USB-to-serial bridge obsolescence: which parts are EOL & how to cross them
The USB-to-serial bridge, the USB-to-UART converter that presents a USB device to the host as a virtual COM port, is one of the most widely designed-in interface parts in electronics, and two of the most common are now obsolete. What decides each swap is the host driver and the USB VID/PID the host software binds to: a same-vendor successor keeps them; every cross-vendor part changes them. The footprint match does not tell you that. Suggested candidates to validate in your own design, not drop-in equivalents.
The family map
Every legacy part routes to a same-vendor go-forward that keeps the driver and USB identifiers, or out to a shared cross-vendor pool that changes all of them. The arrow is the whole decision.
VID/PID
family
Two ways it reaches end-of-life
USB bridges do not all reach end-of-life the same way, and the difference decides whether a last-time-buy even helps. Diagnose the mode first.
Foundry end-of-life
FTDI’s R-series notice places it at EOL with a last-time-buy of 1 June 2024 (passed), traces the discontinuation to the chip foundry (legacy silicon, a wafer-supplier price increase), and names the FT232RNL as the replacement. DigiKey lists it obsolete and no longer manufactured.
A last-time-buy bridges itDriver-support end-of-life
The older chip versions (HXA, TA, TB, HXD, RA, SA, EA) are EOL, and Prolific’s Windows driver refuses to load them on Windows 11, reporting a Code 10 fault: “PL2303TA DO NOT SUPPORT WINDOWS 11 OR LATER.” The maker-recommended part is the PL-2303GC.
A last-time-buy does NOT helpA last-time-buy bridges a foundry EOL; it does not help a driver-support EOL, where the block travels with the operating system, not with your stock. The fix is a supported part, not more inventory.
The four catches that decide every cross
Host driver & USB VID/PID Load-bearing
Anything keyed to a vendor’s Vendor ID (FTDI 0x0403, Prolific 0x067B, Silicon Labs 0x10C4), a device path built from it, or a vendor driver stack stops matching the moment the silicon changes vendor. Only a same-vendor successor keeps the binding.
Package & pinout
Only same-vendor successors keep the footprint (the FT232RNL stays SSOP-28). Every cross-vendor part is a board redesign: the CP2102N is QFN-28, the CH340G is SOP-16, and no cross-vendor part is a pin-for-pin match.
Which obsolescence mode
Foundry EOL versus driver-support EOL, which decides whether a last-time-buy actually bridges the gap, or buys stock that the operating system will still refuse.
Counterfeit exposure
The FT232R and PL2303 are among the most cloned USB bridges, and both vendors have shipped Windows drivers that acted on non-genuine parts. Buy only from franchised or traceable sources. A clone is a separate requalification.
The USB-to-serial bridge landscape
The common parts a cross-reference returns, mapped by the axis that actually decides the swap, the host driver and USB VID/PID, plus package and lifecycle. Confirm every lifecycle standing at the maker field before you design a part in.
| Part | Vendor | Host driver / USB VID | Package | Lifecycle |
|---|---|---|---|---|
| FT232RL | FTDI | FTDI VCP/D2XX, 0x0403 | SSOP-28 | Obsolete foundry EOL, LTB Jun 2024 |
| FT232RNL | FTDI | FTDI (same as RL) | SSOP-28 | Active FTDI’s named replacement |
| PL2303 HXA/TA/HXD… | Prolific | Prolific, 0x067B; blocked on Win 11 | varies | EOL chip versions |
| PL2303GC | Prolific | Prolific family, 0x067B | SSOP-28 / QFN / SSOP-16 | Active maker-recommended |
| CP2102N | Silicon Labs | SiLabs VCP, 0x10C4 | QFN-28 | Active |
| CH340G | WCH | WCH driver, own VID | SOP-16 | Active |
| CY7C65213 | Infineon | Infineon, own VID | SSOP-28 / QFN-32 | Active FT232R pin-match |
The load-bearing column is the host driver and USB VID: the same-vendor successors (FT232RNL, PL2303GC) keep it; every cross-vendor part changes both the driver and the identifiers the host binds to. Lifecycle, package and driver data is vendor/distributor-reported, not independently audited.
The per-part teardowns
Each legacy part has a full forensic teardown: the fit-catch on every cross, the comparison table, and the procurement path:
FTDI · foundry EOL
FT232RL replacement: FT232RNL, CP2102N, CH340 & CY7C65213
The end-of-life FTDI part, the same-vendor silicon-revision cross, and the cross-vendor driver change, compared in full.
Prolific · driver-support EOL
PL2303 replacement: the FT232RNL, CP2102N, CH340G & PL2303GC crosses
The driver-support EOL, the Windows 11 block, and why a last-time-buy does not fix it, covered in the USB-UART cross comparison.
Silicon Labs · contested NRND
CP2102 vs CH340, FT232RL and PL2303: the USB-UART bridge crosses compared
The distributor lifecycle flag reads NRND while Silicon Labs commits to production through March 2029, a status disagreement worth reading before any of the three crosses.
How to approach a USB bridge obsolescence
No single winner across the family, a fit to your design and, above all, your host software:
Stay same-vendor Lowest friction
The FT232RNL keeps FTDI’s driver and identifiers (FTDI calls it pin-compatible in nearly 100% of designs; still a silicon-revision requalification); the PL2303GC keeps Prolific’s driver family. Neither disturbs host software keyed to the vendor VID/PID.
Cross-vendor is a software project
Not a part swap. Budget a new driver, new USB VID/PID, a re-created GPIO/config scheme, and a board redesign. Footprint and pinout differ on every cross-vendor part.
Diagnose the mode first
Foundry versus driver-support: a last-time-buy only helps the foundry case (FT232RL), not the Windows 11 driver block on the PL2303.
Buy franchised or traceable
Both families are heavily counterfeited and both vendors’ drivers act on non-genuine parts; a clone is a separate requalification, not a like-for-like source.
Requalify in your own system
A matching function and data rate do not qualify the part; validate host enumeration and any software keyed to the old VID/PID or vendor API.
Frequently asked questions
Which USB-to-serial bridges are obsolete?
The FTDI FT232RL is end-of-life (last-time-buy 1 June 2024 passed, obsolete at distribution), and the Prolific PL2303’s older chip versions (HXA, TA, TB, HXD, RA, SA, EA) are end-of-life with their driver blocked on Windows 11. Confirm the exact part and version you use; the FT232RNL and the PL-2303GC are the makers’ own current parts.
Are USB-to-UART bridges interchangeable?
Not as drop-ins. They share a function, but each vendor’s part uses its own host driver and USB VID/PID, and the packages and pinouts differ, so any host software keyed to the old vendor’s identifiers or driver must be reworked and cross-vendor swaps are a board redesign.
What is the difference between FTDI, Prolific, Silicon Labs and WCH USB bridges?
They perform the same USB-to-serial function, but they differ where it matters for a swap: the host driver and the USB Vendor ID (FTDI 0x0403, Prolific 0x067B, Silicon Labs 0x10C4, WCH its own), plus the package. Host software binds to those identifiers and the vendor driver, so moving between vendors is a driver and provisioning change, not just a new part.
Why did my PL2303 stop working on Windows 11?
Because the older PL2303 chip versions are end-of-life and Prolific’s Windows 11 driver refuses to load them, showing a Code 10 fault with the message “DO NOT SUPPORT WINDOWS 11 OR LATER.” This is a driver-support decision that travels with the operating system, so buying more of the same EOL chip does not fix it; Prolific directs affected designs to the PL-2303GC.
What replaces the FT232RL and the PL2303?
Each has its own set of real crosses with a specific catch; the two teardowns linked above work them in full. In short: the same-vendor go-forward parts are the FTDI FT232RNL and the Prolific PL-2303GC (each keeps its vendor driver and VID/PID), and the shared cross-vendor options are the Silicon Labs CP2102N and the WCH CH340G (both change driver, identifiers and footprint). Treat all as candidates to validate.
Keep reading
Sources · public, primary
- FTDI R-series end-of-life notice, last-time-buy 1 June 2024 (PDF)
- FTDI FT232RNL product page + AN_560 migration note
- DigiKey FT232RL-REEL (part status: obsolete, no longer manufactured)
- Prolific PL2303 Windows driver / EOL chip-version support notice
- CHIRP bug 9509: “PL2303TA DO NOT SUPPORT WINDOWS 11 OR LATER” (driver error string)
- Silicon Labs CP2102N datasheet (PDF)
- WCH CH340G (LCSC product page, in production)
- Infineon: Replacing FT232R with CY7C65213
- Ars Technica: Windows Update bricking counterfeit FTDI chips (2014)
- Hackaday: FTDI drivers break fake chips again (2016)
Lifecycle, package, driver and status data is vendor- and distributor-reported, not independently audited. Alternates are candidates to validate in your own design; confirm every detail with the manufacturer before committing.
About the author