• Laboratory prototype for performing parallel electrolysis reactions in 96 wells.
• Electrical readout is not required; experiment readout is expected to be color change.
• Wetted environment: buffer/saline.
• The board is not intended for medical use or sterilization-critical use.
• Preferred manufacturer: PCBWay.

• Provide 96 liquid wells or 96 gasketed reaction areas.
• Expose three electrode types per well to liquid:
  • Working electrode, globally shared across all wells.
  • Reference or pseudo-reference electrode, globally shared across all wells.
  • Counter electrode, globally shared across all wells.
• Keep non-electrode copper/traces isolated from liquid by solder mask or mechanical sealing.
• Support a future AAA-powered controller that applies approximately 0.1 V to 1.5 V DC in 0.1 V increments.
• Keep expected total current below about 1 mA unless later testing changes the requirement.

• 8 x 12 well array.
• Approximate ANSI/SBS/Bio-Rad style plate footprint.
• 9.00 mm well pitch preferred unless the mechanical/sealing approach requires adjustment.
• Hard gold target on exposed electrodes, user preference: 20 microinch hard gold if PCBWay supports it for this geometry.
• A1 to H12 labeling.
• Off-board connection through three robust contacts/connectors for WE, RE, and CE.
• Two mechanical variants should be considered:
  1. Preferred manufacturable prototype: flat electrode PCB plus gasket/O-ring or top well plate.
  2. Experimental monolithic version: thick PCB with controlled-depth machined wells.

Diagram

• 96 repeated electrode cells.
• Each cell should expose only the intended electrode metals to liquid.
• Preferred top geometry per user direction:
  • Working electrode: annular or large edge-region feature, about 1 mm width.
  • Reference electrode: annular or large edge-region feature, about 1 mm width.
  • Counter electrode: central circular pad, target diameter 2 mm or 3 mm, intended for later silver paint deposition.
• The exact ordering of the two annuli must be finalized before layout because it affects reference proximity and sealing.

• Only three nets are required: WE_GLOBAL, RE_GLOBAL, CE_GLOBAL.
• Because current is very low, full internal planes are not electrically necessary, but large pours/busses may be useful for symmetry, low resistance, and easier review.
• A 4-layer stack is preferred for the main design.
• Avoid vias inside wetted regions unless they are filled/capped and protected.

• User suggested SMB connectors.
• Engineering preference: use either three SMB connectors for robust coax-style lab wiring, or a compact keyed 3-pin connector plus test pads if cost/space matters.
• Connector choice should be finalized when the companion AAA regulator/controller is designed.

• External controller not yet designed.
• Future power source: AAA battery.
• Future output range: 0.1 V to 1.5 V DC, 0.1 V increments.
• Current target: below 1 mA total.
• No onboard active electronics are currently required on the passive electrode plate.

Diagram

• Preferred fab: PCBWay.
• Maximum layer count: 4 layers.
• Preferred material: FR4.
• Overall thickness target: less than 8 mm.
• User originally targeted about 6 mm total thickness.
• Hard gold target: 20 microinch if available and compatible with selective electrode exposure.
• User suggested 4 mil trace/space and 0.3 mm vias; these are reasonable as initial layout constraints, but the design is mechanically/electrochemically limited rather than trace-width limited.
• Blind/buried vias are allowed by user but should be avoided for cost and reliability.

• Use a relatively normal PCB thickness such as 1.6 mm to 2.4 mm unless a thicker stiffener is required.
• Electrode cells are exposed on the top face.
• Wells are created by a gasket sheet, O-rings, or separate machined plate clamped above the PCB.
• This avoids raw machined FR4 sidewalls contacting saline.
• This is the preferred first manufacturable route.

• Use 5 mm to 6.4 mm thick FR4, still below 8 mm total.
• CNC/control-depth machine wells 2 mm to 4 mm deep.
• Keep internal copper far away from well sidewalls.
• Require fab confirmation before ordering.
• Risk: raw FR4 sidewalls and machining debris contact liquid unless sealed.

• Rebuild cleanly rather than repair the Claude/KiCad dummy.
• Use the Claude/KiCad dummy mainly for dimensions and labels.
• Treat the requested three channels as global nets, not per-well independent channels.
• Prefer a flat/gasketed first prototype over monolithic machined FR4 wells because saline exposure of raw FR4 is a major risk.
• Document the reference electrode as a pseudo-reference unless Ag/AgCl chemistry is added.
• Counter electrode can be a central pad intended for silver paint deposition, but this should be validated chemically.

1. Electrode ordering: which annulus is working and which is reference should be finalized.
2. Reference electrode chemistry: gold reference is a pseudo-reference; Ag/AgCl is better if a stable reference matters.
3. Surface finish: PCBWay must confirm hard gold thickness, selective plating limitations, nickel underlayer, and saline exposure suitability.
4. Wet FR4 risk: monolithic CNC wells expose raw FR4 unless sealed.
5. Sealing: 96 isolated wells require uniform compression and leak testing.
6. Connector choice: three SMBs are possible but bulky; a keyed 3-pin connector may be cheaper and easier.
7. Current limiting: future controller should include current limiting; the passive plate should at least include test pads and possibly optional series resistors if integrated.
8. Prototype strategy: a 3 x 3 or 4 x 4 coupon is strongly recommended before ordering the full 96-well version.

• SBS-style 9 mm pitch is acceptable.
• Bio-Rad outer footprint dimensions are the mechanical starting point.
• Total current target is less than 1 mA for the whole array, not per well.
• The design is for short-duration lab prototype testing, not production biomedical use.
• User wants outputs viewable/checkable in KiCad eventually.

• Initial specification created from the user answers and research review.
• Project renamed from draft placeholder to 96-Well Electrode Plate.
• The attached KiCad dummy was inspected and found useful for Bio-Rad-like outline/pitch, but not suitable electrically or mechanically as-is.
• Rebuild started in Flux: created WE_GLOBAL, RE_GLOBAL, and CE_GLOBAL nets; added a 3-pin Molex 0022232031 prototype connector J1; configured a 127.76 mm x 85.48 mm Bio-Rad-style board outline with a 4-layer stackup; placed the 96-well electrode field on 9.00 mm pitch.
• Current electrode implementation for review: CE is 96 individual 2.5 mm circular exposed SMD pads, one per well. WE and RE are implemented as shared 1.0 mm wide horizontal exposed row bars above and below each well row, respectively. This is a manufacturable flat-gasket prototype simplification of the requested annular WE/RE concept and should be reviewed before routing/export.
• Routing is intentionally not complete yet. The current checkpoint is placement/geometry review before adding copper interconnect traces or distribution planes.