ADOpt: Automated Signal Delivery for Biosensor Deposition

ADOpt in the lab environment: module, deck space, cables, fixtures, and the workflow around the signal.

Schematic (PDF) Use procedure table (PNG)

Intro

ADOpt came from a practical bottleneck in the lab. Biosensor deposition involved too many manual steps, too much setup time, and bench equipment that was not shaped for the workflow. The project asked a narrower question: can one reliable signal, delivered in the right physical format, save researchers time and increase throughput?

This was my first substantial hardware project. It went through many iterations, taught me how much mechanical integration matters, and eventually produced two modules that were sent to Grenada for field testing.

Purpose

The goal was not to build a general-purpose function generator. The goal was to make one specific deposition signal well, package it into a compact module, and make it easy for another researcher to use.

ADOpt was designed around an Opentrons-compatible workflow, though in practice it was often used as a benchtop device. That still mattered: the geometry, cabling, manifold access, and chip contact strategy were all shaped by the automation workflow around it.

Base information

555-timer signal-generation circuit
50 Hz, 10 Vpp square-wave target
PCB designed in KiCad
CAD enclosure and pressure fixtures
Pogo-pin, copper-tape, and spring-contact delivery experiments
Fluorescent and magnetic bead deposition workflow
Two modules shipped to Grenada

Implementation

The electrical design generated a repeatable deposition waveform. The mechanical design made that waveform usable: chips had to sit under micro assay manifolds, seals had to hold pressure, contacts had to route from single-sided Ti/Au chips, and the workflow had to stay understandable in the lab.

The original chip-contact method used copper tape to route the working and counter electrode contacts from the bottom of a glass slide to the top. Pogo pins then touched those routed contacts. Later designs explored spring contacts and overhead pogo-pin approaches to reduce manual prep.

Design decisions

Decision	Selected direction	Why it fit the lab workflow	Tradeoff / next step
Signal source	555-timer circuit	Simple, compact, repeatable, and dedicated to one deposition signal	Not a flexible lab function generator
Output target	50 Hz, about 10 Vpp square wave	Matched the deposition condition the workflow needed	Tune only around the useful operating region
Contact strategy	Copper routing, pogo pins, spring-contact tests	Kept the glass-chip geometry usable without redesigning the chip	Manual contact prep should be reduced
Mechanical pressure	Manifold clamp, then three-block fixture	Directly addressed gasket sealing and workflow repeatability	Pressure fixture still needs serviceability polish
Deployment style	Benchtop first, Opentrons-compatible geometry	Let the tool work in the real lab while preserving automation compatibility	Full robotic integration can come after reliability

Validation

The module was validated electrically with an oscilloscope and experimentally through bead deposition. In testing, the ADOpt signal produced brighter fluorescent bead outputs than the Rigol benchtop generator being used previously. The likely explanation was that the purpose-built signal had characteristics better suited to the deposition process than the broader spectrum or behavior of the bench generator.

The optical validation used microscope imaging with fluorescence. Testing included molar concentration ranges around 10^-4 M to 10^-7 M, with many chips prepared, labeled, imaged, and reviewed.

The linked schematic and use-procedure table show the handoff trail behind the tool: electrical behavior, use steps, and the practical details needed by another researcher.

Lessons from deployment

The Grenada-deployed fixture revealed a mechanical issue. The one-piece pressure structure could let the side compartments prop up the center, preventing enough force from sealing the middle manifold. The workaround was to run two manifolds at a time. The later design split the fixture into three independent pressure blocks so each manifold and gasket could seal properly.

That lesson is the heart of the project: the circuit can work and still fail the workflow if the physical interface is not right.

Results

Built a compact signal-delivery module for a specific deposition waveform
Replaced a benchtop generator in the deposition workflow
Supported Opentrons-compatible lab automation constraints
Shipped two units for Grenada field testing
Produced useful optical deposition results and informed later SDU work

Future work

Standardize assembly and use documentation
Reduce manual chip contact preparation
Improve serviceability for field deployments
Carry the contact, manifold, and pressure lessons into the Sensor Deposition Unit

A bottleneck reducer, not a general function generator

ADOpt was built around one lab need: deliver one deposition signal reliably enough that biosensor bead deposition no longer depended on a bulky bench generator and a fragile manual setup. If the goal had been a flexible instrument, a 555 timer would have been a strange center of gravity. But that was not the goal.

The goal was to remove a bottleneck. The lab needed a repeatable 50 Hz, about 10 Vpp square-wave condition in a physical workflow with chips, manifolds, contacts, gaskets, cables, and a person trying to get through the assay without rebuilding the setup every time.

That is why ADOpt became useful. It did not win by being electrically fancy. It won by turning one annoying bench condition into an object that could sit in the workflow.

Built ADOpt module in lab context. The tool lived in a real bench workflow, not only as a schematic.

The signal had to become a usable object

The circuit is the simple part to describe: a 555-timer square-wave generator. The harder part was making that waveform land on the chip in a way the lab could repeat. The chips were single-sided, the assay geometry was constrained by the manifold, and the contacts had to reach the electrodes without turning every run into delicate hand work.

That is why the project moved through copper tape, pogo pins, connector boards, and spring-contact ideas. Each version was really asking the same question: how do you deliver the signal without making the chip preparation worse than the problem you are trying to solve?

This is also where my final hardware changed the most from the first idea. The circuit stayed narrow on purpose, but the fixture and contact strategy had to keep adapting to the real shape of the assay.

ADOpt 555 timer schematic and PCB — 555-timer signal-generation design for the deposition waveform.

Validation was optical and electrical

The oscilloscope mattered first because the module had to prove it was producing the intended waveform. But a clean square wave on a scope is not the same as a successful deposition workflow. It only says the circuit is doing its narrow job.

The second validation layer was optical. Chips were prepared, labeled, imaged, and compared so the circuit behavior could be tied to bead-deposition output. Concentration work around 10^-4 M to 10^-7 M belongs in that context: not as a decorative number, but as part of checking whether the workflow produced usable assay results.

That distinction matters. The project was not "I made a square wave." It was "I made the signal and physical interface reliable enough that deposition improved in the workflow."

Close view of bead deposition result — A strong bead-deposition result from later testing, connecting the circuit behavior to assay output.

Grenada exposed the mechanical lesson

Two units were sent to Grenada for field testing, which is where the project stopped being a neat benchtop story. The deployment surfaced a pressure-fixture issue: the one-piece clamp could let the side compartments prop up the center, so the middle manifold did not seal with enough force.

That failure is useful because it is specific. The circuit could work. The deposition condition could be right. The workflow could still fail because a gasket did not see the pressure it needed. The workaround was to run two manifolds at a time, but the better design lesson was to separate the pressure zones.

The later three-block fixture does exactly that. Each manifold gets its own pressure structure, so the fixture is less likely to hide a sealing problem inside a single rigid piece.

ADOpt three-block fixture render — Later three-block pressure fixture, designed in response to sealing problems from the one-piece clamp.

What ADOpt taught the later hardware

ADOpt fed directly into the Sensor Deposition Unit. It made the deposition signal smaller, more repeatable, and easier to hand to another researcher, but it also showed where the next problems lived: contact repeatability, gasket pressure, and the boundary between a tool that works once and a tool that survives a workflow.

That is why I do not think of ADOpt as only my first big PCB project. It was the project where I learned that hardware success is often hiding in the parts that look secondary: the clamp, the contact, the cable path, the label, the procedure, and the person who has to use the thing after I walk away.

ADOpt module with three manifold positions — Three-position ADOpt module view. The later lessons were pressure, contact repeatability, and handoff.