inductor problem

I have a pulse generator circuit that needs a roughly 350 nH inductor. RMS current will be about 6 amps, and I used four Coilcraft MiniSpring air-coil inductors series-parallel, running about 3 amps RMS each, which should have been conservative. But they get very hot at well below max current, which I'm blaming on skin effect. The waveform is pulses, but rougly equivalent to 30 MHz maybe.

I wound my own single inductor from #22 bus wire, about 0.2" dia x 1" long, and it gets pretty hot too. DC-to-output power efficiency is a little better but not much.

So, for you RF guys,

I used regular tinned #22 bus wire. Would bare copper, or magnet wire, be much better as regards skin effect heating?

Would Litz be much better at this sort of frequency? It would be hard to wind as a self-supporting air core inductor.

Might any core material help? That would trade copper loss for core loss. I might have an old MicroMetals powdered-iron toroid sample kit around here somewhere.

I might try making four air-wound inductors from #20 bus wire, to replace the four separate Coilcrafts. Might help. Probably not much.

Maybe I just need a huge inductor. I'll have to spin the PCB layout for that.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin
Loading thread data ...

The skin depth isn't much (65 um); you might consider using the center conductor wire from RG59 coax, which is a copper-clad steel wire. The steel will hold shape even if the copper gets toasty, and doesn't hurt the RF conductirity. Air-cooling and high temperature tolerance is a potential winning combo.

Reply to
whit3rd

I copied this RF (~70 MHz) Rb lamp driver. The original used silver coated copper for the coils... (~7 turns #20 wire, ~1/2" diam.) I used magnet wire and it worked fine.

Phenolic? How about some sort of copper tape?

George H.

Reply to
George Herold

Litz will make a _big_ difference... run the numbers... solid wire cooks you all the way up to the 9th harmonic being significant.

Wind over polystyrene or similar form. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
 Click to see the full signature
Reply to
Jim Thompson

They publish models for those things. You didn't SPICE it?

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

Silver plate the wire. Or just try winding it with silver wire.

Reply to
tom

Winding with silver wire might work. Silver-plating was popular early on, until somebody measured the conductivity of electroplated silver, which is less than that of solid copper.

Laser annealing of plated silver might work.

--
Bill Sloman, Sydney
Reply to
bill.sloman

At 30Mhz? Skin depth is only 12um. Even a #48 wire is overkill, and you would need 660 or 1200 wires in the bundle. I'd go with a 12 gauge wire it was 3 times more area at a 12um depth.

1/3 the resistance to heat and more surface area to dissipate it. Although, it probably can't be wound as tight and will take a longer length of wire, so drop the 1/3 less R, but probably less R. That's the opinion of a 3/4 retired shrimp seller.

Ok, I didn't run the numbers for the number in the litz bundle, could be less, might be more.

Mikek

Reply to
amdx

So, wait, you're pulsing >200V across this poor thing? More than 600VAR? And you're worried that it's getting "very hot"?

If it's not glowing red hot in ten seconds, I'm not disappointed, I'm impressed!

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

The resistivity & tempco of copper & silver are so close as to make no significant difference afaic tell. Parallel thinner wires give more surface area per xsa.

NT

Reply to
tabbypurr

Polystyrene would melt. If I use any form, it will be ceramic, preferably AlN.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Oops, I got the decimal point wrong; it's 0.65 um. Not sure about Litz wire, though, because you can get good results by just using a larger diameter of wire (or tubing); I've never seen fat Litz wire.

Reply to
whit3rd

(must have forgotten to hit the recalc button)

Lower resistance comes from larger diameter or Litz wire, or both: but, I've never seen fat Litz wire. This would be a good app for that copper-over-aluminum that folk complain about in motors...

Reply to
whit3rd

Litz loses its efficiency above the MF range.

Get an ARRL handbook and have a look at the HF final amplifier coils. They are usually made from copper tubing and maybe silver- plated.

--

-TV
Reply to
Tauno Voipio

Getting rid of the tin plating would help. I'd add my vote for some nice fat magnet wire.

How are you handling the EMC issues?

Cheers

Phil Hobbs

Reply to
pcdhobbs

A little more. #22 wire has an area of 11,945 um^2 if you assume 12um skin effect. #12 wire has an area of 38,509 um^2 if you assume 12um skin effect. #44 wire has an area of 848 um^2 if you assume 12um skin effect. So, 45-#44 wires would equal one #12 wire in skin effect surface area. However litz has some additional losses, so if you go with 100-#44 wire litz, you would gain a lot. A 100/44 litz wire is 0.031" in diameter. Making a small coil with litz is a pain, I would try the 12 gauge wire first. I have 660/46, it is 0.70" in diameter, a little smaller than 12 gauge wire. I'll send you a few feet if you send me an email. My address is good. Mikek PS. I have read many times that litz runs out much above MW frequencies. Don't know if that's because of proximity effect or what. But, you certainly get a surface area increase. RE my previous post, I didn't find #48, so did the calcs on #44.

Reply to
amdx

Yes, but I may just need cubic inches to work in, which means a board spin. Well, we sort of expected that. I could maybe use a forest of old-fashioned axial-lead RF inductors.

I'm not! That's the customer's problem; I'm just furnishing a PC board.

I tried to measure the top temperature of one fet; it felt pretty hot but my Omega thermocouple meter insisted it was 3 degrees C.

Easy for you to say.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Huh, that's weird. You figure the first guy to try silver plating copper would measure the resistance.

George H.

Reply to
George Herold

#22 is like trying to power a toaster with door bell wire :-)

Litz isn't enough here, you should have a flat conductor and most of all large. Either trace inductance (with or without core) or as parts.

It would help but first you need to get some serious copper on there.

How about something like this, two in series?

formatting link

Even more serious:

formatting link

IMO one of the best sources for high-power inductors is Coilcraft but they sell only direct:

formatting link

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

The tin is far lossier than the copper, and it's in exactly the wrong spot. Even #22 magnet wire might be good enough. Bare copper would be fine at first, but get worse as it tarnished.

And repeat!

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

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.