0
\$\begingroup\$

I have found multiple RS232 to RS485 converter dongles (from different brands, models, and makes) that do not need external power to operate.

Connect the RS232 DB9 on one side, and three wires for the RS485 out the other side, there is even a small LED in the converter that shows connectivity.

I assume (guess) the power comes from the RS485 side of things as the LED goes dark when I only have the RS232 side connected.

  • Where is the power coming from to power this converter? (I assume RS485 A/B)
  • How much current is available?
  • Is there enough current to run a microcontroller (like SAMD21)?
  • Is there a reference design for capturing power from RS485 or RS232?

Examples:

\$\endgroup\$
9
  • \$\begingroup\$ you are asking technical questions about an unknown device \$\endgroup\$
    – jsotola
    Commented Jul 11 at 21:10
  • \$\begingroup\$ @jsotola this seems to be common across multiple vendors, brands and makes of RS232 to RS485 converters. \$\endgroup\$
    – Steven Smethurst
    Commented Jul 11 at 21:12
  • \$\begingroup\$ @StevenSmethurst or multiple brands selling the same proprietary design. \$\endgroup\$
    – Justme
    Commented Jul 11 at 21:13
  • 2
    \$\begingroup\$ I quote directly from the serialcomm.com page: "The unit is enclosed in a rugged ABS housing, and is powered from the RS232 data lines; no external power is required." \$\endgroup\$
    – MrGerber
    Commented Jul 11 at 21:26
  • 1
    \$\begingroup\$ I've seen and used stuff like this too and in every case it seemed like bad engineering. They are assuming things about the serial lines when they should assume nothing. Also it puts your serial line in a ground loop so there's current flowing in the signal ground, which is an open invitation to all manner of intricate EMC problems. \$\endgroup\$
    – Lundin
    Commented Jul 12 at 6:57

3 Answers 3

Reset to default
5
\$\begingroup\$

Where is the power coming from to power this converter? (I assume RS485 A/B)

your serialcomm.com link says on the specification page, under "electrical", that it's powered from the RS232, not RS485. antaira.com.tw is the same.

How much current is available?

Impossible to know, because that must depend on the device driving the RS232 line with power. However, again, reading docs is key, since there we learn that the device uses less than 10 mA power. So, the designers felt it safe to assume that it works with 10 mA. antaira.com.tw says 30 mA at 5V. A bit of googling around suggests that Analog Devices thinks that you can only assume 10 mA of current capacity from PC equipment.

Is there enough current to run a microcontroller (like SAMD21)?

Is 10 mA enough? Probably, yes. But whether it suffices for your application depends on what you do with that microcontroller.

Is there a reference design for capturing power from RS485 or RS232?

I'm not aware of one, but somewhere, someone might have done one. But after all, it's just RS-232. If in doubt, you'd just attach a bridge rectifier to the data lines, and add current-limiting to that before you add bulk capacitance.

\$\endgroup\$
1
2
\$\begingroup\$

RS485 has impedance about 100 ohms and signal voltage about 3 to 5V

so while it's active you have about 250mW best case 100mW worst case

probably enough to run a small microcontroller if you replace the termination with something that powers your microcontroller.

The problem is that rs485 is a half duplex bus and there will be no power available while the bus is idle.

finding enough energy to transmit a reply over the RS485 bus could be tricky.

Rs232 has a few spare wires than can be used to power devices, for a reference design there's probably an old PC mouse circuit schematic out there somewhere.

also this RS232 powered relay:

schematic

simulate this circuit – Schematic created using CircuitLab

here the C1 accumulates charge which is dumped through the pulse (latching) relay to switch its state when the state of the RTS pin changes

\$\endgroup\$
1
\$\begingroup\$

Edit: one of the devices you linked to says it is powered from RS-232 side. Which makes more sense as ther is no power available from a half duplex RS-485 bus anyway.

As mentioned, most of your questions cannot be answered without reverse engineering the device. We cannot know how a random device works, why it works the way it does, or should it even work at all, and what was someone thinking when it was designed. We can only guess.

  1. From where the power really comes from is unknown. I would guess from RS-232 side because it has at least some handshake lines availabe for using parasite power, while a half duplex RS-485 should literally have none available.

  2. From RS-232, maybe 20mA, but it depends. Ancient RS-232 mice were standard and worked fine.

  3. Ancient RS-232 mice had microcontrollers. So yes. Many hobby projects also use RS-232 for power.

  4. Define a reference design first. Many chip manufacturers have published designs for something that just happens to be powered from RS-232 as if is a perfectly normal thing. From RS-485 I firmly believe nobody can suck any power out of a half duplex bus.

Powering things from RS-232 seems to be sort of a lost are now that we have this USB that gives power too. Most hobbyist projects simply used a 78L05 regulator with a diode or two, or a 5.1V Zener diode.

Only if a RS-485 trasmitter is always on, it is possible to suck power from it.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.