WS2812B-V5
1-wire RGB · WorldSemi · ws281x · revision of WS2812B
- Channels
- RGB
- Bits / channel
- 8
- Supply ≠
- 3.7–5.3 V
- PWM
- 2 kHz
- Backup data
- no
- Packages
- 7
Datasheets disagree
The sources we hold state different values for the facts below. Rather than pick one quietly, both sides are listed with where each was read. Check the datasheet for the exact part number and revision you are buying. How this is shown.
Supply voltage
V5 and V6/V7 state different supply ranges, so the minimum depends on which revision you have.
| Source | Says | Page |
|---|---|---|
| ws2812b-v5-c114586 | VDD +3.7 to +5.3 V | 2 |
| ws2812b-v6-c52917433 | VDD +3.3 to +5.3 V | 3 |
| ws2812b-v7-c55109525 | VDD +3.3 to +5.3 V | 3 |
What we show. 3.7 V (the V5 sheet) is shown because it is the safer minimum; V6 and V7 variants state 3.3 V on their own sheets.
A silicon revision of the WS2812B, modelled here as its own protocol rather than a footnote because it changes timing. Same part number on the reel, same footprint, different behaviour - which is exactly the kind of thing this catalogue exists to make visible.
Refresh & flicker
The V5 sheet states “scan frequency is of 2KHz” and “Refresh Frequency updates to 2KHz, Low Frame Frequency and No Flicker appear in HD Video Camera”; the V6 sheet says the same and the V7 sheet gives the port scanning frequency as 2 kHz. Pre-V5 silicon is described as “not less than 400 Hz”. Supply is 3.7-5.3 V in the V5 sheet, 3.3-5.3 V in V6 and V7 (V7 also raises the per-channel drive from 12 mA to 16 mA).
Timing
Sources disagree. This is normal and worth understanding: FastLED's numbers are tuned to drive real parts reliably, including clones, while the datasheet gives the manufacturer's nominal figures. Both are recorded here rather than picking one and hiding the conflict.
| Source | T0H | T0L | T1H | T1L | Period | Rate | Reset | ± |
|---|---|---|---|---|---|---|---|---|
| fastled ↗ | 225 ns | 1000 ns | 580 ns | 645 ns | 1225 ns | 816 kHz | 280 µs | — |
| FastLED carries a dedicated V5 entry because the timings genuinely differ from earlier silicon. The datasheet set below agrees: its windows are wider and the reset is above 280 us. | ||||||||
| datasheet ↗ | 300 ns | 790 ns | 790 ns | 790 ns | 1250 ns | 800 kHz | 280 µs | — |
| The datasheet gives RANGES and the values here are their midpoints: T0H 220-380 ns, T0L 580 ns-1 us, T1H 580 ns-1 us, T1L 580 ns-1 us, reset above 280 us. The 1250 ns period is the minimum data cycle stated in the V6/V7 sheets (the V5 sheet is silent). FastLED’s 225/1000/580/645 ns figures all sit inside these windows. Note the old WS2812B sheet’s 400/850/800/450 ns +/-150 ns is NOT the same table: V5 widens every window and extends the reset from 50 us to 280 us. | ||||||||
Clones & compatibility
“Compatible” is never stated bare here — each relationship carries how strong the claim is, and what bites if you assume too much.
-
near clone
WS2812B
Same package, pinout and command set, but the reset/latch period is extended to more than 280 us (V5, V6 and V7 sheets; FastLED also uses 280 us). Code written against the old 50 us figure can glitch. This is the most common cause of "my new reel of the same part flickers". source ↗
-
protocol compatible of this
AOP single-wire IC (high speed mode, 280 us reset)
T0H 220-380 ns, T0L 580-1000 ns, T1H 580-1000 ns and a 280 us reset are the windows the ws2812b-v5 protocol records; the AOP T1L window is wider (0.22-1.0 us against 580 ns-1 us). The sheets never mention WorldSemi.
-
near clone of this
WS2812D-V6
Same 280 us reset, 2 kHz scan frequency and the same T0H window (220-380 ns) as the WS2812B-V5, whose sheet the catalogue now holds. The V6 dome's T1L window is SHORT (220-380 ns) where the V5's is 580 ns to 1 us (and its T1H and T0L windows are 750 ns to 1 us against the V5's 580 ns to 1 us), so it is not the same timing table. The windows overlap, so a V5 driver may work, but that is not a tested claim.
-
protocol compatible of this
WS2812E
Identical datasheet timing windows, pin count (4) and supply range. Independent measurement has not been done; E and B differ in IC drive current (12-12.5 mA) and packages (the E exists in 1313, 2020, 2427, 3535-class MINI and 4020).
Library support
- FastLED
WS2812B_V5↗- Adafruit_NeoPixel
NEO_GRB + NEO_KHZ800- WLED
WS281x
FastLED exposes this as a distinct chipset. Adafruit_NeoPixel does not distinguish it, relying on its own conservative latch period.
Parts using this protocol
Dating evidence
Evidence, not a release date. A library commit proves the part existed by that date and says nothing about how long before.