Recording vs Stimulation Electrodes: Design Requirements and Boundaries

JASPER certifications: ISO 9001, ISO 13485, IATF 16949, and ISO 14001.

Scope boundary: This engineering comparison covers noninvasive, skin-contact printed electrode pads. It does not give treatment instructions, electrode-placement advice, patient-safety limits, or regulatory approval conclusions. Implantable-electrode literature is cited only for electrode-interface principles that also require project-specific verification at the skin.

1. Quick Verdict: Decide What Crosses the Interface

The practical verdict is simple. Choose a biopotential recording electrode when the system must preserve a small signal for ECG/EKG, surface EMG, or surface EEG acquisition. Choose an electrical stimulation electrode when the system must distribute a specified waveform for a TENS, EMS, or NMES device. If one physical electrode must do both, treat that as a third architecture—not as a free combination of the first two.

No column “wins” universally. Recording construction is not the best choice when it cannot carry the intended stimulation waveform without unacceptable polarization or current concentration. A large stimulation pad is not the best choice when spatial selectivity, array density, or a compact recording site controls the design. Separate electrodes are usually easier to specify when simultaneous operation, rapid recovery, or independent replacement matters.

2. Two Electrical Jobs, Two Energy Paths

A skin electrode is an electrochemical and mechanical interface among electronic conductors, an ionic contact medium, skin, and the equipment. Stuart Cogan’s 2008 review in the Annual Review of Biomedical Engineering models it with a half-cell potential, double-layer capacitance, charge-transfer resistance, and series/solution resistance. Values vary with frequency, material, area, electrolyte, skin preparation, pressure, hydration, time, and manufacturing history (Cogan, 2008).

Recording path: body signal to converter

Biopotential source in tissue
        ↓ ionic conduction
Skin + electrolyte/contact medium
        ↓ electrode–electrolyte transduction
Exposed recording interface
        ↓ electronic conduction
Printed trace → connector → protected analog front end
        ↓ amplification / filtering / conversion
Recorded ECG, EMG, or EEG data

The recording channel receives rather than drives. Its interface should not add enough offset, noise, drift, impedance imbalance, motion artifact, or environmental pickup to obscure the signal of interest. Stephen Lee and John Kruse’s Analog Devices review places electrode behavior inside the full ECG/EEG/EMG acquisition chain rather than treating the pad as an isolated resistor (Analog Devices). That systems view matters: an impedance value without frequency, electrode pairing, amplifier input characteristics, and test fixture is incomplete.

Stimulation path: generator to tissue and back

Controlled waveform generator
        ↓ lead / connector / printed trace
Current-spreading conductor
        ↓ exposed stimulation interface
Electrolyte/contact medium + skin
        ↓ intended current path through tissue
Return electrode and return conductor
        ↓ monitored output stage
Generator

The stimulation channel drives energy through a complete circuit. Current amplitude alone does not characterize that transfer. The device team must define pulse shape, pulse width, frequency, duty cycle, phase sequence, maximum compliance voltage, active area, return path, fault cases, and the state of the interface over use and aging. Merrill, Bikson, and Jefferys explain why waveform and charge balance matter in neural stimulation, while also showing that electrochemical response depends on the electrode system rather than one universal number (Merrill et al., 2005).

A useful model, not a complete simulation

Electronic conductor ── [interface impedance Z(f)] ── electrolyte/skin
                              │
                              ├─ half-cell potential
                              ├─ double-layer capacitance
                              └─ charge-transfer resistance

For recording, Z(f) affects noise, attenuation, mismatch, and common-mode conversion in the acquisition band. For stimulation, the same interface experiences driven current and time-varying polarization. The model is useful for requirements and test fixtures, but it does not predict local skin current distribution, adhesive lift, gel dehydration, or edge effects by itself.

3. Recording vs Stimulation Electrodes: Side-by-Side Design Matrix

The central electrode pad design difference is the electrical envelope imposed on the same physical stack. Every drawing should connect geometry and materials to that envelope.

Geometry must name the electrically active boundary

The external die-cut outline may include a handling tab, connector island, adhesive border, liner split, and strain-relief region. None of those necessarily equals active area. The drawing should hatch or otherwise identify the exposed conductive/contact region and give its area. For a stimulation concept, current divided by nominal exposed area provides an early geometric current-density calculation, but Cogan and Merrill show why this cannot be treated as a local distribution or safety conclusion. Edge shape, lift, wrinkles, gel thickness, anatomical contour, and return placement can make the interface nonuniform.

4. Where Recording-Electrode Design Takes Priority

Recording design takes priority when the wanted output is a faithful representation of body-generated potential over a defined bandwidth. The electrode, cable, front end, mechanics, and algorithm form one acquisition chain.

Control impedance as a spectrum and a pair

A single resistance measurement can catch an open trace, but it does not characterize a skin-contact recording interface. The requirement should state frequency or frequency sweep, excitation method, fixture/electrolyte or skin model, stabilization time, temperature, electrode pairing, and whether the criterion concerns absolute impedance or mismatch. This is especially important for ECG/EKG electrode patches, where lead pairs and common-mode behavior affect acquisition.

ANSI/AAMI EC12 addresses disposable electrodes used for diagnostic electrocardiography and includes electrical performance methods. It does not make an Ag/AgCl print, hydrogel roll, or converted patch conforming by material name. FDA’s ECG electrode special-controls guidance also separates electrical performance, adhesive performance, biological evaluation, shelf life, and labeling as finished-device evidence topics (FDA ECG electrode guidance).

Treat motion as an electrical input

Cable tug, substrate bending, gel shear, and adhesive lift can change contact conditions and create a signal inside the acquisition band. A lower bench impedance measured in a static saline fixture does not guarantee lower motion artifact on the intended body location. Recording sample approval should therefore pair electrical tests with controlled bend, lead movement, dwell, humidity, temperature, and aging conditions selected by the device team.

Keep the front end in the requirement loop

Input impedance, input bias path, protection components, driven-reference strategy, cable shielding, sampling, filtering, saturation limit, and recovery behavior determine whether an electrode is acceptable. A supplier can verify print continuity and a defined coupon or assembly impedance method. The legal manufacturer still owns system signal quality under the intended acquisition chain and use environment.

Recording-optimized Ag/AgCl is not automatically the best construction when the project needs repeated high-current stimulation, a dry reusable interface, high stretch, dense spatial mapping, or chemistry incompatible with the selected gel and process. Those cases require a fresh materials and architecture decision.

5. Where Stimulation-Electrode Design Takes Priority

Stimulation design takes priority when current delivery—not passive acquisition—is the primary job. The pad must be considered with the generator, cable, return path, waveform, body interface, and foreseeable contact faults.

Specify charge and polarization, not a marketing current rating

For a rectangular phase, nominal charge per phase is current multiplied by phase duration. That number becomes meaningful only with the exposed area, material, interface condition, polarity sequence, interphase interval, repetition rate, duty cycle, and voltage transient. Reversible charge storage differs from irreversible faradaic processes, and charge-injection capability is material- and test-dependent. A generic “supports X mA” statement therefore cannot replace polarization and waveform evidence.

Nominally charge-balanced biphasic output can reduce net DC, yet the generator’s programmed balance does not prove zero residual electrode polarization. Component mismatch, coupling capacitors, switching, timing error, interface asymmetry, and partial contact can change the result. The system requirement should define what is measured, where it is measured, over how many pulses, and what recovery criterion applies.

Area and edge condition are functional dimensions

A stimulation electrode often uses a broader exposed region to spread current. Larger is not always better. Increased area can conflict with placement envelope, anatomical selectivity, adhesion over curvature, cable load, packaging, or the desired field. Rounded transitions, controlled dielectric overlap, uniform contact medium, and a defined return geometry may matter more than merely increasing the die-cut outline.

For TENS, EMS, and NMES electrode pads, the OEM should provide the maximum normal and fault waveforms, not just the device category. IEC 60601-2-10 covers particular requirements for nerve and muscle stimulators at the medical electrical equipment level. It does not certify an unconnected printed pad or substitute for the finished system’s risk controls, essential-performance analysis, and market-specific regulatory work (IEC 60601-2-10).

Design the return path at the same time

Current leaves through one interface and returns through another. If the return electrode, lead, connector, or contact condition differs, the pair may not share voltage or current stress equally. Draw both paths, define polarity or alternation, identify connector pins, and evaluate open, shorted, lifted-edge, partially attached, dried, folded, and wrong-pad conditions under the device risk process.

A broad stimulation construction is not the best choice when the main objective is low-noise recording, fine spatial discrimination, a dense multi-channel array, or rapid post-pulse acquisition. In those cases, separate sensing sites or a dual-region architecture may reduce compromise.

6. Material and Layer-Stack Choices Follow the Function

Material selection starts with the interface job, then proceeds through process compatibility. “Ag/AgCl,” “carbon,” “PET,” “TPU,” and “hydrogel” each describe a family. Ink ratio, binder, particle loading, dry-film thickness, cure history, surface treatment, dielectric overlap, gel chemistry, adhesive, liner, and aging state can change the result.

Recording-oriented printed stack

Release liner
Conductive electrolyte / hydrogel or qualified dry interface
Exposed Ag/AgCl or other qualified recording interface
Printed conductor trace under dielectric
PET, TPU, or other qualified substrate
Connector reinforcement / snap / tail termination
Backing, label, or carrier as required

Ag/AgCl is a common recording-interface candidate because the Ag/AgCl–chloride reaction can behave relatively nonpolarizably under suitable electrolyte conditions. That does not make every Ag/AgCl ink equivalent. DuPont 5874, for example, is a 65:35 Ag:AgCl composition specified for print-treated polyester; its data sheet describes a formulation and cure process, not a universal electrode property (DuPont 5874 technical data sheet). The selected interface still needs compatibility work with the conductor below it, gel above it, dielectric edge, connector process, packaging, and acquisition circuit.

A dry electrode may be preferable where gel dehydration, residue, packaging, or reuse dominates. It may also have different contact impedance, pressure dependence, motion sensitivity, and surface-wear behavior. The trade is real. “Dry” is an architecture branch, not an automatic upgrade.

For a source-backed recording-side planning point, Axelgaard's November 2022 AG625 sensing-gel TDS specifies 0.6 ± 0.1 mm thickness, maximum 600 Ω·cm volume resistivity, minimum 240 g/in skin-side peel on stainless steel, pH 3.5 ± 0.5, and bulk-roll shelf life of 12 months non-slit or 6 months slit (Axelgaard AG625 TDS). These are supplier-grade data, not JASPER results, skin-adhesion evidence, or finished-electrode shelf life.

Stimulation-oriented printed stack

Release liner
Qualified conductive contact medium
Exposed current-spreading interface
Low-resistance bus / printed conductor
Dielectric with controlled active-area opening
Dimensionally stable or extensible substrate
Lead wire, snap, tab, or tail with strain relief
Backing and skin adhesive zones as specified

Carbon, conductive rubber or polymer, silver-based conductors, and other material systems may support stimulation designs. Their roles should be separated. A low-resistance silver bus can carry current beneath dielectric while a different exposed layer controls the interface. Carbon may spread current or isolate another chemistry, but a carbon ink’s sheet resistance does not by itself state pulse polarization, gel compatibility, or local current distribution.

PET is often selected when dimensional registration and a stable printed circuit are primary. A stretchable TPU stack may fit motion and contour better, but the ink and dielectric must survive the specified strain and cure. Henkel ECI 1014 is one example of a stretchable silver ink; its existence does not prove compatibility with a different supplier’s TPU, dielectric, gel, or adhesive (Henkel ECI 1014).

For a stimulation-side planning point, Axelgaard's November 2022 AG735 TDS specifies 0.9 ± 0.1 mm thickness, maximum 1,500 Ω·cm volume resistivity, minimum 160 g/in skin-side peel on stainless steel, pH 4.2 ± 1.0, and bulk-roll shelf life of 18 months non-slit or 6 months slit. Axelgaard identifies AG735 as a stimulating gel designed for conductive film (AG735 product page; AG735 TDS). This is not a pulse-performance result, wear claim, or approved JASPER material.

ISO 10993-1:2025 frames biological evaluation within risk management for a device with direct or indirect body contact (ISO 10993-1). It is not a certificate attached to a raw ink or adhesive. The legal manufacturer must evaluate the finished contacting materials, processing residues, contact category and duration, packaging, aging, and other relevant evidence. ASTM D3330/D3330M-04(2025) can support a defined pressure-sensitive-adhesive peel comparison, but it does not demonstrate skin safety, clinical wear, or comfort (ASTM D3330).

7. Can One Pad Record and Stimulate?

One electrode can perform both functions, but “can” and “should” are different decisions. A 2026 Scientific Reports paper described a dry electrode tested for EMG recording and transcutaneous electrical stimulation. It established feasibility for that construction, not universal interchangeability (Scientific Reports, 2026).

A dual-function design must answer at least five extra questions:

1. Operating state: Are recording and stimulation sequential, interleaved, or simultaneous?
2. Switching: What protects and disconnects the sensitive acquisition front end during output pulses?
3. Artifact: How much saturation occurs, and how long until the electrode plus electronics recover to the recording criterion?
4. Polarization: What residual interface voltage remains after the pulse sequence, including component and contact mismatch?
5. Fault containment: What happens if switching fails, one electrode lifts, a connector is reversed, or the wrong accessory is attached?

stimulation produces artifact on recording electrodes in its microelectrode context (NeuroNexus technical note). A surface-pad system has different geometry and tissue, yet the integration lesson holds: artifact and recovery must be measured in the intended signal chain.

Use separate electrodes when simultaneous operation, rapid post-pulse measurement, independent geometry, easier replacement, or clean verification boundaries outweigh package reduction. Use a dual-region electrode when shared placement is valuable but the recording and stimulation sites need different exposed materials or areas. A single shared active area is not the best choice merely because it reduces part count.

8. Follow the Failure Chain, Then Assign the Test

An electrode problem often appears one layer away from its cause. A lifted adhesive edge can change contact area; changed area can alter interface impedance or current distribution; the electronics may then report noise, saturation, compliance-limit operation, or a user-visible fault. Troubleshooting only the printed trace misses that chain.

Verification and validation ownership matrix

The word “test” is too vague for a sourcing package. Name the specimen, fixture, stimulus, environment, sample plan, acceptance criterion, record, and owner. A component manufacturer can control print and conversion characteristics, but only the device design authority can approve the system against its intended use.

The project’s quality and testing framework can support manufacturing discussions, but the actual test list and acceptance criteria must be confirmed for the quoted construction. ISO 14971:2019 provides the life-cycle risk-management framework for assigning hazards, controls, and evidence; it does not set one acceptable-risk level for every device (ISO 14971).

9. Drawing and Sample-Approval Checklist

A useful RFQ does not begin with “make this pad in the attached shape.” Start with the electrical function. Tie each critical drawing feature to a requirement. The related custom printed medical electrode design guide provides the broader development sequence; the checklist below isolates the recording-versus-stimulation handoff.

Drawing checklist

• [ ] Declare record, stimulate, or both, including whether dual functions are simultaneous or sequential.
• [ ] Identify modality: ECG/EKG, EMG, EEG, TENS, EMS, NMES, or another defined system function.
• [ ] Dimension the external outline and the exposed active area separately; state area units.
• [ ] Mark conductor buses, sensing sites, return sites, dielectric openings, overlaps, keep-outs, and edge clearances.
• [ ] Define substrate grade and thickness; identify bend, stretch, or strain-restricted zones.
• [ ] Name each conductive and interface material by controlled specification, not generic color or chemistry alone.
• [ ] Define contact medium, skin adhesive zones, nonadhesive handling areas, backing, and release-liner split.
• [ ] Specify connector type, pinout, channel identity, polarity/state, retention, reinforcement, and cable exit direction.
• [ ] Provide normal and fault electrical envelopes: acquisition bandwidth or maximum waveform, pulse width, frequency, duty cycle, compliance voltage, phase sequence, and return path as applicable.
• [ ] Define artwork revision, datums, critical-to-function tolerances, labeling, packaging configuration, and change-notification rules.

Sample-approval checklist

Now test the stack, not a coupon alone. A 2026 design review should retain the exact PET or TPU grade, Ag/AgCl or carbon formulation, hydrogel lot, IEC 60601 equipment context, ISO 10993-1 contact rationale, connector revision, and firmware build with the results. Change one controlled input, and the device team decides what evidence must be repeated.

Do not approve samples from one attractive oscilloscope trace. Use predetermined criteria, representative specimens, documented fixtures, relevant conditioning, and the intended electronics or a justified equivalent. A supplier first-article approval is not the finished-device validation report.

10. Decision Matrix: Record, Stimulate, or Both?

The next step is not to select a catalog shape. Identify whether the project records, stimulates, or performs both, then attach the signal or waveform envelope, active-area drawing, connector definition, and validation-owner matrix. JASPER may be evaluated as one printed electrode-pad component source after those inputs exist. The finished-device developer retains responsibility for application suitability, verification, validation, regulatory strategy, and approval.

11. Frequently Asked Questions

What is the main difference between recording vs stimulation electrodes?

A recording electrode transfers body-generated electrical potential into a measurement circuit; a stimulation electrode delivers generator-controlled current into tissue. That direction of transfer changes the critical requirements: recording emphasizes signal fidelity, impedance/mismatch, noise, and motion artifact, while stimulation emphasizes waveform, charge per phase, polarization, active area, current distribution, and fault behavior.

Can the same electrode pad record and stimulate?

Yes, a dual-function electrode can be feasible, but it requires a defined switching state, acquisition-front-end protection, artifact and recovery limits, polarization evidence, and fault analysis. Separate electrodes are often the better choice when recording must continue during stimulation, rapid recovery matters, or the two functions need different geometries and materials.

Why is Ag/AgCl common in a biopotential recording electrode?

Ag/AgCl can provide a relatively stable, comparatively nonpolarizable ionic-to-electronic interface under suitable chloride and electrolyte conditions. Performance still depends on the exact ink, Ag:AgCl ratio, cure, exposed area, gel chemistry, aging, skin/contact condition, and acquisition circuit. The label ‘Ag/AgCl’ alone does not approve a recording design.

Does an electrical stimulation electrode need low impedance?

It needs an interface compatible with the intended waveform and output stage, but ‘low impedance’ without frequency, pulse conditions, area, fixture, and acceptance criterion is incomplete. The device team should examine voltage transient, polarization, charge per phase, distribution, contact faults, and generator compliance—not only a single resistance or impedance value.

What is the most important electrode pad design difference on the drawing?

The drawing must separate the exposed electrically active area from the external die-cut outline and identify its material, conductor path, dielectric boundary, contact medium, adhesive zones, and connector. Recording drawings also need channel/reference identity; stimulation drawings need active/return paths and the electrical envelope tied to those areas.

Which standards apply to recording and stimulation electrode pads?

ANSI/AAMI EC12 is relevant to disposable diagnostic ECG recording electrodes, while IEC 60601-2-10 addresses nerve and muscle stimulator equipment. IEC 60601-1, ISO 10993-1, ISO 14971, and market-specific rules may also apply. Applicability and edition must be determined for the defined finished device; a printed component does not inherit approval.

Who validates a custom printed electrode pad?

The component manufacturer verifies agreed material, print, geometry, continuity, resistance, dielectric, converting, connector, and packaging characteristics. The finished-device developer or legal manufacturer owns signal or stimulation performance, biological evaluation, aging rationale, medical electrical equipment testing, usability, labeling, regulatory submissions, and intended-use validation.

What information should an OEM send before requesting samples?

Send the record/stimulate/both declaration; modality; signal bandwidth or complete normal/fault waveform; active and return-area drawing; materials and layer stack; connector/pinout; contact and wear conditions; packaging/aging plan; test methods and acceptance criteria; and a responsibility matrix. Replace every initial planning value before design-input approval. For a recording sample, identify ECG/EKG, EMG, or EEG plus the intended front end. For a stimulation sample, identify TENS, EMS, NMES, or the defined application plus pulse width, frequency, current, compliance voltage, phase sequence, and return path. For both, identify recovery time, blanking, switching, and artifact criteria.

Technical References

• Source: FDA Use of ISO 10993-1 Biological Evaluation Guidance. Accessed 2026.
• Source: ISO 10993-1:2025 Biological Evaluation of Medical Devices. Accessed 2026.
• Source: ISO 14971:2019 Medical Device Risk Management. Accessed 2026.
• Source: FDA Design Control Guidance for Medical Device Manufacturers. Accessed 2026.
• Source: Cogan, 2008. Accessed 2026.
• Source: Analog Devices. Accessed 2026.
• Source: Merrill et al., 2005. Accessed 2026.
• Source: FDA ECG electrode guidance. Accessed 2026.
• Source: IEC 60601-2-10. Accessed 2026.
• Source: DuPont 5874 technical data sheet. Accessed 2026.
• Source: Axelgaard AG625 TDS. Accessed 2026.
• Source: Henkel ECI 1014. Accessed 2026.
• Source: AG735 product page. Accessed 2026.
• Source: AG735 TDS. Accessed 2026.
• Source: ISO 10993-1. Accessed 2026.
• Source: ASTM D3330. Accessed 2026.
• Source: Scientific Reports, 2026. Accessed 2026.
• Source: NeuroNexus technical note. Accessed 2026.
• Source: ISO 14971. Accessed 2026.