Designers use Litz copper wire for high frequency windings. That rapidly increases the diameter of wire needed and decreases the stacking factor.
Should’t it be possible to make a copper wire with distributed gaps made from insulator over all wire length?
Edit:
Cut through wire:
Shouldn’t it be possible to make a copper wire with distributed gaps made from insulator over all wire length?
Your proposal will not prevent current hogging the low-inductance path around the periphery of the conductor. That's called skin effect. My addition to your image (in red) indicates current density at high frequencies: -
The redder the colour the higher the current density. Note that current density in the middle is zero and, this is wasted copper.
The main reason to use Litz wire is this; by design it ensures that a wire "strand" at the periphery will move into the centre of the "bunch" (and then out) every few cm thus, the net inductance is increased and the vast wilderness of unused copper in the centre is much better utilized for the same amount of copper used.
All strands take turns at moving to the centre. In fact if you could cut through a section of Litz wire that is carrying a high-frequency current, all the strands are carrying the same current thus, even though the overall cross-sectional copper area is a bit smaller, the avoidance of skin effect will more than counter this issue.
The technical downside of Litz wire is that if you are both passing a high current at DC and high frequency currents, you lose out on the DC losses. So, you have to decide what is best for the application.
You can also consider copper pipe as a better alternative to what you propose because, although it leaves a big hole in the centre, it doesn't disrupt the mid to outer regions. In fact small copper pipe is used for this very reason.
Sure. The cable can be crushed or drawn, forming it into a dense, often rectangular, form.
A quick googling of relevant keywords finds:
Why Preformed Litz Wire Is Best for High Current Magnetic Devices | NEWT, however, amusingly they don't mention numbers at all -- this article is no help to the engineer; perhaps it only exists for SEO?
Hofer Powertrain gives a claim of up to 65% fill factor, though it sounds like they're positioned as a motor engineering and winding production machinery design company, so maybe aren't much help to us on the more power electronics side of things. There are additional issues with forming windings into 3D shapes required for motor application, which I imagine they have specialized solutions and machinery for.
It's not clear to me just how much stranding is called for in motor application, but I've heard that high frequencies and extremely high RPM are a viable route (including designs like ferrite-cored reluctance motors) for reasonable power density at very low losses, and perhaps that is such an application they are targeting. Stranding or layering is also relevant in very large motors (MW+), and generators, where for conductors sized for currents of 1kA+, even mains frequency experiences significant skin and proximity effects. And anything inbetween; EV applications in the 100s kW, and at somewhat higher frequencies, might easily run into the same limitations of solid wire.
Litz Wire: Round, Rectangular, Formed and Compacted Profiles | OSCO is more-or-less NEWT's copy, but does include some numbers at the end, claiming up to 85% fill factor.
On their inventory page, they list an oddball range of types. I imagine this changes from time to time, but at the moment I only see a couple that would be useful for modest-scale commercial power conversion applications (say, something near 100 x 40AWG for example).
Most of the other results are either connected NEWT pages, alternate domains of theirs, or academic articles describing the use of such materials (which, I suppose I should check a few of them to see if they mention disclosures; with the degree of SEO here, I wouldn't be surprised if some are sponsored...).
My experience with regular (uncompacted round served) litz is, about a 30% fill factor is typical.
Note that a higher fill factor means more strands in a given area, and more strands means more proximity effect; in turn requiring even more, of finer, strands. So, expect to use something a couple AWG finer still, than would otherwise be calculated.
The take-away here is, I think:
These materials are special purpose. There are far too many parameters to just produce every dimension imaginable; and if a given cable doesn't fit exactly into your winding window, the awkward gap left at the end may well be worse than if you had used uncompacted round wire in the first place(!).
These are low production, specialized materials; few make them. Clearly, NEWT is one such manufacturer, and has taken the effort to dominate SEO, at least in my search results; YMMV.
(I imagine there are Chinese competitors, but they are hard to search for -- not just because of the language barrier, but many have piecemeal if any presence on the internet, and you more or less have to be local, on the street, to find them, as I understand it.)
These are specialized, highly customizable materials; it also requires production machinery to make it. It's not something you can just drop in some samples and test a dozen prototypes; NRE (non-recurring engineering) costs most likely dictate production runs of thousands to amortize the cost over. That's going to be a bigger win in something like automotive (where 100k production runs are "a market failure"), than in low-rate or specialized application.
I recall NEWT isn't (or, wasn't as of like a decade ago I last quoted something there) interested in production runs under 2km or so, and in larger sizes at that; more like 10 or 20km for light/commercial duty wire.
To combat eddy losses the wire needs to have individual small diameter strands that are insulated from each other.
To reduce skin effect, Litz wire is used where the wire strands meander from the inside of the bundle to the outside and are twisted.
To reduce proximity effect, bunched insulated conductors simply twisted will do the job.
Proximity effect is the dominant AC loss mechanism in magnetics.
An interesting article was published using stranded wire that used uninsulated strands. They demonstrated that eddy losses are reduced, but not as efficiently as using Litz wire.
