Litz wire at VHF frequencies and best core material?

[gbugh]

Per equations and charts like this one:

**broken link removed**

is it true that Litz wire won't work at VHF frequencies, like around 110MHz?

If yes, is it still good to use something like silver plated flat copper braid for high power RF transformers?

Would it work almost as well to use something like 1/8th to 1/4 inch diameter soft copper tubing for a primary of a 200-300W RF transformer? I probably will need just need a single turn.

I'm thinking of using a -17 powered iron core else a type 61 ferrite. Is there a core material better than either of these for a very low loss VHF DC-DC converter or a VHF power amplifier? I know powdered iron is repeatedly recommended in all the ARRL literature but the right size of type 61 ferrite, should not saturate and could be smaller, yes?

73
AF5IE

 

[FvM]

"won't work at VHF frequencies" is probably the right answer to your question.

The skin depth at VHF frequencies is low, about 7 µm at 100 MHz, so th effective cross section of a 1/4" tube would be only 0.13 square millimeters. https://en.wikipedia.org/wiki/Skin_effect

I don't know the intended impedance of your transformer and can't say if it' still acceptable. You get at least lower AC resistance with Litz wire. Just a matter of calculation.

 

[vfone]

There are small air core inductors dedicated for VHF/UHF, using thin Litz wire. Now depends by application.
https://www.coilcraft-cps.com/products/air-core.cfm

For a high power VHF/UHF power amplifier have to use relative thick copper wire, about 18 AWG (1mm), enamel or silver plated.

 

[gbugh]

Thanks to you both.

Over night I started remembering my past experimenting with magloop antennas. This morning I searched the yahoo magloop group and found this:

"On the interesting and little understood subject of woven coax braid as an
RF current carrying conductor I’ll give you a couple of sobering anecdotal
examples from personal experience. In a former life whilst developing a
prototype STL for a military application I once upon a time used a couple of
short fly-lead sections of braided heavy Cu conductor to temporarily connect
from the fat copper tubing main loop element to the vacuum tuning capacitor;
and the braid quickly got quite hot under heavy Tx key-down CW excitation!
Losses are actually much higher in braid at high RF currents than for solid
Cu wire, flat strip, or tubing. What does this empirical observation tell us
about the Ohmic loss at RF?

At radio frequencies stranded / woven / braided wire can have several times
the loss resistance of a smooth solid conductor. Our military loops were
fabricated from polished large-bore copper tubing, electroplated with high
purity electrical grade silver to one or two skin depths, and coated with
two-pack epoxy paint for robust environmental protection. That of course is
an extreme example that an amateur ham radio application can compromise
without throwing away too much performance and satisfaction from STL antenna
construction.

Another example: In building a QRO RF amplifier tank circuit on 10 meters,
flattened braid interconnect removed from an RG-213 cable typically will get
so hot as to melt the solder, whereas a number 8 or 10 gauge solid Cu wire
or 1/4 inch polished Cu tubing or flat Cu strap runs as cool as a cucumber!
This is how dramatic the loss differences are. This is also why high power
Tx tank circuits carrying large RF circulating currents commonly use silver
plated copper conductors. This is analogous to the very large RF circulating
currents flowing in a high-Q STL resonator.

The ohmic loss in an RF carrying conductor is considerably increased by
surface roughness (in the order of the skin depth, lateral: current hot
spots, longitudinal: long current path) due to woven braid or multi-stranded
conductors. It's instructive to consider why this is so.

Skin effect occurs because the outer areas of a conductor have less magnetic
flux surrounding them than do the inner layers. While we often consider the
effect on materials without weaves, think about how this affects woven
conductors like braiding or Litz wire. Any individual conductor has HIGHER
impedance when it moves into the centre areas, and if current finds an
alternate path it moves out to the lower-impedance outer layer. If strands
touch with high resistance connections, it effectively adds unnecessary
resistance to the path. The fact the surface is rough also decreases
effective surface area for current for a given occupied physical area,
because there are non-conductive air gaps in the occupied area of the braid
weave.

Smooth round copper has lower RF resistance per unit length for a given
surface area, but wide flat copper has less reactance and lower overall
impedance. This is because fewer magnetic flux lines encircle any given area
of wide strip than enclose the surface area of a compact conductor. In
effect the magnetic field is "spread out", reducing inductance....a relevant
factor in obtaining a maximally large enclosed area loop conductor for an
STL that tunes both 40m and 80m without becoming self-resonant below 7 MHz.
Braided and woven conductors have high surface resistance to RF currents and
is the reason why lossy braids are never used for low inductance lightning
grounds which are always made with smooth flat copper strap conductors.

The increased RF resistance (and impedance) is why braiding should never be
used for high frequency high current applications. Parallel stranding
decreases effective current carrying surface area at high frequencies, but
at least avoids the weave problems. The weave construction is what causes
most of the loss increase when typical coaxial cables are contaminated by
water. The problem occurs, even after the cable is dry, because the hundreds
of weave contact points corrode and make poor connections, increasing the
resistance. Something similar happens when we "unpack" a braid from a
jacket. The pressure between weaves decreases, and series resistance (and
impedance) increases.

Skin effect pushes the current to the outside surface, so every time a
strand weaves inside the adjacent surface conductors it feeds current to it
via contact pressure. The current flows through thousands of low-pressure
unsoldered joints along the surface. If the braid has pressure between
strands, and if the braid is clean and shiny, the contact resistance has a
small effect. If we let the braid tarnish or lose contact pressure, the
surface resistance goes way up.

For example typical new shiny copper coax braid is a long parallel weave
only one or two layers thick. It can have around 4 times the resistance of a
solid conductor coax when clean, shiny, and when lays are packed together
tightly in weaves. If you take the braid out into open air and let it make
poor contact between weaves, or let it tarnish, the RF resistance goes way
up! This is actually the cause of nearly all of the loss in cables that have
been wet inside after they drain. This is why, if we carelessly let water
wick into a typical braid shielded cable, the loss goes way up. The loss
does not recover when the cable dries, because the braid has been tarnished.

However, a cable with Cu or Al foil under the braid such as RG6 etc will
often nearly fully recover from water ingress, but a cable with woven braid
only will be ruined.

Having said all this, at the end of the day STL loop construction is always
a compromise to varying degrees. One must always make judicious trade-offs
appropriate to the application at hand. The design and construction of
portable STL antennas represents quite a challenge.

The golden rule for realizing an efficient STL is to eliminate every
milliohm of ohmic loss counts; it counts big time, lest one risks a middling
or wet-noodle antenna!

The short answer to your question is to always interconnect your STL loop
conductor to the tuning capacitor using wide flat Cu strip, ideally silver
plated.

73

Leigh VK5KLT"

 

[FvM]

I think, the quoted article is treating the problem on a "prescientific" level. There are many good points like an explanation why strands of unisolated wires are worse than massive copper conductors, but it seems not to understand the functional principle of litz wire.

I won't exclude that there additional effects that reduce litz wire performance at high MHz frequency like non-ideal stranding, but they aren't derived in the article and I didn't yet see a serious paper about the topic.

 

[WimRFP]

At VHF, you will never reach the saturation limit as at relative low flux density, volumetric loss is high. Even at around 50 MHz, material 61 has Q factor of around 1. So you may go for material 67.

Do you need a wide band transformer, are there strict leakage inductance requirements, insulation (safety) requirements, impedance levels, etc?

If it is narrow band, maybe you can live with an air core transformer.

Very low-loss design takes time to find a good trade-off between core loss and copper/dielectric loss. When using ferrite, you need to test the transformer at the flux density that is used in the real application as ferrite material is non-linear. From my own experience I had a transformer at HF (with 61 material), the large power core loss was twice as high compared to the small signal loss! In other words the Q factor fo the primary winding halved

At VHF, forget real litz wire (so I mean the wire where each strand is insulated), it doesn't bring you anything but high cost.

Regarding flat strip or round wire, this depends on your transformer design. With flat strip I can generally get less leakage inductance (but increased capacitance), and the handling is easier (in my opinion).

Regarding braid, the only reason to use it is where flexibility is of importance

 

[gbugh]

Do you need a wide band transformer, are there strict leakage inductance requirements, insulation (safety) requirements, impedance levels, etc?

If it is narrow band, maybe you can live with an air core transformer.

It is for a narrow band but one reason I haven't considered an air core is that I want to minimize radiated EMI. Whatever core material I use, it needs to contain the flux locally with low loss rather than letting some radiate away. I'll look at type 67 ferrite, thanks.

 

[WimRFP]

As long as the actual size is well below say 0.05 lambda, the radiation loss of the prim and sec inductor will very likely be less the the ferrite loss. You may of course expect radiaton loss due to common mode issues, and then ferrite material can be helpful (such as 43 material).

If you are afraid of near field loss in nearby objects, you may use copper foil/plate screening.

 

[gbugh]

Yes, I'm worried mainly about passing FCC emissions/interference testing. That's why I wanted to reduce radiation. I guess I could have an air core loop with ferrite material/foil shielding around the loop rather than through the eye of the loop.

 

[vfone]

I see you are an amateur radio.
From what I know, for almost 100 years of radio experiments, the last thing that an amateur radio was thinking during his experiments was how to pass FCC emissions.
First they try to "tune for maximum smoke", and latter check for any complains :-D

 

[gbugh]

LOL! I know what you mean. But I'm hoping I can design something to sell to supplement a limited retirement income so I'm hoping I can control the emissions. :wink: