2
\$\begingroup\$

I'm putting together a portable equipment rack for field use - the usual sort of thing where a collection of off-the-shelf items are mounted together into a peli-case as a one-off assembly. All the devices are 24v powered, so I'm planning to incorporate a generic switch-mode power brick from RS - something like https://uk.rs-online.com/web/p/ac-dc-adapters/2047301 - with a little row of blade fuses for the various devices.

Under some circumstances it would be useful to power the box directly from a 24v vehicle battery (all devices are good to at least 30v) instead of 230v. I can rely on the operators not to connect both 24v and 230v at the same time (they won't typically have both available anyway), but will applying 24v to the output of the power supply (with no AC on its input) cause any problems?

Obviously I could just include a 24v / 230v switch on the back panel, but if the power supply doesn't mind being back-fed then it would be simpler not to.

New contributor
Pete Verdon is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
\$\endgroup\$
5
  • 4
    \$\begingroup\$ Without knowing the details of the circuit inside your specific power brick, this question is unanswerable. Some can be safely back-fed, some can't. \$\endgroup\$ Commented yesterday
  • 2
    \$\begingroup\$ In addition to what @brhans says, I'll add that "I can rely on the operators" is a recipe for disaster. People forget, they take a phone call halfway through a procedure, someone who isn't qualified thinks they know what they're doing, etc etc etc. Do yourself the favor and use a switch, diodes, relay, whatever, but make it bulletproof. This also relieves the burden on you or someone down the line to find a replacement brick that can be backfed in case it needs to be replaced or upgraded. \$\endgroup\$ Commented yesterday
  • \$\begingroup\$ I'd be surprised if it would cause a problem. The real problem is no one will say whether it is a problem -- unless there's some particular reason to write that in the spec. \$\endgroup\$ Commented yesterday
  • \$\begingroup\$ @TimWilliams I agree that a power supply not being able to withstand its own voltage at the output would be strange, if only for the reason that any load with input capacitors could damage it. However, if "vehicle battery" means "vehicle power" and not just a bare battery, then you have the whole slew of vehicle power defects to contend with, as well as potential interactions between vehicle power and the power supply's regulator. Though - to your point - if the supply is "good to at least 30V" (I see nothing in the data sheet to support this), then it might not be a concern. \$\endgroup\$ Commented yesterday
  • \$\begingroup\$ @vir True, if this is in fact a portable, automotive application -- not just "a vehicle battery" (which could be read as: a battery, used for such purpose, standing alone) but in the vehicle (running on alternator or battery power), that's a very different regime -- the OP may wish to familiarize themselves with automotive power quality and related standards. Not just for sake of what a power brick might be exposed to but the entire rack itself. \$\endgroup\$ Commented yesterday

3 Answers 3

Reset to default
6
\$\begingroup\$

This is kind of what @brhans said, only stated in reverse:

In general, no, you cannot expect an off-the-shelf brick to behave nicely if you back-feed it, either with or without AC power applied. In general you can expect results that range anywhere from exactly the behavior you want to smoke, and possibly even flame.

You can expect your brick to accept back-feeding if the manufacturer specifically states that it can be done in the data sheet. Personally, I wouldn't trust such a statement without doing my own testing before I committed to a final design -- but I'd at least buy samples based on said statement.

If you don't mind a 4% efficiency hit, put sufficient-sized Schottky diodes in line from both the brick and your 24V input -- that way whichever one happens to be supplying more voltage wins the "who gets to feed power" contest.

If you do mind the efficiency hit and you don't mind the trouble and expense, you can get "0V diode" chips that either have an internal FET or manage an external one, and that end up acting like a diode with 50mV or so voltage drop. It's more design complexity and a higher BOM cost -- but you don't have one diode burning up 4% of your input power.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks Tim. The datasheet doesn't say, so I'll either take one of your suggestions or simply add the physical switch. \$\endgroup\$ Commented yesterday
  • \$\begingroup\$ I would be cautious with the Schottky diode suggestion too. They leak in reverse much more current than e.g. standard diodes. It might not be a problem in this application but in general it may come as a surprise because people only look at the forward voltage being lower. \$\endgroup\$ Commented 17 hours ago
2
\$\begingroup\$

Since you state it's a "switch-mode power brick," almost certainly it has voltage regulation. Connecting it directly in parallel to another supply is likely to cause issues.

There are a few ways around that issue:

  1. Use a power connector with internal switch, such as the barrel connector shown below, from Pololu.

    Barrel connector with switch

    When the male connector is inserted, the switch should disconnect the internal supply.

  2. If the voltage requirements are not precise, two ordinary silicon diodes can be inserted in the connection to the internal and external supplies. That allows power to flow from each supply to the device, but not between the two supplies. Ordinary power silicon diodes have a drop of ~0.7V, so the supply would be ~23.3VDC, rather than 24.0VDC. Schottky diodes have slightly less forward voltage drop, between 0.2V to 0.5V but are more expensive.

  3. Use an automatic power changeover switch. It can be a simple DPDT relay, which firsts disconnects the internal power before applying 24VDC from the outside, or a commercial power transfer switch, such as this example from Amazon.

N.B.: Devices listed are generic examples. You will need to select one that suits the device requirements.

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

Provide only one DC power input connector and have the operator choose whether to connect the battery cable or the DC supply there.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ This doesn't answer the asked question if backfeeding the supply is an issue or not. But to be fair, this is a solution that removes the need to ask it. \$\endgroup\$ Commented 17 hours ago
  • \$\begingroup\$ I did consider this, but I’m building the power brick into the rack. If it was loose, it’s guaranteed to get lost or swapped for a different one with the wrong voltage or power. So there would have to be a short lead coming out of the rack and back in - doable, but a bit messy. I’ve added a switch instead: 230v-off-24v \$\endgroup\$ Commented 7 hours ago

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.