Testing the Isolated RS485 Building Block Board

Testing the Isolated RS485 Building Block Board#

Over the past few days, I finally got around to testing the Isolated RS485 Building Block Board.

As a quick summary, the board uses a TI ISO1430 isolated RS485 transceiver together with an onboard 3.3 V-to-5 V isolated DC-DC converter. It also includes reverse-polarity and overvoltage protection on the 3.3 V input.

Initial testing has gone reasonably well, but it has also identified a problem with the input protection circuit that will make a Rev B of the board necessary.

Initial Power Testing#

I started with basic continuity checks to confirm that the different sections of the board were connected as expected. Once those checks passed, I powered up the board.

The 3.3 V indicator LED lit as expected, and I measured approximately 3.21 V on the input-side supply rail. On the isolated side, the output from the DC-DC converter measured approximately 5.21 V.

The converter used on this revision is a YLPTEC B0305S-1WR2, rated for a 3.3 V input, 5 V output and up to 200 mA of output current. It appears to be intended as an alternative to the similarly specified Mornsun B0305S-1WR2, which was unavailable when I ordered the components.

Because this is an unregulated converter, some variation in its output voltage is expected depending on its input voltage and load.

Reverse-Polarity Protection#

The reverse-polarity protection uses an AO3401 P-channel MOSFET, with the drain connected to the power input, the source connected to the protected supply rail and the gate connected to ground.

To test it, I deliberately reversed the input connections by connecting 3.3 V to ground and ground to the 3.3 V input.

The MOSFET protection worked as intended and prevented the board from powering up.

Below is a picture of the Rev A board:

Finding a Problem with the Overvoltage Protection#

The next test was the input overvoltage protection, which consists of a 3.6 V Zener diode and a Littelfuse 1206L050/15YR PPTC resettable fuse.

I applied 5 V to the nominally 3.3 V input and the board survived, which was good—but it appears to have survived despite the protection circuit rather than because the circuit operated as intended.

The Zener diode clamped the input rail to approximately 4 V, while the input current reached around 500 mA. I had originally selected the PPTC on the assumption that it would trip at 500 mA. However, 500 mA is actually its hold-current rating; its specified trip current is 1 A.

This explains why the PPTC continued passing approximately 500 mA instead of disconnecting the supply.

With approximately 4 V reaching the input of the unregulated DC-DC converter, its output rose to around 6.4 V. This was then applied to VCC2 on the bus side of the ISO1430.

The recommended maximum operating voltage for VCC2 is 5.5 V, while the absolute maximum is 6 V. Feeding it approximately 6.4 V therefore exceeded the device’s absolute maximum rating.

The ISO1430 survived the brief test, but that should be considered luck rather than evidence of an effective protection design. The overvoltage protection circuit will need to be redesigned for Rev B.

For Rev B, I will probably redesign the power stage around a regulated isolated DC-DC converter. This would provide a more predictable 5 V supply for the isolated side and avoid relying on the load regulation of the current converter, although the final decision will depend partly on BOM cost.

Another possibility would be to retain the unregulated converter and add properly sized secondary-side voltage limiting. However, simply adding a nominal 5.5 V Zener would leave very little margin below the ISO1430’s 6 V absolute maximum and would not account for the Zener’s tolerance or dynamic clamping voltage.

The input protection also needs to be reconsidered. A lower-current PPTC may help limit sustained fault current, but its hold current, trip current, operating temperature and trip time will need to be selected against the board’s normal current consumption. It should not be treated as a fast or precise overvoltage protection device.

Measured at the 3.3 V input, the board currently draws approximately 100 mA while idle and 140 mA with the driver enabled and a 120 Ω termination resistor connected across the A and B lines.

I will repeat these measurements when the board is connected to the complete test harness. That should give me a better understanding of its normal and worst-case current draw before I select a more appropriate PPTC, while still allowing sufficient margin for start-up current and temperature-related derating.

Testing the RS485 Driver#

I then tested the driver output, partly to confirm that the board was still operational after the overvoltage test.

I set DE high to enable the driver and /RE high to disable the receiver, then switched the DI input between 3.3 V and 0 V.

With DI high, I measured:

With DI low, I measured:

These results were in line with what I expected from the RS485 driver, confirming that the device was still operating and correctly switching the differential bus outputs.

That seemed like a good point to finish the day on a high note.

Next Steps#

The next step will be to connect the board to approximately 20 metres of CAT5 cable and reuse my existing test harness, consisting of Raspberry Pi Pico boards, an OLED display and an EC11 rotary encoder.

This will allow me to test actual data transmission over a longer cable and confirm that the isolated board works correctly as part of a complete RS485 system.

After that, I will use the results from these tests to finalise the power and protection changes for Rev B.