Can You Still Build the BBC micro:bit V1?
Can you still build the BBC micro:bit V1?
The Micro:bit Educational Foundation revised this board in 2018 to get off an end-of-life magnetometer, the NXP MAG3110, published the reason, and laid out the replacement footprint for a second manufacturer’s part. Neither published V1 revision is buildable as drawn today. The lines whose lifecycle status stops it are the two Everlight LED lines and the 16 MHz timing reference, all three carried through the 2018 revision unchanged.
The three numbers
The counts behind the verdict, per revision:
Gone on V1.3B
audited lines obsolete, end of life or no longer manufactured. On V1.5 it is 3 of 11, and all three are lines V1.5 inherited unchanged from V1.3B.
Replacement NXP named
orderable for the MAG3110, in its own 2018 discontinuation notice: the FXOS8700CQR1. That replacement now reads Obsolete at the distributor.
GPIO drive the display
of the nRF51822’s GPIO. All 31 carry a signal net, and the twelve that drive the display drive an end-of-life part.
The ledger · 13 lines
Thirteen distinct lines across the two revisions. Lines 5 and 6 are V1.3B only, line 7 is V1.5 only. Three lines carry the blocking statuses: the 25 display LEDs are one Everlight orderable placed twenty-five times, discontinued at DigiKey with both listed substitutes at 0 available; the yellow indicator carries the same PCN field; and the 16 MHz reference reads Obsolete. Three further lines cannot be bought through either major distributor.
| # | Desig | Component | Role | Rev | Read from | Status |
|---|
Tap any flagged row to open its finding; use the filter to isolate a status. The Rev column marks which of the two published revisions carries the line, and the Read from column names the kind of source each status came from. Method note 01 below sets out which BOM lines are counted.
The same 13 lines, on the boards they came off
Every audited line above is drawn where it sits, on whichever of the three faces carries it. Hover, tap or tab to a part to read its finding. Eleven of the thirteen are on the boards; the two no photograph could place are in the strip underneath.
The keystone
Nordic prints “Not Recommended for New Designs” on the nRF51822 page. Its “Where to Buy” block is headed with the orderable NRF51822-QFAA-R7 where both BOMs name NRF51822-QFAA-R, and no document read here relates the two part numbers. The ledger’s Obsolete and End of life entries sit elsewhere, on the two Everlight LED lines and the timing reference.
The SoC carries 31 GPIO, P0.00 through P0.30, and every one carries a named signal net. Nine 220 ohm series resistors sit on the columns with no LED driver between the SoC and the matrix, a 5×5 array wired as a 3×9 matrix, rows of 9, 7 and 9 anodes.
The full GPIO assignment, parsed from the V1.3B schematic
| Function | Schematic net names | nRF51822 GPIO | Count |
|---|---|---|---|
| Display columns | COL1 to COL9 | P0.04 to P0.12 | 9 |
| Display rows | ROW1 to ROW3 | P0.13 to P0.15 | 3 |
| Edge connector only, no second function | P0.16, P0.18, P0.20 | P0.16, P0.18, P0.20 | 3 |
| Large ring pads | PAD1, PAD2, PAD3 | P0.03, P0.02, P0.01 | 3 |
| SPI | MOSI, MISO, SCK | P0.21, P0.22, P0.23 | 3 |
| Sensor interrupts | MMA8653FC_INT2, MMA8653FC_INT1, MAG3110_INT1 | P0.27, P0.28, P0.29 | 3 |
| I2C | SCL, SDA | P0.00, P0.30 | 2 |
| Buttons | BTN_A, BTN_B | P0.17, P0.26 | 2 |
| Interface UART | TGT_TX, TGT_RX | P0.24, P0.25 | 2 |
| Target reset | TGT_NRESET | P0.19 | 1 |
| Total GPIO | 31 | ||
Six of the nine columns also land on the edge connector, as COL1R, COL2R, COL3R, COL7R, COL8R and COL9R. The Foundation states the effect: “Some of the Columns appear on the edge connector, so if you want to use extra GPIO pins, you have to disable the display in software.” The matrix is also the light sensor. It senses ambient light “by repeatedly switching some of the LED drive pins into inputs and sampling the voltage decay time”, so a different LED in those 25 positions changes a sensor as well as an indicator.
The forecast
The Foundation documented its response in public: V1.5 “was created to address the end-of-life of the micro:bit’s motion sensors.” NXP’s notification 201804001DN gave the MAG3110FCR1 a last-time-buy of 28 July 2018. The revised board shipped that October, after the window had closed.
A single V1.5 footprint is laid out for two motion sensors from different makers: ST’s LSM303AGR and NXP’s FXOS8700CQ, resolved in software at runtime. The Foundation records one leg as fitted: “To date, all V1.5 boards have been manufactured with the LSM303AGR.”
LSM303AGR
The leg the Foundation records every V1.5 board as carrying to date. Reads Active at the distributor, and its orderable is named in ST’s 2025 manufacturing-reshaping early notification.
FXOS8700CQ
The leg the footprint was laid out to accept, and which the Foundation records no board as carrying to date. NXP now publishes it as “NO LONGER MANUFACTURED”.
The NXP leg of that footprint is now no longer manufactured, and so is the orderable NXP named as the MAG3110 replacement. Notification 201804001DN names FXOS8700CQR1; NXP now publishes the FXOS8700CQ as “NO LONGER MANUFACTURED”, with no date given for that second discontinuation.
The fitted ST LSM303AGRTR reads Active, alongside a PCN field “Manufacturing reshaping program 30/MAY/2025”. The notification behind that field is ST PCI CRP/25/15383, which names the orderable, puts footprint reshaping at end 2027 and records “no impact” on form, fit and function. No termination notice against it was located here.
Only the sensor lines changed in the revision; every other audited line carried over unchanged. Both Everlight lines picked up a PCN Obsolescence/EOL field dated 14 December 2021, about three years after the revised board shipped. The repository publishes no V1 revision after V1.5.
NXP names no MAG3110 replacement on its product page, though the accelerometer beside it gets the FXLS8974CF. For magnetic sensing the FXOS8700CQ page points only to the NMH1000. NXP’s NMH1000 datasheet opens “The NMH1000 is a hall effect magnetic field switch” and describes a bi-state output, a different class of device from the 3D magnetometer it stands in for.
Verdict
Changing the display is a sensor change, for the reasons section 03 sets out. A replacement LED is qualified on forward voltage at the drive current the 220 ohm resistors set, on multiplexed brightness across a 3 by 9 refresh, and on the ambient-light decay behaviour the same pins implement.
A move off the nRF51822 is a new board. Of the two changes, it is the one no current status forces, and a replacement is qualified against all 31 GPIO the design uses, the radio, the three-inductor matching network between the SoC’s ANT1 and ANT2 pins and the PCB antenna, and the SWD path to the interface processor. The Foundation’s own migration went to a different nRF52 part on a new design.
Sheet 4 carries the annotation “Needs to be very low Vf” against each Schottky diode. They perform the supply OR-ing between the interface rail and the battery rail, so forward drop at the board’s operating current is the parameter a substitution is judged on.
For a buyer, two lines carry the action
The display line
Both DigiKey substitutes read 0 available, so a build depends on independent-market stock the buyer qualifies, at 25 placements per board, and known only through a distributor’s PCN field.
The ST sensor on V1.5
Active, but named in ST PCI CRP/25/15383 (reshaping by end 2027). Re-check LSM303AGRTR for a PCN or PTN on the 2026-11-04 cadence.
What this audit does not claim. No part named is validated as a form-fit-function replacement; each is a candidate to qualify in your own hardware, firmware and test. Statuses are dated, and the missing listings are audit-date sourcing observations, not proof of discontinuation. Nothing here was measured on a bench.
Frequently asked
Which micro:bit revisions does this cover?
The two V1 revisions the BBC publishes, V1.3B and V1.5. The repository README states that “There are three versions of the BBC micro:bit board that have been produced, V1.3 and V1.5 and V2.” The README writes V1.3 where the directory, the BOM title block and every schematic sheet write V1.3B, and this audit uses the schematic’s designation. Sheet 5 carries a note reading “Changes from V1.04”, so a V1.04 existed and the repository does not publish it. V2 is out of scope here.
Does NXP name a MAG3110 replacement?
Not on the product page, which reads “NO LONGER MANUFACTURED” and offers nothing in its place. NXP did name one in the 2018 discontinuation notification that retired the part: the notice’s “Replacement Orderable Part#” column gives FXOS8700CQR1. That replacement orderable is itself Obsolete at the distributor, and NXP has separately discontinued the FXOS8700CQ. By contrast the MMA8653FC accelerometer page on the same site names a specific part in its place, the FXLS8974CF, whose own lifecycle status was not read in this audit.
Does the nRF52832 replace the nRF51822 on this board?
Nothing located in this audit presents it as a replacement. Nordic’s nRF51822 page carries the NRND marker and names no successor outside the nRF51 series. The two devices are built on different Arm cores, a Cortex-M0 against a Cortex-M4. Any move off the nRF51822 is qualified in the target design. The Foundation put the nRF52833 on V2 instead, and no revised V1 followed.
Why is the LED display a risk line when the resistors are not?
The display is one manufacturer’s orderable placed 25 times and its distributor record carries a PCN Obsolescence/EOL field. The 220 ohm resistors are a value, a tolerance and a package, and this board demonstrates the difference itself: the 10 ohm 0402 resistor is a Vishay part in the V1.3B BOM (CRCW040210R0JNEDHP) and a ROHM part in the V1.5 BOM (ESR01MZPJ100), at the same value, tolerance and package.
Did the V1.5 revision fix the obsolescence problem, and what happened to its second source?
It fixed the magnetometer, the part NXP named in its 2018 discontinuation notification, replacing both NXP motion sensors with one ST part and laying the footprint out to accept an NXP combined sensor as well, the FXOS8700CQ. The NXP option is now no longer manufactured, where the fitted ST LSM303AGRTR reads Active. No other audited line changed. Three years on, in December 2021, both Everlight LED lines acquired the distributor PCN Obsolescence/EOL field their ledger rows quote, and the 16 MHz timing reference reads Obsolete with no date given for when that status was set.
Can the board be built with the parts still available?
No. A build needs 25 of the display LED, whose distributor record carries a PCN Obsolescence/EOL field dated 14 December 2021 and shows both listed substitutes at 0 available, one of the yellow interface indicator, which carries the same PCN field and whose one listed substitute also shows 0 available, and two of the 16 MHz timing reference, which reads Obsolete. Three further lines have no authorized listing or cannot be ordered through either major distributor.
Why does the accelerometer get a named replacement on its page and the magnetometer does not?
Neither page, read on 2026-08-06, states a reason. What the pages record is that NXP names a replacement part for the accelerometer and names none for the standalone magnetometer, which matches the reason the Foundation gave for its own change: “many new designs are choosing combined magnetometer and accelerator chips, rather than standalone devices.”
Method & sources
01Scope
A part is audited if its BOM line resolves to one manufacturer’s orderable part number, whatever the part does. Every micro:bit BOM line names a manufacturer orderable, so the test is applied to the line’s function: a part is bucketed where the same function is bought from several manufacturers interchangeably. One of these BOMs buys its 0402 ceramic capacitors from Murata, Multicomp and Yageo at once. Bucketed on that basis: chip resistors, ceramic capacitors and chip inductors. The PCB, the label and the anti-static bag are not electronic parts.
02Counts
The pin, net and matrix figures were parsed programmatically from the 6-sheet V1.3B schematic PDF rather than counted by eye: 48 pins on the SoC symbol, 31 GPIO each with a named net, grouped into the section 03 table, and 9 + 7 + 9 = 25 anodes across the three rows.
03Provenance
The Read from column on the ledger, shown at tablet width and wider, records per line whether the status came from the manufacturer’s own page, an authorized distributor’s status field, or neither. The Nexperia ESD array is reported as an availability finding and is not converted into a lifecycle status. Where a distributor exposes a PCN Obsolescence/EOL field, that field is named as what it is; the Everlight PCN behind those two fields never surfaced, and nothing is quoted from it here. The Maker column reports three attribution discrepancies as they appear in the documents; none is reconciled here.
04Unconfirmed
NXP publishes MKL26Z128VFM4 as ACTIVE, and the BOM names MKL26Z128VFM4R, which has no separate manufacturer page and whose relationship to the base part no document read here states. The same gap applies to the NRF51822-QFAA-R7 on Nordic’s stock block against the BOM’s -R. Two further pairs are named at different levels by their own sources and are not related by any document read here: LSM303AGR in the Foundation’s text against LSM303AGRTR in the BOM and at the distributor, and FXOS8700CQ on NXP’s product page against FXOS8700CQR1 in the 2018 notice.
Re-check. Every lifecycle status can change after the audit date.
Sources · public, primary
- micro:bit Foundation, open hardware repository. Schematics and BOMs for both published V1 revisions: github.com/bbcmicrobit/hardware
- micro:bit Foundation, hardware documentation. V1.5 hardware description, schematic notes, the stated reason for the motion sensor change, and the V2 application processor
- Manufacturer notices and datasheets. NXP discontinuation notification 201804001DN (PDF), ST PCI CRP/25/15383 (PDF), and the NXP NMH1000 datasheet
Every lifecycle status in the ledger links to the record it was read from. Open a row to reach it.
Lifecycle, package and availability data here is vendor- and distributor-reported and has not been independently audited. The switches and the battery connector returned no distributor listing on the audit date, which records what a search returned and does not establish that either part is discontinued. No part or board named is validated as a form-fit-function replacement; each is a candidate to qualify in the reader’s own hardware, firmware and test. Statuses are dated and can change without notice.
About the author
Related reading
Editorially independent and from public sources only; lifecycle, pin and rating data is vendor-reported, not independently audited. Alternates are candidates to validate in your own design. Confirm every detail with the manufacturer before committing.