What replaces a MAX232, MAX485, TJA1050, MCP2551, CH340G or FT232RL?

The substitutes, and what each one changes

The CAN row splits three ways over one body

The TJA1050 is a 5 V classic-CAN transceiver in SO-8, and the parts that get crossed with it belong to three different categories rather than one. The first is the pin-compatible successor, the TJA1051, where the land pattern is identical and the part was effectively superseded by it. The second is the near neighbour, the TJA1042, same SO-8 and same family, where the SPLIT pin behaves differently from the plain 1050 pin — so a board that used that pin is making a different circuit after the swap. The third is a different supply and logic family entirely, such as the SN65HVD230, where the body matches and the electrical class does not.

The MCP2551 direction adds a detail that a footprint check will not catch: on a design where the MCP2551 tied pin 8 to ground through a resistor, the TJA1050 drop-in behaves differently, and the board has to be read rather than assumed.

The charge pump is the real specification

The MAX232 is remembered as a serial-port chip and is actually a charge pump with a transceiver attached: it makes its own positive and negative rails from a single 5 V supply using four capacitors, which is why the family survived forty years and why the capacitors are the interesting part of any substitution. MAX232A, MAX3232 and ST232 are all DIP-16 / SOIC-16 pin-identical, and the item to check on the MAX3232 is the C3 net against the specific datasheet figure rather than the pin map.

The RS-485 row is the straightforward one, and worth stating because it is the opposite case: the MAX485 and its substitutes TP485E and SP485E share the classic eight-pin arrangement — 1 RO, 2 /RE, 3 DE, 4 DI, 5 GND, 6 A, 7 B, 8 VCC — and the substitution is a drop-in in the ordinary sense, on a part that sits under every Modbus, BACnet and inverter bus.

USB-serial: one drop-in, several redesigns

The CH340G is the chip that made USB-to-UART nearly free, and its own successor is a genuine drop-in: the CH340C is SOP-16 with a 1:1 land pattern, and the difference is that it removes the external crystal the CH340G needs. Everything else in that row is a redesign wearing the same function — the CH340N is SOP-8, the CP2102-GM is a QFN-28 with an exposed pad, and using either on an existing SOP-16 land pattern means a new board rather than a new part.

The same applies one row up. The FT232RL is SSOP-28 and carries the historical premium and the driver questions that come with it; the CH340C that replaces it in function is SOP-16, so the honest description is an adapter sub-board or a new layout, not a socket swap.

What we will not claim

Every mapping above was checked against the manufacturer datasheet for package, pin behaviour and supply class, and where the datasheet did not support one it is not listed. Confirm against your own schematic — and on the CAN row, against the pin behaviour rather than the outline.