4
\$\begingroup\$

I have seen some vendors selling 24 VAC transformers to power up swimming pool lights that are submerged in the water. These lights also work with DC but I think transformers are being considered safer by the vendor. Lights shown below for reference:

swimming pool lights

Pool light image link

These are typically electrically insulated but thermally conductive to dissipate heat using the pool water. Many of these even thermally shut down if operated outside water.

I want to understand if an SMPS is a safe alternative to a transformer, for such applications? Something like this:

meanwell PSU

PSU image link

Please note that the transformer or SMPS will not be submerged in the water. These will be installed in a dry ventilated space.

My question is: Is SMPS safe (from a shock hazard point of view) to power up such light fixtures if it works, considering the worst that could happen - light fixture losing insulation and DC wires getting exposed, an electrical fault on the primary side of PSU such as surges, over-voltages, wire disconnects, etc?

There are two reasons that I find concerning. Not sure if it should be a reason to worry or not:

  1. Can the transformers in these power supplies fail causing any input/output short leading to a high voltage, causing shock to a person touching those submerged lights?

  2. Usually these power supplies use a Y capacitor across primary and secondary. Can it fail and lead to a shock if someone touches the submerged lights?

Considering the above points, is a typical AC step-down transformer safer than these power supplies? Or can I expect a SMPS from a reputed brand like Meanwell/Osram to be equally safe?

\$\endgroup\$
17
  • 1
    \$\begingroup\$ Depends on the legislation in the country you’re in but most would require a class 4 transformer. \$\endgroup\$
    – winny
    Commented Nov 6 at 6:59
  • \$\begingroup\$ If you are asking if the supply on the picture can be submerged into water, then the answer is no, absolutely not. \$\endgroup\$
    – Lundin
    Commented Nov 6 at 8:02
  • 5
    \$\begingroup\$ Whatever supply you find acceptable for safety and code requirements, make sure that it is sourced through official channels. There are, for example, dubious copies and look-alikes of Meanwell supplies. \$\endgroup\$
    – Spehro 'speff' Pefhany
    Commented Nov 6 at 15:30
  • 1
    \$\begingroup\$ Another reason for AC supply can be to reduce galvanic corrosion if any moisture gets into contact points. \$\endgroup\$
    – jpa
    Commented Nov 6 at 19:33
  • 1
    \$\begingroup\$ In general, any kind of failure (that I can think of) of such a setup would trip pretty quickly an RCD - the key is that the voltages are low though, with a 1:1 insulation transformer the story is different. \$\endgroup\$
    – Vladimir Cravero
    Commented Nov 7 at 7:46

5 Answers 5

Reset to default
7
\$\begingroup\$

There would have to be an extreme input voltage spike with a lot of energy to cause a short between input and output coils of a ferrite core transformer.
Far less energy is needed for the capacitors to fail, but still the voltage would need to get very high and there are protective elements against that at the inputs of power supplies like yours, although nothing is 100.00% safe.

My practical advice would be to have a separate fuse for the lights supply rated just above the maximum supply input current (better for it to disconnect more often, than for someone to get killed or injured), to install a GFCI to prevent ground currents, AND to have a VERY GOOD and reliable earthing/grounding so that any unwanted voltages/currents are drawn away.
EDIT: I forgot to include probably the most important safety measure - A LIGHTNING ARRESTOR and rod lightning protection system if the building doesn't already have it. Lightning can have and produce voltages so high that no transformer or power adapter is good enough against it.

One thing I don't like about SMPS design is those input to output Y capacitors that always leak some small AC current which can often be felt, and even though it is a safe current, it is sometimes not pleasant to feel.

I am still curious about whether a SMPS failure can lead to electric shock and whether a transformer is undoubtedly safer than SMPS in such use cases.

If you take the precautions I outlined above (I have just added LIGHTNING ARRESTER in addition to lightning rod if the building doesn't already have it), your lights and the pool should be very safe and a power supply failure should cause no harm whatsoever if all the safety measures above are functioning properly.

I would say that the plain, mains frequency iron core transformer is generally safer than the ferrite core one if the primary and secondary windings are horizontally separated with a plastic or other type of insulating separator.

In the end, no measure may be enough against a lightning strike, so the best thing to do is not be in or near the pool during lightning strikes.

\$\endgroup\$
4
\$\begingroup\$

Use a medical-grade DC power supply. Those have better isolation barriers and lower ground leakage current. Everything given in the other answer applies as well.

The supply must be installed in a dry location. They are not rated for wet locations usually! This excludes typical pump shacks, for example.

\$\endgroup\$
3
  • \$\begingroup\$ Noted. This can be done. However I am still curious about whether a SMPS failure can lead to electric shock and whether a transformer is undoubtedly safer than SMPS in such use cases. \$\endgroup\$
    – Whiskeyjack
    Commented Nov 6 at 12:45
  • \$\begingroup\$ A medical transformer has the same safety implications as a medical SMPS. Neither is "safer" than the other. That's why both are medically rated. I would feel uncomfortable installing a non-medical-grade supply for in-pool applications even if most people don't think it necessary. The cost difference is marginal in the grand scheme of things. Labor and everything else dominates the end cost. The primary-to-secondary insulation in medical SMPS'es is usually rather impressive, as it should be. Especially when you disassemble the switching transformer and look at it. So it is in a medical toroid. \$\endgroup\$
    – Kuba hasn't forgotten Monica
    Commented Nov 6 at 12:47
  • \$\begingroup\$ @Whiskeyjack To answer your questions here (will include them in my answer as well): it is very rare but yes, an SMPS failure CAN lead to electric shock, HOWEVER if you follow my advices in my answer, the chances of such shock should be practically zero because the grounding should keep it safe. A plain mains frequency transformer would be safer IF the primary and secondary sides are separated side-by-side instead of over each other. I also forgot to mention the need to keep the power supply as dry as possible. \$\endgroup\$
    – Edin Fifić
    Commented Nov 6 at 15:29
3
\$\begingroup\$

DC promotes corrosion of the conductors in wet places, AC - much less so, if at all.

Use AC unless compelling reasons dictate.

Besides, a simple iron transformer with visibly separated primari and secondary windings has less of a diversity of dangerous failure modes. I am not saying that it is safe by definition or that a reputable ac/dc smps cannot be safe, this is just an informed prejudice.

\$\endgroup\$
4
  • 1
    \$\begingroup\$ I would use AC. A transformer is vastly simpler than a SMPS, and far, far less likely to fail. I can't imagine that a quality SMPS is cheaper than a transformer, and it's unlikely to be more energy-efficient either. (Consider that a transformer is at the heart of virtually all SMPS anyway.) So why use a SMPS? The weight difference shouldn't matter in a stationary application. \$\endgroup\$
    – jimc
    Commented Nov 7 at 17:38
  • \$\begingroup\$ @jimc A good quality SMPS can definitely be cheaper than a transformer. All that iron adds up; the whole point of a switching AC-DC supply is that you can use much smaller magnetics than a mains-frequency transformer, for cost reasons and not just weight reasons. \$\endgroup\$
    – Hearth
    Commented Nov 7 at 20:32
  • 1
    \$\begingroup\$ I agree with the answer, simple iron transformer is harder to mess up. But I also agree with @Hearth, there's no reason that SMPS must be more expensive. The point being -- it's easier to visually inspect a big laminated iron transformer than a black box SMPS. however, jimc is wrong -- transformers are definitely less efficient, if only for their no-load consumption, which is quite likely in a pool light. For instance, when the pool is off and the transformer is humming away. \$\endgroup\$
    – rohmeooo
    Commented Nov 7 at 22:31
  • 1
    \$\begingroup\$ It never occurred to me that the lighting transformer might be switched on the secondary side rather than the primary. Is this common in pool installations? Transformers for LED pool lights would be a lot smaller than for incandescents. Is there really that much iron in question? \$\endgroup\$
    – jimc
    Commented Nov 8 at 21:10
2
\$\begingroup\$

The power supply should be marked SELV (Safety Extra Low Voltage). This says that it contains an isolation transformer between the high voltage input and the (extra) low voltage output.

It may contain a class Y capacitor. Class Y capacitors are designed to fail open circuit if overloaded.

Make sure you buy a reputable brand from a reputable retailer. You don't want a fake power supply.

\$\endgroup\$
0
\$\begingroup\$

Look at the datasheets to determine which supply has the highest isolation. It is in Kilovolts.If you can't find that info about the supply don't use it.

\$\endgroup\$
1
  • 3
    \$\begingroup\$ Anyone can write fake specifications in a datasheet—"Does it come from a reputable company?" is another essential question. \$\endgroup\$
    – Andrew Morton
    Commented Nov 6 at 19:45

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.