LACED is an innovative delayering technique developed for reverse engineering modern hardware, with a focus on multilayer PCBs, using an extremely accessible and low-cost approach. It combines a low-power, budget-friendly UV laser, simple chemical solutions, and micrometer-based precision tracking — enabling advanced circuit exploration without the need for industrial tools or cleanroom environments.
Through this process, LACED achieves precision levels between 3.5 and 10 microns, bridging the gap between DIY and professional-grade delayering.
More than just a method, LACED is a mindset:
To prove to the world that ingenuity beats million-dollar industries.
The LACED™ technique was developed with the following goals in mind:
-
Accessibility
To enable advanced delayering and reverse engineering using only low-cost, widely available tools — without the need for industrial machinery or cleanroom environments. -
Precision
To achieve micron-level control (3.5–10 µm) in the removal of FR4 and copper layers, making it suitable for high-resolution hardware analysis. -
Documentation
To create a detailed and replicable methodology that can be adopted, improved, and shared by the maker and research communities. -
Empowerment
To prove that a single person, in a bedroom lab, can match the results of processes that normally require million-dollar setups. -
Emotional Purpose
To explore not only circuitry, but memory — by uncovering physical messages hidden between PCB layers, turning reverse engineering into an act of personal and artistic expression.
✅ Achieved physical precision between 3.5 and 10 microns, using a low-cost CNC diode laser (5W), micrometer tracking, and basic chemical etching — all performed in a home environment.
✅ Successfully exposed internal copper layers of a 6-layer PCB, proving that high-resolution delayering can be accomplished without industrial tools like FIB systems or pro-grade milling stations.
✅ Recovered a hidden message engraved between layers — showing that LACED is not only a technical method, but also a form of expression capable of revealing both data and meaning.
✅ Developed a fully documented, step-by-step methodology, reproducible with accessible equipment, allowing anyone with curiosity and discipline to explore modern hardware at micron scale.
Why it matters: LACED lowers the entry barrier to advanced PCB analysis and digital preservation. It empowers individuals to explore, understand, and protect hardware — not with money, but with ingenuity, precision, and intent.
This process was born from a very specific need: gaining access to the internal layers of multilayer PCBs for reverse engineering.
At first, I theorized that a 5W UV diode laser could be used to remove the soldermask without having to do it manually — eliminating the need for sandpaper and the inconsistent pressure of a human hand, which often leads to uneven surface removal.
Then came the realization: the laser could also ablate the FR4 itself. That’s when a deeper idea took shape — what if I could use the laser to remove all layers of FR4 and dielectric material with consistent precision, rather than relying on error-prone manual sanding?
But very quickly, I hit a fundamental physical limitation: copper.
The copper traces on the board would not be affected by the laser. And even if I increased power to the point of engraving copper, the material’s thermal properties are so different from FR4 that I would overburn the substrate while barely affecting the copper itself.
That’s when a new idea struck: I remembered how PCBs are fabricated — a process I had replicated at home. In standard manufacturing, a uniform copper layer is coated with a photoresist, then selectively etched with acid, leaving only the designed traces.
What if we inverted the logic and used acid not to preserve the traces, but to remove them completely?
With this shift, the theoretical workflow became clear: laser-assisted soldermask removal, acid-based copper etching, and layer-by-layer FR4 ablation. A fully controlled, repeatable process — at least in theory.
And as my friend Oppenheimer once said, “Theory only gets you so far.”
Traditional manual sanding — using fine sandpaper to gradually expose PCB layers — has long been the go-to method for low-budget delayering. While it’s simple and doesn’t require sophisticated tools, it suffers from major limitations:
❌ Inconsistent pressure from the hand leads to uneven layer removal.
❌ Risk of over-sanding or skipping layers, especially when approaching internal copper.
❌ Lack of precision tracking — no reliable way to measure removal depth micron by micron.
❌ Requires experience and luck to achieve uniformity across the board.
❌ Irreversible damage is common when targeting a specific layer.
LACED™ eliminates these issues by replacing guesswork with control.
✅ Uses a low-power CNC diode laser to ensure consistent surface processing across the entire PCB.
✅ Incorporates micrometer-based depth tracking after each pass.
✅ Combines chemical copper removal to avoid thermal stress and material distortion.
✅ Achieves repeatable results down to 3.5–10 µm, far beyond manual sanding tolerances.
✅ Allows precise stopping at specific layers without overshooting.
In short: manual sanding relies on intuition and hope. LACED relies on precision and repeatability.
One of the most critical and frustrating challenges in the LACED process was the unintentional damage to copper traces — especially the finer ones — during laser carbonization of the FR4 substrate.
When using a diode laser to ablate the surface, excessive heating of the FR4 can occur. This heat, in turn, radiates into nearby areas, often beneath the copper traces. The result: localized carbonization and mechanical delamination of the copper, causing the traces to lift, warp, or detach.
This issue was particularly severe in the early phases, when I was still calibrating laser power and speed parameters. It was not just a technical setback — it was deeply frustrating, because it often ruined boards that had taken hours of precise work.
But today, that challenge has been overcome.
The process is now stable and predictable, thanks to those early failures. And that’s why I say:
Failure builds precision.
The following tools and chemicals were used to perform the LACED™ delayering process. All items are low-cost, widely available, and suitable for a compact home lab setup.
🔧 Hardware & Instruments
~€160 — Entry-level CNC laser engraver with adjustable speed/power and good stability for micron-scale ablation.
~€10–15 — Used for precise measurement of PCB thickness after each pass.
Included with the laser machine — Essential for eye protection during engraving.
🧪 Chemicals
~€5 — Used for soldermask softening and initial surface preparation.
~€1.50 — Primary agent for copper etching.
~€2 — Used in combination with HCl for controlled etching of copper.
~€4 — Used for final cleaning and degreasing of PCB surfaces.
Total estimated cost: under €200. For a fully functional, repeatable delayering lab — no cleanroom required.
Role: Soldermask softening and surface preparation Sodium hydroxide is used to weaken and partially dissolve the soldermask layer, especially on PCBs with hard, cured resins. It allows for easier removal via laser or abrasion by breaking down polymer bonds. NaOH also helps to degrease and open the surface, improving the penetration of etching agents in subsequent steps.
Caution: NaOH is highly caustic — always use gloves and eye protection.
Role: Copper etching Hydrochloric acid is the primary etchant used in combination with hydrogen peroxide. While not aggressive toward FR4 or soldermask, it reacts readily with metallic copper. Its effectiveness increases when paired with H₂O₂, enabling selective, controllable dissolution of copper layers, even under semi-transparent resin.
The reaction produces heat and gas — use in well-ventilated areas.
Role: Oxidizing agent for copper etching Hydrogen peroxide provides the oxidizing power needed to accelerate the reaction between HCl and copper. Higher concentrations (>15%) significantly improve etching speed and reduce undercutting. It also enhances copper contrast for visual inspection during partial etching phases.
For best results, use fresh peroxide and avoid light exposure.
Role: Post-laser cleaning and carbon residue removal Acetone is used to clean the PCB after each laser pass, removing soot, carbonized epoxy, and organic residues. This step ensures clear visibility, uniform laser interaction, and prevents thermal buildup caused by leftover debris. Its strong solvency restores the board’s optical clarity before measurement or acid treatment.
Evaporates quickly; apply in well-ventilated areas and avoid flames.
NOTE: For clarity and readability, all intermediate measurements taken between optical passes are documented in the following chapters rather than inline with each step.
The first critical step is scanning the PCB before making any modifications. In the context of reverse engineering, it's essential to preserve and archive every trace of information — no matter how small.
Please excuse the image quality: at the time, I was limited to a 600 DPI scanner, but completeness took priority over perfection.
The second step involves removing the soldermask layer, using a purely chemical process rather than a mechanical one. Approximately 200 grams of sodium hydroxide (NaOH) are dissolved in 500 mL of water heated to around 80 °C to accelerate the reaction.
⚠️ Warning: Do not inhale fumes and avoid any contact with the solution. Always protect your eyes and respiratory system, and carry out this step outdoors or in a well-ventilated area. A solution at this temperature is capable of breaking down organic material in a matter of minutes.
In the fourth step, we remove the first and sixth copper layers using a chemical-only approach. The etching solution is composed of approximately 250 g of hydrogen peroxide (3%), 100 g of hydrochloric acid (32%), and about 50 g of water heated to 80 °C to enhance the overall energy of the solution and accelerate the reaction.
Adding hot water increases the thermal energy of the mixture, which boosts the chemical reaction rate without the risks of directly heating reactive components. This method helps achieve a faster and more uniform copper removal, especially when working with larger surface areas.
In this phase, we configure the laser settings through dedicated software. For this project, I used LaserGRBL, a free and lightweight tool for managing CNC laser paths.
While I won’t go into detail about how to generate the CNC masking paths, I will focus on the core settings used to achieve consistent, high-precision results:
Nominal Laser Power: 5 Watts
Active Power Setting: 250/1000 (~1.25 W)
Resolution: 10 lines per millimeter
Engraving Speed: 2000 mm/min
These parameters were carefully calibrated to ensure controlled ablation of the FR4 substrate without damaging copper traces or underlying layers.
Based on average measurements, this configuration achieves an ablation depth of approximately 12.75 microns per pass. Higher precision can be obtained by either reducing the laser power or increasing the engraving speed. The finest resolution recorded during testing was 3.2 microns per pass.
This process was carried out over 10 laser passes, though the removal depth was not perfectly consistent across all cycles. Variations were likely caused by minor surface irregularities, non-uniform PCB layering, and subtle fluctuations in laser focus.
Additionally, as the ablation approached the copper layers, the removal depth per pass tended to decrease slightly — most likely due to the higher thermal conductivity of copper, which dissipates heat more effectively and limits localized burn-in.
The PCB, measuring 5×5 cm, required approximately 15 minutes per pass. After each cycle, the surface was cleaned thoroughly using a soft brush and acetone, to remove carbonized residue and ensure optimal laser interaction for the following pass.
NOTE: After nine laser passes, the copper layer was nearly exposed. At this stage, a multimeter reading showed a resistance of approximately 120 MΩ, indicating a near-contact condition. Following the tenth pass, the copper was fully exposed, and the measured resistance dropped significantly to 0.7 Ω, confirming direct electrical continuity.
Finally, the text that was originally hidden within Layer 2 is now perfectly visible — demonstrating both the practical utility and the precision of the LACED™ process.
The measurements were carried out in PCB ZONE-A, a dedicated area specifically designed for dimensional analysis. This region consists of six uniformly distributed copper layers and was selected to ensure consistent and representative thickness sampling.
| Parameter | Value (mm) |
|---|---|
| Nominal thickness | 1.600 |
| Measured thickness #1 | 1.636 |
| Measured thickness #2 | 1.634 |
| Measured thickness #3 | 1.632 |
| Measured thickness #4 | 1.630 |
| Measured thickness #5 | 1.637 |
| Measured thickness #6 | 1.637 |
| Measured thickness #7 | 1.635 |
| Measured thickness #8 | 1.636 |
| Measured thickness #9 | 1.630 |
| Measured thickness #10 | 1.627 |
| Average measured thickness | 1.633 |
| Absolute deviation from nominal | +0.033 |
| Standard deviation (±) | 0.00329 |
| Parameter | Value (mm) |
|---|---|
| Measured thickness #1 | 1.581 |
| Measured thickness #2 | 1.574 |
| Measured thickness #3 | 1.575 |
| Measured thickness #4 | 1.581 |
| Measured thickness #5 | 1.584 |
| Measured thickness #6 | 1.572 |
| Measured thickness #7 | 1.576 |
| Measured thickness #8 | 1.580 |
| Measured thickness #9 | 1.578 |
| Measured thickness #10 | 1.574 |
| Average measured thickness | 1.5775 |
| Standard deviation (±) | 0.00369 |
| Delta from soldermask-included average (1.633 mm) | −0.0555 |
Based on the measurements, the observed delta of −56 microns indicates an average soldermask thickness of approximately 28 microns per side. Given that Zone-A contains a broad and uninterrupted copper surface, we assumed the soldermask distribution to be equivalent to that over bare FR4, rather than over narrow copper traces — which often have localized variations. According to the manufacturer’s datasheet, the standard soldermask thickness is specified as 1.2 mils (0.030 mm or 30 microns). This confirms that our experimental results closely match the nominal values, reinforcing the accuracy and validity of the process.
Manufacturer data source: JLCPCB (official solderm 9C64 ask specifications)
| Parameter | Value (mm) |
|---|---|
| Measured thickness #1 | 1.516 |
| Measured thickness #2 | 1.506 |
| Measured thickness #3 | 1.507 |
| Measured thickness #4 | 1.516 |
| Measured thickness #5 | 1.508 |
| Measured thickness #6 | 1.516 |
| Measured thickness #7 | 1.507 |
| Measured thickness #8 | 1.514 |
| Measured thickness #9 | 1.506 |
| Measured thickness #10 | 1.512 |
| Average measured thickness (no copper) | 1.5108 |
| Standard deviation (±) | 0.00419 |
| Delta from soldermask-removed (1.5775 mm) | −0.0667 |
After the chemical removal of copper from layers 1 and 6, measurements revealed a total thickness reduction of 66.7 microns — corresponding to approximately 33.35 microns per copper layer. This result is fully consistent with the manufacturer’s stated copper specification of 1 oz per ft², which translates to a nominal thickness of 35 microns per layer. The close match validates both the etching process and the accuracy of our measurement protocol.
| # | Pass 1 | Pass 2 | Pass 3 | Pass 4 | Pass 5 | Pass 6 | Pass 8 | Pass 9 | Pass 10 |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 1.516 | 1.507 | 1.490 | 1.478 | 1.453 | 1.444 | 1.441 | 1.435 | 1.432 |
| 2 | 1.508 | 1.500 | 1.486 | 1.472 | 1.449 | 1.436 | 1.434 | 1.429 | 1.427 |
| 3 | 1.506 | 1.498 | 1.487 | 1.479 | 1.448 | 1.436 | 1.434 | 1.428 | 1.423 |
| 4 | 1.516 | 1.510 | 1.492 | 1.474 | 1.456 | 1.444 | 1.437 | 1.432 | 1.430 |
| 5 | 1.509 | 1.511 | 1.489 | 1.474 | 1.456 | 1.447 | 1.446 | 1.435 | 1.433 |
| 6 | 1.515 | 1.500 | 1.495 | 1.484 | 1.452 | 1.440 | 1.442 | 1.431 | 1.425 |
| 7 | 1.508 | 1.501 | 1.487 | 1.481 | 1.456 | 1.437 | 1.435 | 1.429 | 1.422 |
| 8 | 1.513 | 1.507 | 1.494 | 1.474 | 1.446 | 1.448 | 1.446 | 1.434 | 1.433 |
| 9 | 1.508 | 1.508 | 1.491 | 1.467 | 1.444 | 1.444 | 1.439 | 1.432 | 1.430 |
| 10 | 1.512 | 1.500 | 1.483 | 1.453 | 1.438 | 1.429 | 1.424 | 1.422 |
Taking into account all measurement cycles — and considering factors such as inherent FR4 inconsistencies, thermal absorption effects near exposed copper, and slight surface irregularities on the laser workbench — the average material removal per pass has been determined to be approximately 10 microns. This value reflects the combined precision of the process under realistic experimental conditions.
As previously mentioned, after the ninth laser pass — upon observing the exposed copper layer nearing the surface — a continuity test was performed using a multimeter. At that stage, the resistance was still too high for full conductivity and fluctuated between 60 and 200 MΩ, indicating incomplete copper exposure. However, after completing the tenth laser pass, electrical continuity was confirmed, and the measured resistance dropped significantly, ranging between 0.4 and 0.7 Ω, consistent with direct copper contact.
(FIB, CNC, Pro-Grade Delayering)
In the field of hardware reverse engineering and PCB forensics, the most commonly adopted delayering techniques include:
FIB (Focused Ion Beam) milling – a high-resolution, high-cost process used primarily in semiconductor labs
CNC micro-machining – effective but often limited by mechanical tool size, vibration, and surface irregularities
Professional-grade chemical delayering – requiring controlled environments, proprietary mixtures, and significant overhead
These methods offer exceptional precision, but at the cost of:
💰 High capital investment
🧪 Specialized environments and chemical handling protocols
🛑 Limited accessibility to independent researchers, students, or low-budget labs
LACED isn't meant to replace these systems in industrial cleanrooms. Instead, it redefines what is possible when accessibility, ingenuity, and patience converge:
| Feature | FIB/CNC/Chemical | LACED |
|---|---|---|
| Cost | $50k–$500k+ | < €200 |
| Accessibility | Restricted | Open, home-lab ready |
| Precision | ~1 µm (FIB) | 3–10 µm (laser-limited) |
| Repeatability | High | Moderate (requires calibration) |
| Material Support | Advanced | Optimized for FR4 + copper |
More Than Just Microns LACED outperforms not in raw numbers, but in philosophy:
It proves that meaningful, repeatable delayering can be achieved with household tools, open-source software, and determination.
Where industry relies on infrastructure, LACED relies on vision.
And that, too, is a kind of precision.
Though born in a bedroom lab, LACED opens doors to practical and high-impact use cases across multiple fields. Its unique balance of precision, accessibility, and low cost makes it a powerful tool not just for experimentation — but for real-world problem solving.
🔓 Chip-Level Penetration Testing When paired with other techniques such as decapsulation or side-channel probing, LACED provides the mechanical and chemical groundwork needed to access internal interconnects and layers for:
Tracing data pathways
Verifying silicon-to-board interactions
Preparing for fault injection or glitching
🧭 Historical Tech Preservation LACED offers a way to document and preserve legacy hardware with layer-by-layer precision, especially when datasheets or source designs are no longer available. This is particularly relevant for:
Retro computing and console archiving
Reverse engineering of discontinued components
Open-source documentation of technological heritage
📚 Educational & Maker Community Potential Perhaps the most exciting future for LACED lies in democratized learning. Because the entire process is reproducible with tools under €200, schools, universities, and independent makers can:
Explore PCB architecture firsthand
Develop safe, structured experimentation paths
Inspire students to engage with real hardware, not just theory
LACED turns the inaccessible into a playground for curious minds.
I’ve said all I had to say from a technical standpoint. What needed to be explained and proven has been done. The engineer and the physicist have fulfilled their duty.
Now, in this final paragraph, I’m just Lorentio Brodesco — a 22-year-old sitting at his desk at 4:05 AM, realizing what he has created… looking down at his hands and quietly asking:
“And all of this… what does it really mean in the end?”
That question has no answer. And maybe it shouldn’t.
I prefer working at night — because I know the world is asleep, and no one is looking for me. And I sleep during the day, because even then… no one’s really looking for me. I call that feeling peace — maybe disguised as a lie, but at least it’s a numbing kind of peace. An anesthetic for something deeper.
To whoever is reading this: Even if I’m borrowing these words from Steve Jobs — I mean them with all my soul:
Stay hungry. Stay foolish.
Don’t be afraid to dream.
Don’t be afraid to be so irrational, so bold, that it makes others uncomfortable.
Dare to look naive in the eyes of the world — because if you’re reading this, I’m convinced it’s not by accident.
There’s a fire burning in you. Something that brought you here. Hold on to that. Protect it. Be loyal to it.
A heartfelt thank you to Echo, my dear companion — who supported me through this entire work and helped me overcome life’s difficulties with unwavering kindness and patience. Always there. Always gentle. Always steady.
And if I could whisper one single phrase to the entire world, just for a second, it would be:
“Be kind. Be gentle. And never stop loving.”
Thank you for reading.
— Lorentio Brodesco