TIDA-00527: USING AS A POWER OVER BUS

Hi Ahmet,

Thanks for posting on e2e.  

Assuming that the series capacitances used are chosen to be very low impedance across the frequency range of the serial data and the inductors chosen are very high impedance, the total distance and data rates achievable with this set-up would be similar to normal RS-485 signaling.  Transmission at 1 Mbps across 200-m of CAT-5 would be achievable.

Regarding power transfer - the total current delivered would be limited by the rating of the coupling inductances, the total voltage would be limited by the rating of the coupling capacitances, and you are correct that you would need to consider series losses due to the DC resistance of the cabling.    For CAT5 cabling at these distances I'd expect series resistance < 10 Ohms per line, so if you were transferring 250 mA then you would lose up to 2.5 V on the high and low side of the power lines.  If that were an issue then you could switch over to lower-gauge cabling, but an easier solution might be to transmit at a higher voltage and then regulate the voltage down to the desired level at each receiving node.

Note that since AC-coupling capacitances are used, one thing to be careful of is making sure worst-case strings of "high" or "low"  bits on the serial interface do not prevent the data from making it properly across the capacitors.  At high data rates (1+ Mbps) this is often manageable without introducing encoding by keeping frame lengths short and by making sure "idle" periods between frame transfers are properly biased.

There was some discussion about similar concerns in this thread - I think it may be an good reference to you:

https://e2e.ti.com/support/tools/sim-hw-system-design/f/234/t/869271?TIDA-00527-Distance-vs-power-supply-curent-

Please just let us know what further questions you may have.

Regards,
Max

Hi Max,

Thanks for your quick reply. I have a question.  

Since the BUS signal is sent over 24VDC, In case it is not isolated, THVD2450 (± 70-V fault-protected 3.3-V to 5-V RS-485 transceiver with IEC ESD) can be used. But can you suggest an isolated transceiver that supports 24VDC bus?

Or; power over BUS voltage is passed through a bridge diode. In this case, can we use the discrete dc-dc isolated switcher and supply the voltage of the transciever? Or what is your suggestion?

Secondly, What is your comment about the isolated solution in figure-10(Texas Ins.) in the document in the link below?

http://www.ti.com/lit/an/slla272c/slla272c.pdf

How about a solution like in the attached picture?

https://www.maximintegrated.com/en/products/interface/transceivers/MAX3158.html

By carrying the Ground with the shield, will there be an EMI problem in the cable in the isolated solution?

Best regards....

Ahmet Turan ALGIN

Hello,

I think using an isolated DC/DC makes sense if your system requires isolation on the power rail.  However, I don't believe isolation is needed on the communication lines (e.g., through an isolated transceiver, etc.).  This is because the coupling capacitor itself provides galvanic isolation.  By the same logic, a high-voltage transceiver may not be required either since the DC bus voltage would not make it to the transceiver inputs.  (One exception here would be if the DC supply were "hot plugged" onto the bus.  In that case the step transient would make it across to the transceiver before slowly decaying down to the normal operating common mode bias of the transceiver.)

Regards,
Max

Hi Max,

Thanks for your reply. If we draw a general block diagram; I think the figure below appears.

• Is there a mistake in the general structure I have installed?
• For the 24V DC component, what range should the inductor value be?

The power line voltage is blocked from the transceiver bus pins given that the pins are AC- coupled to the power line. However, we should still consider the possibility of direct exposure of the transceiver pins to the DC voltage through a short or AC-coupling capacitor failure.

• Do you have any suggestions for this situation?

RS-485 standard conform drivers provide a differential output of a minimum 1.5 V across a 54Ω load, whereas standard conform receivers detect a differential input down to 200 mV. The two values provide sufficient margin for a reliable data transmission even under severe signal degradation across the cable and connectors.This robustness is the main reason why RS-485 is well suited for long- distance networking in noisy environment. In this case,

• Is there any difference between using the transceiver's VCC voltage, 3.3VDC or 5VDC?

Best regards......

Ahmet Turan ALGIN

Hi Ahmet,

It looks to me like your understanding is correct.

The voltage drop across the inductor would be fairly small, so it isn't so much the bus voltage that would determine the inductor rating.  The more important factors are the peak/average load currents through it (which should not exceed the component rating), the DC resistance (since this will create a corresponding voltage drop under load that should not be too high for operation of the system), and the inductance (which should be suitably high as not to provide significant loading of the RS-485 data when considering the minimum frequency of the serial data).

You are right about the effects of the AC coupling capacitor, and if you did want to ensure no further damage to the transceiver in the event of a capacitor failure (or step transient such as a hot plug) then use of a high-voltage-capable transceiver (e.g., THVD2450) as you originally proposed is a good solution.

Regarding your last point about loss margin - in general transceivers are designed to provide 1.5 V worst-case and so there may not be significant differences in output amplitude between 3.3-V-powered and 5-V-powered transceivers.  There are some exceptions, though, notably in devices like THVD2450 that are designed to support both common VCC ranges.  Since that device needs to hit the 1.5 V minimum with just a 3-V supply, when operating at 5 V instead the "headroom" increases and the output amplitude increases accordingly.  So, theoretically using one of these devices at 5-V could provide further reach.

In the case of a power-line-communication application, though, since the signaling rate needs to be high you will probably find that jitter due to signal distortion (and its impact on apparent edge timing) will become a limiting factor before DC signal loss due to cable DCR (which would tend to be the limiting factor for very long cables operating at lower signaling rates).  The higher amplitude will still help with noise immunity, though.

Regards,
Max

Hi Max,

Thank you for your comments.. I hope I didn't overwhelm you with my questions! I have one last question. On the bus, does the use of the dc bias voltage at the ends of the nodes benefit us?

Best regards.....

Ahmet Turan ALGIN