1
\$\begingroup\$

I primarily do VHDL for work, so please forgive if this question is dumb.

I started teaching myself PCB layout and am using a buck converter as the application. I want my buck converter to handle 50 W. Output voltage is 2.5 V. Therefore output current is 20 A. Load is capacitive.

When I look for switching transformers on DigiKey, they do not seem to list a rated current, saturation current, etc. as a search parameter. That makes it really hard to find a suitable transformer.

On the other hand, the AC transformers have current ratings, which makes it easy to find a transformer that will not saturate, light on fire, etc.

Pros of Using a Power Transformer:

  • Easy to find a transformer that will not saturate, light on fire, etc.
  • Step down widely available, so it's easy to make the output current mode which makes it easier to not short out converter when driving a capacitive load.

Cons of Using a Power Transformer:

  • Core tends to be much larger than we prefer for a switching converter
  • Ideal operating frequency is much lower than we usually like for switching converters. This problem should be solvable by simply operating the transformer close to its preferred frequency. For instance, if transformer is designed for 60 Hz, and switcher is fixed on-time, variable frequency, I can operate from 1-120 Hz.
  • Harmonics of square wave will be attenuated (not sure if this is a good or a bad thing)
  • Bulky
  • Heavy
  • Expensive

Here is my question:

Are there any problems I haven't already thought of? If so, how easy or difficult are they to solve?

\$\endgroup\$
6
  • \$\begingroup\$ Is your input AC mains or another, lower DC voltage? \$\endgroup\$
    – Jens
    Commented Sep 14, 2023 at 12:58
  • \$\begingroup\$ 10-20 VAC, 10-1000 Hz. \$\endgroup\$ Commented Sep 14, 2023 at 13:25
  • \$\begingroup\$ Without airgap, mains-frequency power transformers store very little energy. Within that range, one could be used. But it's not much energy, and at such low frequencies, even less power. Surely there must be a better way? \$\endgroup\$ Commented Sep 14, 2023 at 13:42
  • 3
    \$\begingroup\$ For this use case you don't need a transformer, just an inductor. Low frequency transformers work acceptable up to 400 Hz in a forward converter topology, not in a buck converter. \$\endgroup\$
    – Jens
    Commented Sep 14, 2023 at 13:44
  • 1
    \$\begingroup\$ Technically, all transformers are "AC transformers". Transformers don't work with DC. \$\endgroup\$
    – brhans
    Commented Sep 14, 2023 at 17:13

2 Answers 2

2
\$\begingroup\$

How Bad of An Idea Is It to Use an AC Transformer for a DC-DC Converter?

It is a bad idea.

Are there any problems I haven't already thought of? If so, how easy or difficult are they to solve?

There are a number of DC-DC converter topologies that use coupled inductors or "transformers".

Flyback Topology

The main difference between the coupled inductors of a flyback converter and a classical transformer is this. The coupled inductors in a flyback converter are designed to store energy during part of a cycle, and release that energy during another part of a cycle. Classical mains frequency power transformers, on the other hand, are designed transfer power immediately from the primary to the secondary.

You can think of the coupled inductor in a flyback converter as an inductor, where energy is temporarily stored, but an inductor which just happens to have two (or more) windings. One for storing the energy, and another (or more) where the energy is later discharged. That is why many prefer to call these devices coupled inductors rather than "transformers". However, the term fly-back transformer is also commonly used.

Because the coupled inductor in a flyback converter is designed to store energy, it often has either an physical "air" gap in the core (which is actually often filled with a solid material to reduce stresses and vibrations) or a "distributed air gap". The energy storage capacity in a magnetic core depends on the volume of the core divided by the permeability of the core plus the volume of the air gap divided by the permeability of the "air". Since the relative permeability of the magnetic core is high, it is often the case that the bulk of the stored energy capacity resides in the air gap. For this reason, a traditional power transformer would make a very poor fly-back transformer, even apart from the issue of frequency.

how easy or difficult are they to solve?

I suppose one could increase the energy storage capacity of a traditional transformer by increasing the bulk of iron. However, that seems rather foolish when the same increase in energy storage capacity can be achieved by adding a small amount of gap. Also, increasing the core size will also increase core losses. So, one will end up using a very bulky, lossy core.

\$\endgroup\$
5
  • \$\begingroup\$ I'm not sure how representative the flyback converter is. Most converters that use transformers don't require the transformer to store energy like the flyback does. \$\endgroup\$
    – Hearth
    Commented Sep 14, 2023 at 14:11
  • \$\begingroup\$ @Hearth, it is true that topologies like forward converter use a transformer as a transformer, and not as an inductor. But then they require an extra inductor which stores energy when switching element is off. Further, the forward converter requires an extra winding to remove DC bias. \$\endgroup\$ Commented Sep 14, 2023 at 14:18
  • \$\begingroup\$ Not all forward converters use an additional winding. An active-clamp forward doesn't need one, for instance. But yes, they do have an additional inductor--you need some kind of energy storage device. My point is just that the flyback converter isn't really representative when put beside the forward, push-pull, full-bridge, isolated Cùk, and probably other transformer-coupled topologies I'm not aware of. \$\endgroup\$
    – Hearth
    Commented Sep 14, 2023 at 14:24
  • \$\begingroup\$ Hmm. Seems I need to do a case by case reasoning for why traditional mains frequency power transformers are not suitable for this, that or the other converter topology. \$\endgroup\$ Commented Sep 14, 2023 at 14:35
  • 2
    \$\begingroup\$ I think the argument for all of them boils down to the same: the eddy current and hysteretic losses are going to be much too high at switching converter frequencies. No one uses iron laminate cores for switching converters; you use metal powder or ferrite. \$\endgroup\$
    – Hearth
    Commented Sep 14, 2023 at 14:41
0
\$\begingroup\$

So your real problem is that you don't have a good way to search for transformers by current rating on Digikey.com.

The simple answer is to search somewhere else. Manufacturers often have their own search tools on their website. There are of course also other distributors that do let you search by current rating.

  1. Try another distributor. Mouser.com sells transformers and you can filter by "power" (in your case >50W), and current rating (in your case > 20A).

  2. Try searching manufacturers directly. Wurth Electronics has a wide selection of transformers and they have good search tools on their site. The Red Expert tool is free and is set up to find transformers and inductors for power supply designs.

  3. Go on Digikey.com and search for SMPS transformers. Do what filtering you can (like active, in stock, max voltage, etc). Make a note of which manufacturers are left after the search. Then go to the websites for those manufacturers and search there directly. Some manufacturers actually have good product catalogs or search tools to help you choose the right part. Since we found the list of manufacturers on Digikey, its likely that parts you find on their website will be available on Digikey.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.