1
\$\begingroup\$

I have seen plenty of questions asked and answered about setting up whole-house low-voltage DC via something like PoE (passive over standard Cat5e cable, run through plasterboard walls).

With the cost of electricity going up more and more, I have been looking at this again.

Using an off-the-shelf switch-mode power supply that outputs 48 V (with an efficiency of 85% @ 240 V) and running over Cat5e cable (average length 30' - 50'), at what point does it become more economical to switch to PoE (passive) over multiple plugin 240 V -> 5 V power supplies (or mains sockets with USB) at each piece of equipment?

Main usage for the power is charging devices (phones/tablets), running a few Raspberry Pi's (5 V, 4 A) and some other basic IoT devices.

My thinking is: does using a single PSU save more money than running ten 5 V wall wart PSUs purely in running costs?

Cost of parts is not a big factor (other than the main 48 V PSU) as I have plenty of small switch-mode modules that I would repurpose for my end equipment.

Any suggested calculations/formulas would be great.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ I think this is a bit off-topic to this site. Perhaps the Home Improvement site would be a better fit? \$\endgroup\$
    – jaskij
    Aug 28, 2022 at 22:09
  • 3
    \$\begingroup\$ Resistive losses are probably going to balance out any efficiency gain from a central power supply. Ethernet is a poor choice for a power cable. \$\endgroup\$ Aug 28, 2022 at 22:40
  • 2
    \$\begingroup\$ Unlikely to be cost effective to run in a whole second set of cabling, sockets etc for LV. One or more wall warts in each room is likely to be cheaper. \$\endgroup\$
    – user16324
    Aug 28, 2022 at 23:04
  • \$\begingroup\$ The relevant question is whether your 1 big wall-wart is actually more efficient than many small ones. Possibly yes because of quiescent current, but also possibly no \$\endgroup\$
    – user253751
    Aug 30, 2022 at 1:04
  • 1
    \$\begingroup\$ "Main usage for the power is charging devices (phones/tablets), running a few Raspberry Pi's (5 V, 4 A) and some other basic IoT devices." You realize that running your oven/microwave for 10 seconds probably consumes more current than running all of these devices at once all day long? \$\endgroup\$
    – Lundin
    Sep 1, 2022 at 14:27

2 Answers 2

2
\$\begingroup\$

The game I like to play when trying to make my home more efficient is the one of worst cases:

You have ten of 5 V, 4 A supplies. Let's assume they run at max load all the time, 80 % efficiency. These supplies are providing 200 watts total of electricity and wasting 50 watts of heat doing it - that's

\$0.05\times 24=1.2\text{ kWh}\$ per day of energy, or \$1.2\times 31=37\text{ kWh}\$ per month. If that costs you 20 cents (pence?) per kWh then your monthly bill is $7.44 (insert local currency) converted directly to heat. The solution you propose thus must be better than 80 % end to end (i.e. your 85 % converter plus your POE point of load conversion should do better than my guess of 80 %) but also must be economical at a payback rate of a maximum of less than 8 bucks a month.

Note that I posit (and my worst case relies upon) the amount of power being wasted increases with increasing load, even if the efficiency gets numerically better.

\$\endgroup\$
4
  • \$\begingroup\$ Current energy futures price in Europe is closer to $1.40 per kWh, thus (if these futures prices are realized next year) the calculation gives >$50 per month. \$\endgroup\$
    – user253751
    Aug 29, 2022 at 23:26
  • \$\begingroup\$ also you did not consider that some people may do some things out of interest even if they are not monetarily worth it. \$\endgroup\$
    – user253751
    Aug 29, 2022 at 23:28
  • \$\begingroup\$ @user253751 That's some pricey electricity. Also of course I consider that, I do things that "arent worth it" all the time. The question cites the cost of energy as a motivator: "With the cost of electricity going up more and more, I have been looking at this again." and then calls for analysis techniques: "Any suggested calculations/formulas would be great.". I answered the question with those statements as my guide, and I see how the tone could be interpreted as "this is not worth it don't do it" but that was not my intent. \$\endgroup\$
    – Bryan
    Aug 30, 2022 at 1:00
  • \$\begingroup\$ They could be inspired by the price of electricity, and then want to do it for fun even if it only saves a few dollars. \$\endgroup\$
    – user253751
    Aug 30, 2022 at 1:04
0
\$\begingroup\$

You need to consider the power lost in the cable runs. The PoE supply may be delivering 48V at one end, but unless you use really fat cabling, you might not get anything like 48V at the far end once you connect a load. Any volt drop represents power wasted in the cable run.

Devices that are designed to accept PoE will have a built-in voltage regulator. But most household gadgets aren't designed to run on something that's notionally 48V, but may be less. So you're going to have to add a DC-DC converter at the load end to charge a phone. That's another inefficiency to factor in.

\$\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.