air-wound inductor Q

I want to make a notch filter at 1 MHz, to pull out a fundamental sine wave and leave the harmonics for THD analysis. The signal levels will be pretty high, maybe 50 mA p-p with some DC bias, so I'd rather not use ferrite cores, in case they might add distortion. (I have no idea how much distortion they might add.)

So I could wind air-core inductors. The range might be around 100 uH maybe. I could get a, say, 1" diameter coil form, maybe 0.25" high, and fill it with copper. Maybe Litz wire. This is getting old-timey.

I think some people here have been working with coils in this range. Any guess as to what sort of Q I could expect?

Would putting such a coil near a PCB ground plane much affect the L and Q?

I'll have to try it, but I may as well ask first.

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

From the Radiotron Designer's Handbook, 1940 (so, like, very up to date), with addenda from me:

  • Under 3MHz Litz wire is mentioned.
  • Shield diameter should be twice coil diameter.
  • Shield ends should not come within one diameter of coil.
  • Bakelite forms work with little loss in Q (remember: 1940) You probably want a B&W coil if you're not going to hand-wind from Litz
  • Q increases with coil diameter
  • Q increases with coil length, rapidly up to a length:diameter ratio of 1, slowly after that
  • Optimal wire diameter to pitch ratio is about 0.6 -- B&W coils seem to use this ratio

formatting link

Folks have used polycarbonate forms (pro), and PVC pipe (not-pro), but I don't know how it compares to air -- and the B&W coils are premade and have minimal dielectric in 'em.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

Oh, YES, it will HAVE to be Litz wire at 1 MHz.

The RF guys seem to be able to get VERY high Q values with properly constructed air-core coils. The capacitors need to be high-Q also. Mica for big ones, maybe NPO ceramics for smaller values.

Sure. if the axis of the coil winding is perpendicular to the ground plane, the plane will act as a shorted turn, and kill the Q. If the winding axis is parallel to the plane, it will be better, but still don't want it to be too close.

You have to worry about coupling to anything within several diameters of the coil, if it is not enclosed in a shield box.

Jon

Reply to
Jon Elson

The best old-timey shielded coils used a silver-plated shield box, to increase the Q that much more.

John, you never mentioned your target Q, or how hard you're willing to work to make sure the surrounding circuitry doesn't load the coil -- the more you let the tank be unloaded, the less power you can transmit.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

Like one of these?

Reply to
bitrex
[snip]
[snip]

And I thought my copy was old (1952, Fourth Edition ;-) ...Jim Thompson

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

Looks like I can work with Q=100, but more is better.

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

There are lots of commercial coils like that around, with good Qs, but they are usually nanohenries.

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

Egad. That's going to be a challenge -- you may need cascaded filters, just to get the notch deeper.

Hey -- notching out carriers is a common spectrum analysis thing. Is there any prior art you can lean on? Maybe you should make a cavity resonator -- you got any friends with 300-foot long hallways?

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

Only the best, man, only the best.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

My copy is plastered all over with "Property of the Stanford University Aeronautical Department". I can't imagine why they retired the book.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

I think a ground plane anywhere (near) will spoil the Q. (But that's from a big as commercial coil on a ferrite, resonance in the 10 kHz range. W/ ground plane Q ~20 w/o ~Q=35.. but not particularly high Q).

Air coils pick up lotsa stuff (in the air) are there any ferrites at 1 MHz for a pot core?

George H.

Reply to
George Herold

At Q=100, two series traps get me around 60 dB, if they are properly tuned. My scope FFT seems to be good for about -50 dB on harmonics, so that combo gets me down below -100.

Maybe I should look into RCs, like twin-tees or something.

Or a big chunk of coax.

Crystal filter? Crystal as a series trap?

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

That's prime territory for pot cores, but I'm concerned with ferrites generating harmonics. I need to measure -80 dB THD.

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

Ouch, we make a big air coil.. ~1000 turns of litz wire but resonance is 2-3 kHz, Q ~100

How about a few stages of active (opamp) filters? (I'm assuming you can put two Q=10 stages in series and get ~100, but I've never done it... it's probably fussy to tweak up, but for a one of...?)

George H.

Reply to
George Herold

In the audio range I was looking for the 3rd harmonic of my oscillator, it was much better to put a (2-pole) high pass at the 3rd harmonic than a band pass... Everyone was helping to kill the fundamental.

George H.

Reply to
George Herold

I routinely make high-performance inductors with ferrite cores and with Qs over 100. The cores have air gaps and there's very little distortion at modest signal levels. They are fairly large in size, RM8 or RM10 types, etc.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

You need to talk to Mikek.

Or read this site. Everything that has been discussed here is covered.

formatting link

--

Rick C
Reply to
rickman

OK, I built an coil on a 1" form. the form is 1-5/16 inch long. I wound 20 turns of 660/46 litz. I came back with 20 more, then I turned around and wound 20 more, I had enough wire for 6 more turns. So the coil has about 66 turns with 3 layers and a fourth layer with 6 turns. The coil inductance is 80.4 uH. I would have been pretty close 100uH if I had enough wire for the extra 14 turns. It was about 25 ft. I used the 3db method with my scope and got a Q of 209 at 1MHz. I used decent materials and a high Q air capacitor. You might be able to do a little better but not much. I have a minor Self Resonant Peak at 1.39MHz and a major Self Resonant Peak at 2.8MHz. I don't understand the minor peak. No connections, a one turns loop made with a clip lead as a driver, scope probe just close enough to pick up a signal. I have some 2616 4C6 potcores, that would be interesting. Maybe tomorrow. I have no clue about harmonics. Mikek

--
This email has been checked for viruses by Avast antivirus software. 
https://www.avast.com/antivirus
Reply to
amdx

Your post didn't show up until after I posted. I wound a 80.4uH coil, (I ran short on wire.) and measured it. See my post.

Mikek

--
This email has been checked for viruses by Avast antivirus software. 
https://www.avast.com/antivirus
Reply to
amdx

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.