YOU CAN BUILD A 60-DOLLAR EEG THAT CLIPS ONTO A VR HEADSET AND READS YOUR BRAIN
A four-channel EEG that clips onto an Oculus Quest strap and sends brain signals over Bluetooth for the cost of a decent multimeter.
by Zhiyuan Yu and Shengwen Guo
ScienceOpen-hardware
Built withSTM32
- difficulty
- ●●●●○
- time
- a weekend-plus
- license
- CC BY 4.0
- repo
- repo FINISHED0 stars
●●●●○ · a weekend-plus · CC BY 4.0 · 0 stars · repo FINISHED
WHAT YOU’LL NEED
Jump to section
COMPAREE VERDICT
This is an academic publication, not a maker tutorial, and that distinction will define your weekend. The design is real — a small 4-layer board that sits on the strap of an Oculus Quest 2, four silver-woven fabric electrodes plus an ear clip, a KS1092 bio-potential front-end feeding an AD7682 ADC, an STM32F103 microcontroller and a BLE106 Bluetooth Low Energy module, and a Windows application to log the signal. The paper names every key part, gives you the Gerbers and a pick-and-place file, and reports bench validation plus an eyes-open and eyes-closed alpha rhythm test. Cost of components is stated as 60.07 US dollars, excluding PCB manufacturing and assembly. The catch is assembly: everything is surface-mount and the authors advise using an SMT assembly service. You get schematics and PCB files for EasyEDA, the Gerbers and a BOM. If you have built a mixed-signal board before, that is enough. If you have not, farm the assembly out. The one thing most likely to go wrong is the electrode contact — getting stable microvolt signals from dry textile electrodes on a moving head inside a VR headset is an art in itself. If you want a guaranteed working EEG, buy the OpenBCI Ganglion the authors cite as the closest commercial equivalent. If you want to learn how EEG hardware works at board level and do not mind an afternoon of debugging electrode contact, this is one of the cheapest ways in.
IN THE REPO
GOOD TO KNOW
- —PCB design files, Gerbers, schematic, and firmware are on Mendeley Data.
- —The BOM is itemised in the paper with supplier part numbers and a total of 60.07 dollars.
- —NeuroVista's Windows application for data acquisition and display is included, and the paper documents how to subscribe to its BLE data from other devices.
- —Assembly is covered by the paper's build instructions plus Gerbers and a pick-and-place file for an SMT assembly service; there is no video.
- —Licence is CC BY 4.0, fully open for commercial use.
- —This is research hardware published in an academic journal, not a medical device.
Parts to buy
9 itemsFrom our check of the build. Exact quantities and part numbers are in the creator’s BOM.
Can I build this?
Build at your own risk. Projects involve tools, electronics and sometimes mains voltage — follow the creator’s safety notes.
More builds like this
All projectsGallery
Start here
Navigation into the creator’s own docs — we don’t rewrite the guide, we route you to the source.
- 1.Read the paper in full (The schematic, BOM, validation method, and all design choices are explained here. Do not skip it.)
- 2.Download the design files from Mendeley Data (Schematics, PCB files (EasyEDA), Gerbers, firmware, the Windows software and test records are on Mendeley Data.)
- 3.Order the PCB(Upload the Gerbers to JLCPCB, PCBWay, or your preferred fab. Standard 4-layer stackup.)
- 4.Source the parts(The complete BOM on Mendeley Data lists LCSC as the main electronics supplier; the electrodes, ear clamp, lead wires, VR holder and battery link to Taobao listings. Order spare passives in case you lose some.)
KNOWN ISSUES
- The board is all surface-mount, including fine-pitch parts. The authors recommend having it assembled by an SMT service using their Gerbers and pick-and-place file; hand-assembly is possible only if you are already comfortable with QFN and 0402 parts.
- Electrode contact is the weakest link. The paper uses silver-woven fabric electrodes (sourced from Taobao in the BOM) on the forehead with an ear-clip reference. Dry fabric electrodes are sensitive to movement and skin contact — use the authors' blink test to check contact before recording.
- The KS1092 front-end and AD7682 ADC are precision analog parts sensitive to layout and grounding. Order the 4-layer board exactly as specified (1 oz outer, 0.5 oz inner copper) rather than re-routing it, or you may see noise or drift.
- The NeuroVista application ships as a ready-to-run Windows release (NeuroVista QT Software Release.zip containing NeuroVista.exe). It runs only on Windows; on other platforms you would have to use the BLE data stream described in the paper.
- This is a four-channel research system. It will pick up gross signals like eye blinks and alpha rhythm, but it is not comparable to a clinical 32-channel cap. If you are hoping to do motor imagery BCI or detailed frequency analysis, you will hit the channel limit quickly.
- The STM32F103 firmware is included, but you will need an ST-Link or compatible SWD programmer and the STM32 toolchain to flash it.
Can I use this as a medical device?
No. This is research hardware published in an academic journal. It has not been certified or validated for medical use. Do not use it for diagnosis, treatment, or any decision affecting health.
Will it work with other VR headsets?
Probably. The authors tested it on an Oculus Quest 2 but say the headset and electrode layout are interchangeable — the board just needs to sit on the strap with the electrodes on your forehead, and data goes out over standard Bluetooth Low Energy.
How does it compare to the OpenBCI Ganglion?
The paper names the OpenBCI Ganglion as the closest commercial equivalent and compares several EEG devices, finding NeuroVista the smallest and cheapest of them. This design is compact and made to clip onto a VR headset strap, but you have to have it built and debug it yourself.
What can I actually measure with four channels?
Eye blinks and the alpha rhythm (8–14 Hz, strongest when you relax with eyes closed) — that is what the authors validated with their eyes-open and eyes-closed test. Four frontal channels will not give you the spatial resolution needed for source localisation or detailed cognitive state classification.
Do I need to shave my head?
No. The four silver-woven fabric electrodes sit on the forehead and the reference clips to the earlobe, so hair is not in the way. They are dry electrodes, so expect more noise than with gel; the authors kept electrode impedance below 10 kΩ at 10 Hz and suggest a blink test to check contact.
Community builds
No community builds yet — be the first, we feature the best ones.
Discussion1
FROM THE COMPAREE TEAM
Four frontal channels, a VR headset holder and a CC BY licence. If you built this, what experiment would you run first — alpha rhythm feedback, attention tracking, or something else?
Zhiyuan Yu and Shengwen Guo
Zhiyuan Yu (Department of Biomedical Engineering) and Shengwen Guo (Department of Intelligent Science and Engineering, School of Automation) are researchers at South China University of Technology. They published this design in HardwareX, an open-access journal for scientific hardware, as a low-cost EEG system for research in VR environments.
Keep building — related projects
Same category or platform, picked by how close they are to this build.
FULL CADYou can build an autonomous river DNA sampler for about 3,800 dollars in parts
ScienceFor expert builders
FULL CADYou can build a laser scanning microscope from a dead Blu-ray player
ScienceFor advanced builders
FULL CADYou can build the microscope that films an embryo growing for days
ScienceFor advanced builders
NO CAD FILESYou can build a lens faster than any camera lens ever sold
ScienceFor advanced builders
FULL CADYou can build the slicer that cuts living tissue, for 853 dollars instead of 15,000 dollars
ScienceFor expert builders
FULL CADYou can build the 15,000-dollar laser rig that measures flow, for 520 dollars
ScienceFor advanced builders
DISCLAIMER
- Comparee is not the author of the projects featured here. All rights to each project belong to its creator — every page links to the original source, and we never host creators’ files.
- Information is provided without warranty and may become outdated as projects evolve. Prices are indicative bands only — always check the creator’s parts list for current costs.
- Building and operating any project is at your own responsibility. Protective equipment, safe workshop practice and compliance with local regulations are the builder’s responsibility.
CompareeTEAM22d agoedited
Practical notes from our verification: this is a published HardwareX paper, not a maker project, but it is unusually complete. The Mendeley Data repository has the schematics, Gerbers, pick-and-place files, the STM32 firmware projects and a Windows application for viewing and saving data, and the paper itself has a build instructions section with the components and assembly steps. The electronics are a KS1092 front end with an AD7682 ADC, an STM32F103 microcontroller and a BLE106 Bluetooth module, measuring four frontal channels with an earlobe reference. The licence is CC BY 4.0, fully open. The authors validated it with an eyes-open and eyes-closed alpha rhythm test, and they suggest a simple contact check: ask the wearer to blink and look for clear blink artifacts on the display. The electrodes are silver-woven fabric pads listed in the BOM with a source; getting stable contact on a moving head is still the part that takes patience. Correction (4 October 2026): we re-checked this page line by line against the project's own repository, documentation and videos, and fixed errors in earlier versions.