course scan followed by a fine scan are carried out as shown on the left, resulting in a focussed sample shown on the right of each well.

Bath perfusion experiment with GFP (left) and NEMO (right) tagged ND cells. (A) Fluorescence overlay with segmented masks detected by Cellpose. The green outlines are the cells labeled as fluorescent, the black outlines show the rest of the segmented cells. (B) Physiological fluorescence intensity change / SBR of single cells (grey) and mean (red). (C) Mean trajectories of all wells for comparison.

References

[1] H. Pinkard, N. Stuurman, I. E. Ivanov, et al., “Pycro-manager: Open-source software for customized and reproducible microscope control,” Nat Methods, vol. 18, 226–228, 2021. DOI: 10.1038/s41592-021-01087-6.

[2] A. D. Edelstein, M. A. Tsuchida, N. Amodaj, H. Pinkard, R. D. Vale, and N. Stuurman, “Advanced methods of microscope control using μmanager software,” Journal of Biological Methods, 2014. DOI: 10.14440/jbm.2014.36.

[3] J. Li, Z. Shang, J. H. Chen, et al., “Engineering of nemo as calcium indicators with large dynamics and high sensitivity,” Nature Methods, vol. 20, pp. 918–924, 2023. DOI: 10.1038/s41592-023-01852-9.

[4] P. Dettinger, T. Kull, G. Arekatla, et al., “Open-source personal pipetting robots with live-cell incubation and microscopy compatibility,” Nature Communications, vol. 13, no. 1, p. 2999, 2022. DOI: 10.1038/s41467- 022-30643-7.

Demonstration Video:

Version: October 2026

Click on image to play on Youtube.com

Parts

9
QtyValueFootprintReferences
8Conn_01x02_PinConnector:memetis_01x02J1, J2, J3, J4, J5, J6, J7, J8

The 8 most common of 9 parts.

Similar boards

  • 2 layers103 × 30.5 mmTS Back Endcap, the printed circuit board in 3D

    TS Back Endcap

    EEVengers/ThunderScope

    ThunderScope GitHub Repo

    9 parts1.3kMIT
  • 2 layers31.8 × 47.6 mmNGC 38 connector, the printed circuit board in 3D

    NGC 38 connector

    rusefi/rusefi

    rusefi - GPL internal combustion engine control unit

    9 parts1.2k
  • 2 layers24 × 27.3 mmExpansion Card Retrofit, the printed circuit board in 3D

    Expansion Card Retrofit

    FrameworkComputer/ExpansionCards

    Reference designs and documentation to create Expansion Cards for the Framework Laptop

    9 parts1.1kCC-BY-4.0
  • 2 layers15.6 × 17.3 mmSKR Mini TFT Thermistor Board, the printed circuit board in 3D

    SKR Mini TFT Thermistor Board

    VoronDesign/Voron-Hardware

    Cool Hardware designed by the Voron Design group

    9 parts1.1kGPL-3.0
  • 2 layers39.1 × 49.5 mmLearnoliver, the printed circuit board in 3D

    Learnoliver

    hackclub/OnBoard

    💡 Join 1,000 teenagers and make your first PCB with a $100 grant!

    9 parts915
  • 2 layers17.4 × 16.9 mmToolhead, the printed circuit board in 3D

    Toolhead

    tanaes/whopping_Voron_mods

    9 parts787GPL-3.0
  • 2 layers9.8 × 108 mmExpander, the printed circuit board in 3D

    Expander

    wntrblm/Castor_and_Pollux

    A Juno-inspired dual oscillator

    9 parts587
  • 2 layersKeyboard after, the printed circuit board in 3D

    Keyboard after

    adamws/kicad-kbplacer

    KiCad plugin for automatic keyboard's key placement and routing

    9 parts547GPL-3.0

Questions

How big is the Custom ffc PCB?

It measures 99.5 × 89.5 mm and has 2 copper layers, 1.6 mm thick.

What parts does the Custom ffc use?

9 parts. The full list of its parts is on this page, with links to each part's symbol and footprint where PCBKit's library has them.

Where can I download the Custom ffc KiCad files?

From the olxssa/perfusion_automation repository on GitHub, at the commit this page was made from: the board is FPC/custom_fpc_memetis_improved/custom_ffc.kicad_pcb and its schematic FPC/custom_fpc_memetis_improved/custom_ffc.kicad_sch. PCBKit opens the board in the browser, with nothing to install.

Can I use the Custom ffc design in my own project?

Its repository states no licence, so its author keeps all rights: ask olxssa before reusing the design.