Litz wire Q

30MHz? I hear 3MHz or so is usually where it loses value. There's probably something about capacitance between strands, too -- the eddy voltage (i.e., the voltage between strands, which would've caused eddy currents in a solid conductor) just keeps going up.

Going up much higher requires extremely fine wire to remain valuable. Finest litz I've heard of has strands around #46, which is finer than human hair. I don't think they make much finer than that.

Tim

--
Deep Friar: a very philosophical monk. 
Website: http://seventransistorlabs.com 
 Click to see the full signature
Reply to
Tim Williams
Loading thread data ...

Hi,

formatting link
shows 2.8MHz max recommended for 48AWG litz wire. I guess litz is probably used for a lot higher than that, but the Q will go down. For 30MHz the skin depth is

13.9um! I think 48AWG wire is about 30.48um diameter (0.0012") so you would need some higher gauge litz wire maybe if it is made!

cheers, Jamie

Reply to
Jamie M

Nominally,what is the highest frequency that litz wire been used at? I am guessing 30Mhz is the top end.

Reply to
Robert Baer

What sort of frequency and Q are you targeting?

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

For local oscillator or front ends tuned by a typical 500 pF variable capacitor, the inductance required for the long wave band is about

2000 uH, for medium wave band bout 200 uH and for some HF band 2 uH. Since the inductance is proportional to the square of number of turns, so for similar cores, the MW coil requires 10 times the number of turns than the HF coil and th LF coil 30 times the number of turns than HF. The linear length and hence losses are proportional to the number of turns. If you try to squeeze 10-30 times the number of turns on a coil former, you have to use thinner wires, increasing losses.

Thus the Litz wire is critical for LF/MF, but for HF/VHF, the low number of turns makes it possible to use a thick conductor, but actually only use the surface due to skin effect.

Reply to
upsidedown

I think of it this way: although the skin is getting thinner, it goes as

1/sqrt(F), whereas reactance goes as F. So at higher frequencies, Q is simply high enough not to care.

You can still get some amazing Qs with litz. I've heard of Q over 1000 for a spider wound type with the finest litz. That's up there with polypropylene capacitors, maybe even air variables. I think that was LF/MF, don't recall if there are measurements and comparisons at HF+ for anything.

I don't think this works for transformers though. Transformers are about two things, flux density and current density. Even for optimal core materials, flux density drops (loss limited instead of saturation limited), but it's more like 1/sqrt(F), while flux goes as 1/F. So the core gets smaller, and length of turn gets smaller. But copper is still copper, and if you can only divide it so far, you're actually going to need the same or more!

Tim

--
Deep Friar: a very philosophical monk. 
Website: http://seventransistorlabs.com
Reply to
Tim Williams

I once designed some audio-range FSK modems, which included some pot core/Litz inductors and polystyrene caps with LC Qs around 1000. Recently I designed some

50 MHz burst oscillators and experimented a bit and wound up using stock Coilcraft inductors and ceramic caps with resonant Qs around 80. Surface-mount 1008 inductors were closer to 30.

Burst (gated, instant-on) oscillators are fun.

formatting link

The trick is to have every edge, especially the first few, happen just on time. And to have the prop delay from gate to the first edge be minimal.

Signal-level transformers don't usually care about flux density, but a transformer still needs two (or more) windings.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

None specifically; just a randumb idea that was partly linked to Litz wire. Much stronger link to surface smoothness VS frequency - as seen in Rogers (and other PCBs). Am thinking,all else being kept constant, that a better surface smoothness gives a better frequency frequency response/capability. A little hard to see why the loss at a given high frequency wold be less.

Reply to
Robert Baer

Somehow what you wrote cot completely lost, so with apologies i copy some of it here to give a more coherent response: " Burst (gated, instant-on) oscillators are fun.

formatting link

The trick is to have every edge, especially the first few, happen just on time. And to have the prop delay from gate to the first edge be minimal. "

I must commend you,the turn-on is virtually perfect; only the first half-cycle is slightly wider. This turn-on response is not easy to achieve; i lucked out in trying that when i tried an ECL (NOR?) gate where one input was the on/off control and the other was for the feedback for oscillation. Been over 30 years, but i think that yours is better.

Reply to
Robert Baer

For picosecond stuff, or things like microstrip filters, copper foil smoothness really matters. The price paid is adhesion. Most PCBs have black-oxide copper finish on the copper foil that's supposed to stick to epoxy, and that's awful electrically.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.