8

I want to interface to this device where (sec 4.2) states

can be powered from the DTR and RTS handshaking lines by having one high and one low

Also, Wikipedia states:

No method is specified for sending power to a device. While a small amount of current can be extracted from the DTR and RTS lines, this is only suitable for low-power devices

  1. I would like to understand how RS-232 allows DTR / RTS to be used as power.
  2. What polarity / voltage levels should be observed?
  3. Can they be permanently wired or do they need to observe a protocol?

I'm particularly concerned that if I just wire DTR / RTS permanently to power then the handshaking protocol will screw up (assuming the device I'm interfacing to observes the handshake protocol).

Peter Mortensen
  • 1,679
  • 3
  • 17
  • 23
akellyirl
  • 4,177
  • 1
  • 17
  • 31
  • 2
    There's no hard line of distinction between a "signal" and a "supply" - merely one of degree and intended usage. So if you have something that consumes less power than the hardest-to-drive input your RS232 port is supposed to be able to drive (a spec you can look up), then you can with care use that as a power source. The state of these lines is usually controllable from software, but depending on operating system & configuration may change when the serial port API handle is open/closed. – Chris Stratton Oct 28 '14 at 14:41

2 Answers2

6

I am currently using the DTR lead for two designs. Like all RS232 leads, the voltage varies from a minus voltage (unasserted) to a positive voltage (asserted). The voltage range is anywhere from ±3v to ±12v. Frankly, I have never seen a voltage amplitude lower than ±5, and that was from a PC COMM port.

The spec for your device says to power the RS-232 interface using +12v on the DTR (asserted), and -12v on the RTS lead (unasserted), for a total of 24 volts. It says it needs 20 ma. This may or may not be possible depending on the RS-232 interface at the other end.

When I did this, there appeared to be a resistor in series with the line at the other end, which acted as a current limiter. I "harvested" between 5 and 10 mA from the DTR line. Even that caused the voltage to drop a bit.

I suggest testing this out first, by measuring the voltage on the DTR and RTS leads with a multimeter. If they're not -12 volts, you can forgot supplying power this way.

If it is -12v, then put a 1210 Ω resistor across the DTR and RTS leads, assert the DTR line and measure the correspond leads to see if the voltage holds up (+12v). You can assert the DTR line either in software (attribute of the COMM function in the Windows API), or for this test, use a program like RealTerm, which has the ability to control the various handshake leads under the Pins tab in its user interface.

I wouldn't worry about the handshaking; if they are telling you to power their interface this way, then it must not be an issue.

tcrosley
  • 48,066
  • 5
  • 98
  • 162
5

RS-232 does not provide for any power standard. Using the RTS and DTS pins as power or GPIO is an abuse of the standard, taking advantage of the nature of electronics in the PC/RS-232 driver. The driver's abililiy to source or sink current determines how much power you can provide to the device. Polarity is determined by the pin state (high or low), while the voltage level is determined by the circuit of the RS-232 driver used.

Any handshaking, if used, can cause reverse voltage or brownout problems, but it will be very short and can maybe be negligible.

Just like using a potato for a battery, just because it works doesn't mean it works well or intended to work that way.

Peter Mortensen
  • 1,679
  • 3
  • 17
  • 23
Passerby
  • 73,027
  • 7
  • 93
  • 209