# Electrical engineers: small transformer efficiency?

I have an LED lamp powered by a commercially-made Switched-mode power supply. Putting out 170 watts at 24 Vdc, the manufacturer claims the power supply has an efficiency of 90%.

For this power/voltage output, how does this compare with a conventional iron transformer/rectifier/capacitor setup? By “conventional,” I mean a transformer whose primary receives wall voltage (110 Vac), and whose output passes through a rectifier and capacitor to produce 24 Vdc.

What would the efficiency be for the latter? How about physical size? The SMPS I’m using is 8.5" x 2.5" x 1.5". Would the transformer/rectifier/cap setup be far larger?

The dirt simple power supply like that actually isn’t too bad with respect to efficiency. They are usually up around 90% or so. The problem is that there is no voltage regulation. AC power can vary. It used to be that the power company guaranteed +/- 10%. Most of them these days guarantee +/- 5%. Your power company’s web site probably has what they guarantee somewhere on it, if you are curious.

Since it is unregulated, the AC will vary between say 109 and 121 volts (assuming they are aiming for 115 volts, +/- 5%), which means your 24 volt output will likewise vary from about 22.8 to 25.2 volts. This could make a visible difference in the brightness of your lamp.

The power supply would also be physically larger and heavier, due mostly to the transformer. For smaller power supplies, it would also be cheaper, but for something in the range of 150 to 200 watts or so the transformer is going to be fairly large, which means a lot of expensive metal.

If you add a simple linear voltage regulator, those don’t do so well from an efficiency point of view. Your overall efficiency is likely to drop to somewhere around 50% or so,with the linear regulator producing a lot of waste heat as a result. But then your 24 volts would always be close to 24 volts. Since a linear regulator basically bolts onto the back end of your simple power supply, it adds to both size and complexity, increasing the cost accordingly.

A typical switch mode power supply eliminates the huge bulky transformer and uses much smaller transformers (operating at a much higher frequency, which is how they can be so much smaller and lighter). The circuitry is much more complex, so for smaller power supplies the cost of all of the extra circuitry outweighs the cost of the simple transformer. This is why your typical wall warts are basically the simple power supply with a linear regulator (some of them even just have a rectifier and don’t have the regulator). Once you get above a certain power level, then the extra cost of the switch mode circuitry begins to outweigh the cost of the huge metal transformer, and you also get a huge benefit with respect to waste heat, which also becomes much more of an issue as the wattage goes up.

To e_c_g’s correct information I will add that nearly all switching power supplies have a transformer in them to provide mains isolation and to keep the switching duty cycle within reason. These transformers operate at 10’s to 100’s of KHz so are much smaller and lighter and cheaper than mains frequency transformers but are rarely more efficient and often less so.

I don’t know anything about the SMPSs used for LED lamps, so this may be a dumb question. But are we sure the SMPS does not have output voltage regulation?

My understanding is that commercially-produced SMPSs incorporate a feedback mechanism that governs the switching duty cycle so that output voltage is maintained despite large variations in input voltage or output current. Case in point, the SMPS I’m using (pdf file) can handle input voltages ranging from 90-264 Vac or 127-373 Vdc.

I was talking about his simple transformer/rectifier/capacitor power supply, not the SMPS. This simple power supply is unregulated.

A SMPS almost by definition has to be regulated. That’s how it works.

Got it. Thought you were talking about the SMPS.

One bigger-piture aspect to keep in mind for ‘efficiency’: the 10% waste in that power supply becomes heat, that is released into the surroundings. Depending on your location & the weather, that may not be entirely wasted. For examole, here in a Minnesota winter, that extra heat just goes into the house, and my furnace kicks in very slightly less often. So some portion of that 10% waste isn’t really waste (though still less efficient heating than the furnace). The only time it’s really wasteful is during the summer, which is a short part of the year here in Minnesota.

Sometimes the ‘waste’ heat from appliances is wanted. I know when I replaced my computer CRT monitor with a much smaller & cooler LED monitor, my cats were much dissatisfied.

While I certainly agree, “they” never take that into account.

At any rate, and generally speaking, a SMPS is more efficient than a traditional linear power supply (all else being equal). Of course, that doesn’t mean they’re “better” – it simply means they’re more efficient.

As any designer knows, there’s more to life than efficiency. If you’re designing an analog circuit that requires ultra-low noise & ripple on the rails, for example, you’ll probably want to use a linear power supply. Or a battery.