Common mode chokes high R or L

It will be both. Read some material datasheets, and you'll spot a pattern: material mu is ~inversely proportional to the material cutoff frequency. Typical mu = 10k falls off at 100kHz, mu = 2k falls at maybe 2MHz, mu = 800 falls at 10MHz or so, etc.

The resulting choke will be inductive below the cutoff, and lossy above.

Typically, mu(f) is not a simple pole (i.e., inversely proportional to frequency above cutoff), which would be the case for ideal eddy current loss (which would give Z ~ constant resistance). Rather, impedance continues to rise with frequency, more gradually. Eventually, winding geometry takes over and capacitance and more complex resonances dominate the impedance at high frequencies (10s to 100s MHz, up to a few GHz for chip ferrite beads).

As the designer, it is your task to select a component such that the L doesn't matter much, and the R dominates in the filter transition band so that the filter is damped.

For power line CMC applications, maximum impedance at minimum frequency (particularly, say, 0.3-3MHz) is typically the goal, and so a high-mu material is used. OTOH, for RFI immunity purposes, a ferrite bead has just the impedance and frequency range needed. Besides being physically too large for the job, the former type of CMC would also be a big fat capacitor up there.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
Website: https://www.seventransistorlabs.com/

Lossy is generally best because you don't want the noise sloshing back and forth on the line before the choke. Regarding the power input it is very important to ground the GND wire right at the point of entry into the metal enclosure and provide a ceramic capacitor from V+ to chassis at that very point as well. Not an inch down the line on the board.

For the RF line ground the coax also right at point of entry. If you have a BNC jack there that is already a given (do not use the isolated ones).

I can't count the times where I was called out to an EMC case and found that ground connections had been made via pigtails of various lengths. Or to put it bluntly, pigtails have provided some of my income :-)

--
Regards, Joerg 

http://www.analogconsultants.com/

A boy could get rich off star grounds.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Back in my TV repair days; I encountered a stock fault on the Philips G11 chassis CTV. In the main power board was a choke (not common mode AFAICR) with a Q spoiler resistor in parallel, the Q spoiler resistor often failed O/C and the choke generated destructive spikes at switch off.

My guess is a lossy core would be better than one which facilitates a high Q inductor.

For lowish currents, I ues a resistor instead of an inductor or a bead.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Powdered iron is usually cheaper than ferrite. #26 has a Q of maybe 10 at

100kHz, dropping quickly above there.

Probably less common these days, what with miniaturization and increasing switching frequencies.

Of course, we're not talking about switching, but filtering. But, switching dominates the inductor market, so we have to swing with it.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
Website: https://www.seventransistorlabs.com/

Maybe a little more info is needed.

I have an Active antenna at the top of a 20 ft pole with another 4ft of pvc pipe with the antenna at the top.

I have buzz and want to try choking the feedline to see if that helps. It may be from the antenna, but that's what I'm trying to figure out.

Mikek

Not really finding that, I find about 4 different frequency range beads for coax. The low frequency is what I'm interested in, mostly the Laird beads, link shown below. Disappointed they don't break out L and R, only give Z.

Besides being physically

Ya, I should have given more info.

I have an Active antenna at the top of a 20 ft pole with another 4ft of pvc pipe with the antenna at the top. Freq Range, 10kHZ to 30MHz. Here's a link,

I have buzz and want to try choking the feedline to see if that helps. It may be from the antenna, but that's what I'm trying to figure out.

I find there are several mixes and only Laird give a typical Z per bead of 9 to 61 ohms for there Low frequency material. It has no breakout of L and R. I sent an email asking yesterday, no response yet today. Polomar gives only frequency range of their products.

See pgs 3 and 15

... 200717.pdf

I had to battery operate my radio, that got rid of 2.5 S units of buzz. Still have about 2 more S units. I put seven snap on beads over the AC power cord, no change! I wound 14 turns of small coax in a huge potcore I, have which should be 1120 ohms XL at 500kHz, I put that at the radio antenna connection. It made no change in the noise. That makes me think doing the same at the antenna will not help, but I still need to try it.

Mikek

Well, or good coupling to keep common mode operation up to x Mhz, one wold think the answer is "low loss". BUT... You want the higher frequencies to be adsorbed and perhaps coupled if CM bazzfazz. Best way is to do both and add lossy toroid on each line going into filter, letting those to at least partially saturate.

It's a one-port, so Kramers-Kronig gives L and R.

Example:

I suppose Q is more informative than L for ferrite bead purposes, but Q is particularly difficult to recover accurately from such a blurry plot as given.

In other words: my reconstruction is as accurate as their data. The accuracy is terrible below 10MHz, because the plot is a few pixels there. (If only they'd used log plots...)

Gosh, and that's for 30MHz? Hardly looks like it'll do 100MHz.

I guess it's tuned, also, but it doesn't have variable tuning? What's up with that?

Should be using a ground plane too, otherwise it's more like a...uh... hmm, vertical long wire with an electrically short monopole on top? I suppose it still doesn't matter much when the pole itself is hardly fractional wavelength at 30MHz.

They also give the circuit, which looks silly. Small capacitor into choke into small capacitor, but they're not even using the first element for gain, it's a follower? And then another follower? What?

I've noticed there's some real nonsense out there in ham-world. It's a lot like the audiophile crowd, where novelty in amplifier design is more important than any measured amplifier parameter (except perhaps power output, to the extent that it's enough power to listen with at least), including stability (to the extent that it's not actively squealing during a session, that is).

There are definitely experts among hams, but as with all true experts, they are few and far between. Which makes their info very hard to find, as everyone and their dog can publish the latest antenna craze without providing a single measurement at all.

Notice that the "spec sheet" of this "antenna" has the key weasel words: "Each antenna performs as good as its installation permits it to."

Oh, you got sucky reception? Completely dead bands? Oh, you must not've installed it correctly. Keep trying! We're sorry we didn't make this easier to use, or, you know, actually good at all. Just use the ideal setup, but we won't tell you what that is.

The "specs" are equally conspicuous. It's measured like an amplifier in a

50 ohm system. It claims a frequency range, but conveniently omits a radiation pattern or sensitivity factor.

If your amplitude margin is +/-100dB, you can spec it as "DC to light"!

What's the other limiting factor for receiver operation? Noise floor! Conspicuously absent as well.

The dynamic range is embarassing, anyway. 150mA for -15dBm? (Without a distortion or compression spec at that level, mind.)

The antenna is shit, no matter how you cut it.

This is an optimization problem, like any other. Like optimizing a computer program, or tuning in an industrial process. Right now, you're ignoring the orders of magnitude stuck in that POS antenna, for the few dB of offending noise that you can pick up with it.

Well, it might not be "few dB", but that's just how bad that thing is.

Start with a real antenna, like a multiband dipole. This can offer continuous bandwidth in a modest size (with loading coils, 40m can be crammed into the space of a 20 or even 10m antenna). Lower frequencies ideally should be pursued the same way, but the size can get prohibitive, and anyway the atmospheric noise makes an optimal antenna less useful in the first place: an electrically small loop is as good as anything.

A small loop can be wideband (amplified to hell), or tuned. The antenna factor is shit (say, -40dB or worse), but as long as amplifier noise floor is better than the atmospheric noise floor, you're fine. And that can be easily achieved with some low noise BJTs or JFETs, since the loop impedance itself is low. If tuned, you can get reasonable gain (maybe even enough to be worth transmitting with), no need for a preamp. But the tuner has to be at the antenna, usually a motorized variable cap, and you need to spin that while you're scanning the receiver.

Note that I'm recommending balanced antennas, dipoles and loops: this helps alleviate common mode ambient noise.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
Website: https://www.seventransistorlabs.com/

Sometimes I salvage filter cores for DIY inductors, they seem to work well enough - never noticed any getting hot, maybe I just wasn't pushing enough power.

That's the lossest material I've see, to bad I can't get that 1206 chip around my coax cable.

I suspect you mean 100kHz?

It's not tuned.

Voltage probe?

Where are you getting the schematic?

I don't have the schematic, but are you looking at the DC/AC separator on the antenna pcb?

I don't know if the spec's meet the advertised of Second order output intercept point: > + 70 dBm, Third order output intercept point: > + 30 dBm. But, you don't either. I have a down load of someones fet amp with very high intercepts, and plan on building it to compare to this one.

Weasel words but doesn't mean it's not true.

I'm have done a thorough test, but so far I'm getting beacons in the

200kHz range, good signals on 80 meters, good signals on 10 meters. Haven't tested higher. Been concentrating on getting rid of the buzz.

I'd be happy to build your design. Then I can compare the MiniWhip, Dallas Lankford's active antenna and yours.

This was an $18 antenna I picked up, I'll soon be putting up an 80ft long wire antenna to compare this to.

I have a Digikey order started, my address is good, please send me your active antenna design, I'd like to compare it to the others. I already have Norton amps, so need just work on the high impedance, low noise, high intercept part.

Thanks, Mikek

Sorry, I read your graph wrong, that's all inductive at any frequency I'm interested in. And useless for my purpose. Mikek

Maybe you managed to avoid the yellow ones? Or, if you were only doing a few watts, you wouldn't notice.

Here's a chunk of one, at ~1MHz:

A rather unfair test as the induction is large (think that was around 50W into a ~5/8" i.d. coil -- wire is 12 AWG), and losses go way up once the resin cooks out. It took some minutes to get there.

They're cheap, which is the only redeeming quality about them over other options -- but if you can afford the space and the wire, you can use enough excess inductance to keep ripple tolerably low.

Yellow-white is #26, and green-blue is #52 (with only slightly lower losses). Most any other powdered iron composition is much better on losses, but also much less common.

Ferrite is much better, even in the higher loss grades -- mainly because mu is so high at the same time, so the magnetizing current (and thus its loss) is very small, for transformer service.

I once made a transformer, using a ~1" ferrite toroid at about 0.2T peak,

50kHz. Wound with loose wire, there was little cooling, and as the core heated up, it began to saturate, and heat faster. Replaced with a bigger core, runs cold.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
Website: https://www.seventransistorlabs.com/

I'm placing my order this afternoon, let me know if you have a better design than the Active antenna I bought, so I can add the parts to my order. Mikek

That's a very interesting antenna. Allow me to post a link to the designer's webpage:

Please share your findings with the group.

Thank you,

--
Don Kuenz, KB7RPU

This guy is local here and I often hear him on the monthly (1st wednesday) local roundtable, so when you have questions for him I could try asking them... (and we have a nice topic)

Not sure which guy, the pa0nhc? I doubt it's Lankford, are you in Louisiana?

Mikek

Yes, PA0NHC, the guy who wrote that webpage about the antenna. Lookup PA0 in your prefix list. It is not in the USA.