Tech minute: Toroidal transformers versus Standard ones

[Vince]

Hi Dave,

Just subscribed to the forum so I can post this, but no doubt I will be here for a long time, because I just love you and your blog ! You are just so special and excellent, every aspect of your personality and the contents of your blog, is just perfect. Such a marvelous blend of many little things, all indispensable, that make the whole thing "just work"  and render it so incredibly addictive, it's a (good) drug ! 

I really like you recent video about the ins and outs of scope probes, and all such videos where you highlight and visit a very particular technical point, so we can get the basics right about it, using your white board. Dave the teacher, would have killed to have teachers like you when I was studying electronics, you rock !  (food for thought when you get bored with the blog in some time and want to do something else).

Anyway, so much as an introduction.

There is one topic I would like to suggest as a potential mini-lecture: toroidal transformers. From the little I can recollect, I gathered they are better than conventional transformers, so one would expect to find them everywhere, but somehow  they are rarely seen outside of high-end audio stuff (both the power supply and output stages/drivers), which I don't understand. Even the super expensive accurate test gear (measuring gear and power supplies) you review and tear down (be it current or 80/90's vintage), somehow use "crappy" standard transformers rather than toroidal ones. So it would be interesting if you could review the main points about toroidal transformers.

1) Advantages (I understand the better magnetic coupling allows for better efficiency: more power in a more compact form factor, as well as less heat/losses, no vibrations/noise)
2) Disadvantages (requires more expensive wire winding machines, in turn making the transformers quite a bit more expensive, for a given power)
3)... and everything you could come up with that would be worth knowing about them, so that people can make an informed decision next time they have to choose a transformer for their project.

Keep up the excellent work Dave ! 

PS: sorry if my English leaves somewhat to be desired, I am from France. I must say though that despite your incredibly high rate of speech, I somehow get you 100% with no effort at all.. your crisp, sharp, crystal clear high-pitched voice, makes wonders !

[EEVblog]

1) Advantages (I understand the better magnetic coupling allows for better efficiency: more power in a more compact form factor, as well as less heat/losses, no vibrations/noise)

Also, they can have lower primary-secondary capacitance, and allow shielding between windings, in particular coupled to the primary. This can reduce common mode noise.
This is common construction in isolation transformers for example.

[SeanB]

Main reason is that they take so much longer to wind, as well as having a massive startup inrush current ( as they are very low loss) that requires extra components to handle it.  The wire for the windings as well needs to be extra flexibe, as it will have to be wound twice, once on the threading bobbin then on the final unit, and this can crack the insulation. You also only have a single form factor ( it can vary slightly, but it will always be round) so it can be hard to fit into a design and still have room, as it is generally going to be a much larger volume than a conventional unit, as the inside of the unit is always open.

On the plus side, very good coupling, no stray fields and you can shield both primary and secondary with electrostatic shields. Also has good regulation, and a low temperature rise under load, though the overload performance is poorer as the core cannot dissipate heat easily.

[Doug_in_Minnesota]

SeanB:

I often take apart computers and other devices to harvest parts. I've built a huge collection of (admittedly used) parts. I've probably got 20 pounds of inductors and transformers.

I see toroids in most computer power supplies. Typically about 1.25 inches in diameter.

From strictly a hobbyist viewpoint (not production) I would think that "custom winding" one of these babies by hand would be both easy and fun.

In the power supplies I've seen, they typically only have a dozen or two windings at most.

But I would love to experiment sometime with winding them in different ways and seeing what I can come up with. Multi tap windings would be really interesting!

[theoldwizard1]

I see toroids in most computer power supplies. Typically about 1.25 inches in diameter.

From strictly a hobbyist viewpoint (not production) I would think that "custom winding" one of these babies by hand would be both easy and fun.

In the power supplies I've seen, they typically only have a dozen or two windings at most.

But I would love to experiment sometime with winding them in different ways and seeing what I can come up with. Multi tap windings would be really interesting!

The big win of toroidal transformers is their efficiency and lower cost, especially when operating at a  "switching" frequencies well above line frequency (50/60 Hz).  Common switching controller ICs operate at a couple of hundred Khz up to about 2 Mhz (LM5032).  Higher frequencies are more efficient, but you have to be careful not to saturate the core.

I have stared at some of the equations for designing toroidal transformers and it has been just to many years since I have been out of university for me to wrap my arms around.  Hopefully, someone who has some design experience in this area will boil it down or maybe provide a spreadsheet as there are many inputs to the formula.

[johansen]

A toroidal transformers first efficiency win is the silicon steel can be grain aligned and run at 1.95T, and run at losses of 1-3 watts per kilogram. yes, you read that correctly.
a transformer's energy density is proportional to flux squared.

the second energy density boost is the primary heat producing element, the copper windings, are almost fully exposed, where as an E core or other core topology hides them inside the core.

however, the first problem with toroidal transformers is very high cost of winding the transformer, and a secondary issue is low fill factor. it is very difficult to get more than 50% copper fill factor.
The second problem is 1.9T grain aligned steel costs 3$ per pound. where as the cheap stuff they punch E cores out of is probably 50 cents a pound.
the cheap stuff also has losses on the order of 10 watts per kilogram
Alphacoredirect sells the blank cores for 3-4$/lb price plus shipping.

large industrial three phase E core transformers also run at 1.5-1.9T but the cores are laser cut from grain aligned steel, not punched out from cheap stuff.
due to the size, they can run at 99.8% efficiency and 1KW/KG power density or better.

[AG6QR]

Toroidal transformers and inductors are commonly used in a lot of ham radio applications, including things like antenna tuners to match impedance between antennas and transceivers.  Some applications include kits, where the labor for winding the toroids is "free", being supplied by the customer who bought the kit.  A nice thing about toroidal inductors is that the magnetic field stays in the toroid, and there is little if any stray field present to be picked up by nearby circuitry.

[T3sl4co1l]

Capacitance: I doubt it's ever lower than an equivalent E-core type.  In fact in some cases it can be vastly higher; bank-wound transformers (which have excellent isolation and creepage specs due to the divided bobbin) obviously have very little capacitance between windings.  Contact area with the core is also lower, so self capacitance is lower (eh, dubious given the compact bank wound construction, but probable).

However, increased capacitance is a necessary and beneficial side-effect of reduced leakage inductance.  Together, it is a sign that the transformer has wider bandwidth (better HF response) and tighter regulation.

That said, you're unlikely to find a toroid that's class II impedance limited (useful for applications of a couple watts -- fusing not required), and the power factor of a cap-input rectifier will be poorer (taller, narrower charging current spikes), which is also a greater ground loop hazard.

As for radio applications, it's worth noting that the external field depends on material and construction.  A perfectly symmetrical winding, with one end returning beneath the winding, or in two layers, doubled back on itself, like so, will always have low strays, regardless of material.  But when low value inductors are required, the low permeability materials used do little to constrain stray fields, and the short, uneven windings give off lots of stray fields.

So, an LF/HF balun on #43 ferrite, lots of turns, even all the way around, probably quite good.  An RFC of 0.2uH or so on #2 powdered iron, beware of strays.

Tim

[Richard Head]

The only advantages I can think that toroidal transformers have over convention EI or EE types are low profile design and very low leakage flux.
Otherwise their disadvantages far outweigh their advantages and that's why they aren't used more often.
The problem with winding them is that you have to wind directly over the core whereas EE transformers can be wound on a bobbin. That is much more convenient and cost effective.
Also, creapage and clearance distances are easier to acheive with a bobbin.
Incidently, you also require a special toroidal winding machine.
Getting back to the first post, you describe conventional transformers as "crappy". I would disagree and say neither are crappy, just different.