FANTASTIC tutorial. 30+ year BSEE and Electromagnetic major here. You know your subject and can present it well. Thank you for all the work put into this.
Excellent presentation-accessible to the newcomer but with enough information to whet an Elmer's appetite. It spurred me to buy the Halibut Electronics kit, and now I need to build ALL THE CURRENT CHOKES!
First of all, nice video, one of the best I've seen on building a CM choke. Only one minor comment: "Building something and trying it" is NOT really an "engineering approach". It is actually an "experimental" approach. As an engineer for 43 years, I found that taking an "experimental approach" ended up taking up a lot more time in the "development phase" of a product than working the design out mathematically before laying any lines down on a drawing. Having said that, I frequently approach my amateur radio projects as experimental in nature simply because I like building a lot more than I like math. In amateur radio I am not bound by schedules, budgets, difficult test requirements, etc, so I can enjoy the build to "see what I end up with". So, this is not meant to be a negative comment about your video. I found your explanations were clear and accurate with nice visual aids, and more importantly easier to understand than wading through the math. Good overview on setting up and using a VNA, too. --Mark, KE6BB 👍
That’s a fair point. I think what I was trying to say was “practical” vs “theoretical.” To me, that’s the difference between engineering and science. Engineering is more applied, science is more conceptual. But you’re right, “experimental” is an even better term for what I’m doing here. Thanks for the feedback! :-)
16:00 - outer conductors of the two NanoVNA ports (at least on the old 1.5G one, the one made out of three SA612A mixers) are shorted through the device's common ground, just FYI
True. That would make the plane of testing right at the edge of the VNA, including the cables between the VNA and the test rig, which is ... less than ideal... 🙂
I know I'm kinda late (apparently not getting notifications for many of the comments on this video!), but I will point you to electronics.halibut.com/product/common-mode-current-choke-test-rig/ 🙂
Excellent explanation and test work Mark. As another RF guy I am troubled by 2 things. One at 09:20, how can CM noise get inside a shielded radio when applied between coax shield and transceiver "ground"? I suspect all CM noise entry is either via poor coax or transceiver shielding. Two at 22:03 is the assumption that a CM choke should be X times the cable Zo. I doubt that the surge impedance on the outside of the coax is also 50 Ohms. Anybody wiser?
1: you’re probably right that RF gets in on bad coax too. Heck, it’ll get in on the antenna itself! But any non-perfect ground will also put that noise into the receiver. 2: The goal is to make the common mode impedance as high as possible at the target frequency. Common wisdom as proposed by people smarter than me is “about 1000 ohms.” That’s about 20x a 50 ohm system … and I think I’m realizing I did my dB math wrong in the video. That’s 20x impedance (linear with power), not 20x voltage (square with power). So it’s only 13dB, not 26dB. If I’m right, all the numbers I gave in the video are twice as good as they “need” to be. Huh. Well, we have some mighty fine CMCCs out there I guess.
@@SmittyHalibut Yup I agree with the math. Just the relation to 50 Ohms which is the internal cable Z not the outside "third wire". No matter, we need Z as high as possible, in most cases with a good Rloss to gobble up energy. As far as RX CM I think it is an "all cables" sneak in (line cord, USB, LAN...). CM chokes DO help. But as for TX induced RF chokes should be at a voltage maxima on the line where ever that it is. PS It is amazing how far the ham community has come in understanding lines, VNAs etc. Cheers!
Forgive my late input . I note in a few of the comments that many would love to see a comparison against a Coax bottle type Choke and what the attenuation figures come out like . . Great video by the way . Well done 73. Ron G4DIY
Thanks for the positive feedback! I'm not sure what design you mean by "Coax bottle type choke". A quick google search turned up a design where you wrap coax around a pill bottle? Is that what you mean? If so, that's just an air-core choke. The bottle is only a form to wrap the coax around, but doesn't appreciably impact the magnetic permeability of the choke. If you mean some other design, please point me to a page or article describing what you mean.
@@SmittyHalibut hi there again. Well my “bottle type “ is literally a plastic bottle about 8 inches high and about 4 inch diameter . I have coiled Rg 213 Coax around it with about 20 turns . So its a coaxial choke effectively.
@@ronbennett5591 Ok, yeah. So it's an air-core choke. I've gotten feedback elsewhere in this comments section that I never gave a properly-wound air-core choke a fair shake, and that's true, really. I need to make an update to this video with my CMCC Test Rig kit instead of the hand-built test rig in this video; when I do that, I'll give an air-wound choke a proper test.
“OTA” usually means “On The Air” so it’s just a matter of the first letter(s). POTA is Parks, and SOTA is Summits. Both their websites can be found by searching. (Read: I don’t remember the URLs off hand. :-)
I too have done some research and testing to determine the optimum ferrite materials and number of turns to get the most out of my RFI choke filters. I based my conclusions on ferrite manufacturers equations and my personal tests using signal generator, oscilloscope and nanoVNA to sweep the entire HF band (e.g. building a combined differential and common mode filter for the output of a home-brew DC switching power supply for my shack). I too found that, for a snap-on ferrite, 0.5 inch hole, type 31, 3 turns was optimal before capacitive coupling between turns reduced effectiveness. But for an FT-240-31 I think 5 to 7 was optimal, according to manufacturer's equations and testing (this was 3 years ago so I don't quite recall exactly). However, I also built a 1:1 current balun for use as a wire dipole feed point and I put 11 turns (16 AWG PTFE insulated wire, split and wound in two direction) on that one, and it tested extremely well, -35 dB attenuation across most of the HF band. So it seems the geometry of the toroid makes a big difference in terms of how many turns is optimal. And your tests also showed better braud-band attenuation with the 13 turns around a FT-140. Do you have a general explanation for this? Why does the capacitive coupling add up so much with the snap-ons?
Its more about the turns of cable than the ferrite material. When on a torroid, the turns are spread out, more space between windings. When wrapped in a "Gordian knot", they're immediately adjacent to each other, maximizing the capacitance between turns.
I took issue with the description of a coax as a differential cable early on in the talk (around 4 minutes). Signal Integrity engineers refer to them as single ended for a reason or two - a differential cable, for example requires at least 4 vna ports set into 2 logical port groups; and the shield is not the return conductor, it is a shield. It serves the same purpose as the shield over a twisted pair like in an ethernet cable, which is a true differential system. One more comment is that shields of coax are almost always tied to the chassis ground on both sides, not the signaling elements of a transceiver so there is no push-pull of currents. Any differential currents created are simply inductive coupling between the center conductor and the shield. That all said, I love the data collected on the experimental chokes the nanoVNA plots. I wish we could have seen data for 2 chokes placed on a single gordian knot to see how adding chokes improved the common mode rejection also.
I just re-watched that section to make sure I was thinking about the right thing. You are absolutely correct that there is a world of difference between a Single Ended signal (like on coax) and a Balanced signal (like on twin lead, or twisted pair). But that's more about voltages, not currents.. (Yes, voltages which then drive currents though impedances, yes, so it's not entirely NOT about current... See below...) But even a Single Ended signal will have opposite and equal current flows on the coax. My mistake was referring to those currents as "balanced", little b, the adjective. They are not Balanced, big B, the proper noun describing the signal type. That's why it's called Differential Mode for currents, because that's different than Balanced for voltages. The push-pull of currents part, again, you are correct. That's the impact of a Balanced (big B) signal on the currents that we talked about earlier. But that still doesn't mean that there isn't an equal and opposite current flow on the feed line. If there's no other return path for current, it HAS go to on the feedline. And even if there is another return path for current, like you said, it will induce the return current in the coax and probably prefer that path over the other (assuming your coax is well connected and low impedance compared to the alternate path.) So, you're not wrong in any of your response. But I think those points don't really impact the description of the examples presented.
Great video on checking CM chokes. Thanks for doing all the work determining that 13T on a 31 Mix core is the best, fits right in with G3TXQ work that he did on chokes, baluns and ununs.
Thanks for the video. As I was watching I was wandering about influences of possible phase shifts in the device. The influence on the working of the common mode choke. Signal coming in the choke (on one side) and signal comming in (on the opposite side and from the opposite direction) may have some phase shift. The load is probably not pure resistive....
Are you talking about the Differential Mode response, meaning how transparent it looks to the signals flowing through it? Or the Common Mode response, meaning how the phase affects the resistive/reactive balance of the impedance it imposes? If the former, the Differential Mode view, by definition Differential Mode currents sum to zero through the inductor, so there is no impact at all. When I=0, it doesn't matter what the j component of X is because it will be multiplied by 0 to get your V vector. "What if phase differences result in I!=0?" Then those differences would become Common Mode. By definition. If the latter, the Common Mode view, then yes you're absolutely right. The impedance imposed by a Common Mode Current Choke is not purely resistive across the whole band of interest. It's an RLC complex circuit. An IDEAL choke would be all R and very little LC over the whole band of interest, but that's nearly impossible. In the video I show highly resonant chokes with a very deep attenuation "peak" (shown as a "valley" on the graph because it's a negative number). In those cases, if you measured the Complex Impedance of the CMCC, you would see a high R/low X at that peak, because that's the resonant point of the choke, and resonance at the frequency happens when f*L and f*C are equal but opposite, which cancels them out and leaves you with nothing but R. But outside that peak, the response becomes very reactive.
Awesome stuff! When you talked about calculating or experimenting with Gordian Knots to make a choke for a specific band, I had to comment because I've done exactly that. I fly paramotors, wanted to build an antenna for GMRS (462 MHz) out of a chunk of RG316 with half of the coax sheath removed, then string it up one of the glider lines. 2 problems: 1) "Center/core fed vertical" doesn't resonate well on any frequency. Experimenting showed that the angle between the two legs needs to be around 30° (aIrc) for resonance, and 2) Even if I got it resonant, Instill needed to decouple the antenna from the feed line, otherwise the radio body, the hand holding it, any metal or human near the coax would cause wild swings in the SWR. So I added a Gordian Knot, and experimented until I reach both resonance and decoupled the antenna from the feed line. In my case there was 2 loops at around 6cm diameter. Location of the knot along the coax was also critical. Couple of us flew those antennas for a while and they worked well but wore out quickly. Next version will be with more flexible coax, and the knot up higher on the line so it won't get beat up as badly during handling. Also trying to figure out how to do multibands ...lol
That's awesome! (Sorry for the late reply; not sure why I didn't get a notification for this one.) Gordian knots can work well for mono-band operations, if you tune them. And a rig like described here can help you do that. But they are inherently mono-band.
What would happen if one were to place the common mode choke in an iron or aluminum tube? For instance, if someone were to insert a linear common mode choke into the support mast itself to protect the coaxial cable from weather elements like UV rays? What are the potential impacts, and have you ever measured them?
Maybe I missed this but why multiple common mode chokes are recommended (at feed point, at trx, in middle) and why single one at transceiver isn't enough for eliminating "bad things" entering trx?
Great presentation! I am just wondering , when it is about a current inducted in the shield , then probably this conductor is not anymore a feed line with 50 ohms characteristic impedance ?
Loved the engineering approach Mark. I am playing with common mode chokes now and so I am going to build a test box, everything you say about it makes sense. I was hoping to see more detail on ferrite core testing, going from 1 turn to 20 turns or so. At some point putting more turns on the ferrite the interwinding capacitance increases and so more turns is no help, just the same reason that more turns on a knot is no help. It would have been good to see how you got to 13 turns being the optimum for the size core and size coax you have. I am sure that a different size and type of core and different size coax has different optimum turns. The NanoVNA has opened up so much testing to low budge engineers - and thank you for pointing out the calibration issues. This is gold
Hi, I found similar, that interwinding capacitance appeared to be an issue with too many turns of coax on the core. Controversially perhaps my best wideband common mode chokes are wound with twisted pairs from CAT6 cable. Each twisted pair has a characteristic impedance of nominally 100ohms so effectively I have two in parallel to get to 50 ohms. I am getting much better bandwidth and very low insertion SWR.
Hi....Am I right that you still have a "jumper" between the grounds of the sma connectors on the VNA side, just not visible in the picture? I have in mind to build this one and a frankenstein version that uses the inner and outer across a chock to see what the sweeps compare like 🙂....then of course measure the CMC on coax installations before and after chokes...then get down some operating maybe....thanks for a great presentation
Soooo... Good catch. No, on the second version of the box, I didn't have that jumper between the shields on the two VNA ports. I should have. I encourage you to make your own CMCC Test Rig if you want to. I will provide as a "Throw money at the problem" option if you'd prefer, I do now offer those test rigs as a kit from my company's website. I started Halibut Electronics _AFTER_ recording this video, so it didn't come up. If you're interested, it's available here: electronics.halibut.com/product/common-mode-current-choke-test-rig/
@@SmittyHalibut Thanks Mark...the kit looks like decent value given the connector costs alone, I would certainly purchase one if I was in the US, but tax and duty on top make it a little less attractive for me, that said, I will make up the V2 with a jumper and try to persuade my club we need a proper kit ;-) I do have a question....Most of the test arrangements I have seen connect the output of the VNA (centre conductor) to both inner and outer conductors on the choke to generate CMC....It is not clear to me how feeding just the inner (from VNA) to shield (on choke) creates CMC....the sweeps look the same. What am I missing here? GM0EDJ
@@paultemple5566 good question! Remember from the video that current flows in interesting ways in coax. Differential mode currents, where there is an equal amount of return current, necessarily flows inside the coax between the center conductor and the inside of the shield. And common mode current, that does NOT have a return current, necessarily travels on the outside of the shield. Because: Physics. So the important part isn’t which conductor of the Device Under Test (DUT) the signal is put on, it’s more important that the return current is NOT sent through the DUT. If you use coax, even the center conductor, as a single wire without sending the return current through the coax, that single wire’s current will NECESSARILY travel on the outside of the coax. It’s weird, I know. And I couldn’t explain to you now why that is, but it was explained to me back in college *mumble* decades ago, and it made sense then, so I’ve accepted it on faith ever since. So, in my screw up of the v2 test rig, without jumpering the shields together, it just means the return currents are INSIDE the VNA. Which means the two jumpers connecting the VNA and the test rig are also common mode. It will add noise to the measurements, but it doesn’t invalidate them. Does that make sense?
@@SmittyHalibut Thanks Mark....good enough for me and makes sense, hence the jumper to avoid a return current through the DUT....on with build and testing
I haven't seen the entire video so I apologize but thought I should say that TRX (a channel) has a video where he builds and measures the effectiveness of a common mode choke with a NanoVNA. I will come back when and if I see a measurement like he did. If not, go and see what he shows us.
I'm back. I like what I'm seeing very much. Its much more elegant to make a test device box. I think though, you were both doing the same test. That said, what did you think of his initial test of toroid material ?
That said, I'm only seeing you show poorly wound coax baluns, and nothing properly wound. So far anyway. I'm going to wind a tested design balun with a 43 mix and good enameled copper wire 1 to 1 and see if it does what others have measured. A coax balun is not the way to go. But, I do like your testing set up, and have both a regular nanoVNA and a V2.
@@californiakayaker It depends on what problem you're trying to solve. If you have a perfect dipole antenna with identical shape and size no uneven obstructions in the near field a wire wound balun is better, I think we can all agree on that. A wire wound balun won't solve my problem in my mobile between 160m ~ 40m where I am attempting to capacitively couple with the earth. I need coax balums right behind the unun or it all goes to Hell real fast. You get the antenna perfectly tuned then start applying about 500 watts and it all falls right on its face, as the amplifier kicks out because it sees 60 watts coming back. 🤦 G3TXQ and K0BG discuss this on their websites.
How did the 2 clamp on with one coil work. It would use less coax and may give more blocking. Or I am thinking it would. 4 clamp ons should, in my mind do good on one coil. But I am not sure if I am thinking the right way.
It would almost certainly help, yes. The question is, which "costs more?" In dollars, ferrite is more expensive than cable and only increases the inductance linearly with the number of clamp-ons, where as adding turns of cable increases the inductance to the SQUARE of the number of turns. But if your source of cable is finite (eg: it's already in place and only so long), and you can afford to put more ferrites on it, then yes that sounds like the better option. To answer your direct question: I don't have data for what the two-clamps-on-one-loop looks like. I'll take this opportunity to point you to electronics.halibut.com/product/common-mode-current-choke-test-rig/ and ask you to post the results, though! 🙂
Jim Brown (K9YC) had some useful graphs in A Ham's Guide to RFI and also in Understanding How Ferrites Can Prevent and Eliminate RF Interference. You should be able to search for those titles and his name.
A question just came up in my mind: If you use two ferrites in series, does the separation between the two matter? IOW, does the flux from each one couple with that of the other one? I think I'll start saving up for a VNA. Also, I'm curious about one other thing *(NOT* related to coax). If I buy a pre-made common mode choke, can I connect one of the sets of windings backwards and turn it into a *DIFFERENTIAL* mode choke? How would that compare with using two *separate* inductors (wound on *different* cores) for that purpose? _73 de AF6AS_
That's a GREAT video explaining Common Mode Chokes - One thing, you never talk about power handling ability since these are usually used with TX as well as RX situations. The air core coax wound chokes would be limited only by power handling of coax, or close depending on how much capacitance affected power handling. The ferrite cores would be my concern - I think even the clipon ferrites would handle 100 watts, but when 1000 watts is involved it would require more?
Well done mark. I enjoyed your presentation very much and look forward to more.in my opinion this is what ham radio is about.experimentation and sharing. Unfortunately you have to suffer the snide comments of others who know better (maybe).I hope you are not discouraged my freind.
Thanks for this video, as a returner to the hobby after nearly 40 yrs, it was very useful to me. Forgive me if this is a dumb question but please would you confirm that when you say "outside" you mean outside edge of the shield and not the outside plastic sheath of the coax? I ask this because I know you can get RF burns from the very outside of the plastic sheath of a coax. Thanks again.
Not a dumb question at all! The currents are traveling on the shield, that is true. But RF can capacitively couple through the insulation into your skin. If the currents are high enough, then yes you might get burned (more likely a tingle) from touching the coax. Now, that’s a REALLY BAD common mode current. If that’s happening, something is seriously wrong with the antenna or feed system. Under normal circumstances, with normal levels of reflected common mode current, it’s very unlikely you’ll get an RF burn by touching insulated coax. But once that current hits the metal chassis of the radio, or the shielded-not-insulated microphone that’s in your hand, you’re much more likely to feel even small to moderate amounts of current. Get thee a CMCC post haste! :-)
@@SmittyHalibut Thanks so much for your detailed reply, it all makes sense now. I had not thought about the plastic sheath becoming a capacitor but of course it will. I have a Nano VNA and will use it to make some chokes. Thanks again for your help! 😄😄👍👍👍
I'm new to the hobby . . I have made a choke balun which i want to use near my HF radio, It is made of 240 mix 43 toroid and RG-58, my question is: is the distance between the balun and HF rig critical ? jumper i made will give me 8 inches distance b/w choke and radio
If the foil and braid are touching, then no, consider that a single shield. If there’s a layer of insulation between the inner and outer shield, then you have a special coax that’s designed specifically for keeping differential mode and common mode currents separate. I’m not an expert on these, but I think the goal is to ground the outer shield to the chassis at the transmitter and leave it disconnected at the antenna. It’s basically a shield for the feedline. Then the inner shield can be connected directly to the transmitter amplifier ground reference (not the chassis), and the antenna. It’s a more complex transmitter arrangement. If you don’t have that, then just connect both shields to the transmitter, and only the inner shield at the antenna.
Top presentation, graphics and explanation! I am yet to play with my VNA and just got some 31 mix torroids for this exact purpose. I have not seen anything like that test box but would like to build one. I have bookmarked the video for playlist to build. Perfect timing, thank you!👍🤠🇦🇺73 de VK1DON
Funny you should mention wanting to build a CMCC Test Rig. When I recorded this, I hadn’t yet started my own company. I now sell those kits. electronics.halibut.com/
Thanks very much! I was looking up what S11, S12, S21, S22 means and ended up bonking my head with formulas that explain scatter parameters. I know how and why inductors work but that Snn was new to me..easier then i thought..now it makes sense and i can use the values my vna spits out. Thanks alot.
so on your Gordian resonant curve, (22.49) the left of the peak is capacitive and the right of the peak is inductive? Is the peak then resistive? Thanks
...I think so? I'll be honest, I didn't measure the complex impedance across the frequency range, but that would make sense. The peak is caused by resonance, and resonance is when the reactance of the L and C balance each other out and you're left with nothing but R. So, I'm going to say "Yes!" 🙂
Balanced means the voltages go up and down relative to ground on BOTH wires, with the same signal but inverse of each other. The average voltage is always 0, always ground. This is what you have on twin-lead feedlines like ladder line, or XLR mic cables. Unbalanced means one wire is ground, a fixed voltage, and the other wire goes up and down relative to the ground reference. This is what you have with coax like UHF/BNC/N connectors, or RCA audio cables. There are some special cases that I’ll ignore here (like “pseudo balanced,” and phono (as in record players) signals which can be treated as balanced, despite the use of RCA connectors.)
Balanced and unbalanced has to do with the voltages. This is not to be confused with differential mode and common mode, which has to do with currents. Differential mode currents are when the current flowing in each wire of your transmission line are exactly balanced. The same amount of current flowing in both directions. The sum of the currents is zero. Common mode current is any difference between the currents in the transmission line. If the sum of currents is NOT zero, the bit that is not balanced between the two wires is the common mode. For example: if there’s a transmission line, two wires, whether balanced or unbalanced, and one wire has 2A flowing one direction and the other wire has 3A flowing the other direction, then there is 2A of differential mode current and 1A of common mode current.
@@SmittyHalibut so you're saying that the voltages go up and down relative to ground in both wires. So if I power my transceiver with a center tap transformer so that the ground is between the pl 259 shield and the other side, the shield and the center conductor will be equal voltages 180° apart and therefore balanced? How does that change the operation of my transmission system in the long run? What does it matter where my reference ground is electrically?
@@BusDriverRFI Don't do anything to the AC power input on your radio. That will cause all kinds of damage. A given transmission line is either balanced or unbalanced. Coax is inherently unbalanced, twin-lead is inherently balanced. If you want to convert from one to the other, you use a (real) Balun (not the things a lot of people claim are baluns.) There are a few reasons one would choose a balanced transmission line over unbalanced. Balanced transmission lines have less loss than coax; for VERY high powered systems (100kW and more), coax becomes large copper pipe and becomes incredibly expensive, where balanced transmission lines can still just be large gauge wire. For hams, a well deployed balanced signal can also be better at eliminating common mode noise than coax. This can be useful on receive. The downsides to a balanced signal is that its fields extend past the boundary of the cable itself, whereas coax is entirely contained inside the cable. This means you have to be very careful how you deploy balanced cable: If you put anything close to the cable that interacts with the E and M fields (eg: running it close to a metallic tower), it will impact the signal. So you have to space it away from near by stuff, including other balanced transmission lines. You don't have to do this with coax: You can tie it immediately to the tower, next to other coax runs, power cables, whatever. It doesn't matter because all the RF is contained entirely inside the shield. Weighing these pros and cons, most hams are running low enough power and good enough coax that the disadvantages of balanced feed lines outweigh the advantages. So most of us use coax. But you see balanced feed lines a lot more in much larger installations.
@@SmittyHalibut seems strange that a 1:1 could cause all kinds of damage. I wasn't going to do that anyway because I don't want high voltage on my radio case. It was a hypothetical question and not a practical question as to determining whether the "unbalanced" line would be considered "balanced" if it was fed with the "balanced voltage" input. Seems logical that since we are talking about the voltage being balanced, the thing keeping coax from being balanced is the input biasing.
Great work Mark! Now all you have to do to improve your test box is add a SPDT (3 position, center open) switch to the Port 0 wiring from the NanoVNA. Wire the "pole" pin as a short to ground on one "throw" of the switch and the same "pole" pin thru a precision 50 ohm resistor to ground on the other "throw" of the switch. Then add a SPST switch to bridge the center pins on the 2 output connectors for thru or "bypass" mode. VOILA,! You've now got a built-in calibrator too. Total of 3 switches: DPDT, SPDT (3-way), SPST.
That's not a bad idea. THe concern is, it doesn't take into account the connectors on the test side of the adapter. But, given all the other compromises we're doing, it's probably not really a lot worse. And, it still doesn't stop you from using your own SOL standards. I've put a ticket in my backlog to add that to the next revision of electronics.halibut.com/product/common-mode-current-choke-test-rig/
Neat work, you do need to have an understanding of the basic physics, but really, what better way than showing by practicality what effect any changes would have on your desired output. I have nothing to do with ham radios but even i understood the whole concept as presented. Well done.
Thank you for the kind words. 🙂 And, if you ever decide you want to get your ham radio license, RUclips (and the Internet as a whole) has a whole lot of resources to make that happen. Just let me/us know, and we'll point you in the right direction. Cheers!
121 turns! That seems excessive. Depending on what coax that is, you're probably just seeing the loss of so much added coax in your feedline. If that was supposed to be 12 turns, that's much more reasonable and at that point, I suspect something is wrong with the choke. A properly functioning choke should have zero impact on the normal reception of your antenna system, except the added loss of the additional feed line in the choke itself. But that should be around 1dB for moderate coax, and less than that for good coax. It'll be WAY less than 1 to 2 S units (an S unit is 6dB, so that's 6 to 12dB!)
Of course 12 threads. RG58 on FT240-43 toroid. Attenuation according to VNA 25-35db. There's nothing wrong with that. On ft8 lower report. When SSB calling kiwisdr, I can hear myself much more faintly. I don't understand why that is. Where can I go wrong? @@SmittyHalibut
@@miranovak8098 attenuation, using my test rig to measure the common mode of the choke? Or just connected directly to the VNA? If you’re using the adapter described in the video (that I sell now, electronics.halibut.com) then 25dB common mode loss is good, but that’s not what you need to measure for reception loss. Connect your choke directly to the VNA, no adapter. And measure the S21 LOGMAG on that. That should be 0dB loss, or very close to it. If it’s higher than 0dB, something is very wrong with the choke.
It might be, depending on the cable. Thats why I use RG316, which has a very small minimum bend radius, for winding on FT140 cores. Larger coax on larger cores, you definitely should not abuse the minimum bend radius. It’s possible that’s what’s happening here on @miranovak8098 ‘s choke.
An excellent presentation, I have been experimenting with common mode chokes for a few years. Further down the comments you will see the observations I have made. Greater than 25dB is easily possible from 160m to 10m, I will retest mine on 6m but I suspect it will be fine there too. Spoiler alert, I no longer believe that winding them with coax is a good wideband solution...
@@geekthesteve6215 I couldn't find them either! I have done some experiments using twisted pairs pulled out of CAT5/6 cable, each pair has a characteristic impedance of nominally 100 ohms so two pairs fed in parallel retains a 50 ohm characteristic impedance. I have made many chokes wound with coaxial cable and always found the bandwidth disappointing. I have pictures and plots of this solution, nearly 40dB of attenuation at 7MHz...
@@geekthesteve6215 This was my previous comment: Hi, I found similar, that interwinding capacitance appeared to be an issue with too many turns of coax on the core. Controversially perhaps my best wideband common mode chokes are wound with twisted pairs from CAT6 cable. Each twisted pair has a characteristic impedance of nominally 100ohms so effectively I have two in parallel to get to 50 ohms. I am getting much better bandwidth and very low insertion SWR.
@@SmittyHalibut I believe that a major factor is the reduced capacitance between turns. I made a variety of common mode chokes wound with coax, but I found that putting too many turns on the cores caused the high frequency performance to deteriorate. The braid of even the small coaxes has a lot of cross section so adjacent closely spaced turns will couple significantly. The CAT5/6 conductors have a very small cross section, so should result in reduced capacitance between turns. Certainly the insertion SWR is surprisingly low and the common mode attenuation has a very broad characteristic. I have not tested my first example at beyond 100w carrier...so far...
Hey Smitty thanks for your efforts . Have you done any efficiency tests ? I’ve made a few baluns the latest being a 49 to 1 on a big ft243 core . It measured really good on the nano vna however it got quite warm using 100W and HF which lead me to question its efficiency ? 73 Chuck VK2SS
I haven't. That's a very different thing than what I'm trying to measure here. What you're measuring is an actual differential mode impedance transformer, not common mode choking. (This is why I hate the term "Balun" because it is commonly used to describe so many different things.) I'm not sure I understand the magnetic fields in an impedance transforming un-un (the more accurate name of a 49:1 for an end-fed) to speculate on what's happening. But to measure the performance of the 49:1, you'd need a non-inductive resistor that's (49*50ohms=) 2450ohms, and capable of dissipating the full 100W of your transmitter. You'd put that on the antenna side of the un-un, and transmit into it. Then measure the heat in the TRANSFORMER, not the dummy load. You want to see all the heat being dissipated by the load, not the transformer. If your transformer is cool and the dummy load is hot, then your transformer is good, and I'd look at the antenna for faults. (This is the part where I couldn't tell you what to look for.) If, while transmitting into the dummy load, your transformer still gets hot, then look in the transformer for faults: Broken wires, cold solder joints, a-symmetries in the windings, etc. I hope that helps. The 100W 2450ohm non-inductive load is likely to be a challenge to find. I don't have any pointers on that either. Sorry. You'll probably end up having to make it yourself.
I'm building a similar test jig... But have a question Can you use a single 6 position On-On toggle switch instead of 2 separate 3position switches? I have a 6 pos in the junc box but I'm worried the cables from the two channels on the VNA would be too close together when soldered on the 6pos switches terminals.
*ahem* electronics.halibut.com/product/common-mode-current-choke-test-rig/ ;-) You can, but the risk is exactly what you mentioned: You want to keep the two sides as separate as possible. They will electrically couple, throwing off your readings.
@@halibutelectronics Awesome thanks so much for the help.. As for the link... I am building it myself ... "to save money" ... Of course I'm already at least $60 deep into it ... and now I have to wait on some switches .. BEST HOBBY EVER!!! ;-)
@@halibutelectronics Of course! ... I think its actually just a matter of reminding myself that the value of DIY'ing isn't to "save money" b/c it rarely works out that way... and thats OK... That being said sometimes it is smarter to just buy it (especially tools) ...
@@grumpycat_1 sometimes you want a project, and sometimes you just want the result. There’s no shame in sometimes wanting the result without the project, nor is there shame in wanting the project even though the result is cheaper/easier/quicker/whatever-er. Just, ya know, be aware of which you’re doing and why. :-)
Finally. A video that does a good job of explaining what’s going on with baluns. Question: did you try placing two DIFFERENT types of baluns in series? For example, two different baluns with complementary responses? I hope you plan to do more videos. You do a great job and I’ve subscribed to your channel, so you might as well :)
I haven't! That's a really good idea. I think I spoke briefly about different mixes of cores, using a 31 and a 43 to combine their response curves, but I did not talk about different builds of chokes. As for more videos: I do! I'm neck deep in trying to get SOAR into production right now, so I'm short on time for doing videos. Having said that, I am working on another QSO Today presentation for the September 2022 Expo. It's on a totally different topic than Common Mode Current Chokes, but I hope it'll be interesting too. :-)
You've probably figured this out by now, but when you keep saying you are making the circuit so that it is less resonant in that one place on the spectrum so much, what you are actually doing is lowering the "Q". Low "Q" equals wide bandwidth without the "performance" at any one frequency than you had before you started lowering said "Q". It's the same principal as the tank circuit in your vacuum tube linear amplifier. We shoot for about 12 "Q" because there we get a pretty wide bandwidth in which to work without making it so that when we go to tune it up on our frequency of choice within a band, we don't have so high a "Q" that if we breathe on the tuning capacitor we're not off into another band. I once built an amp with a "Q" of 40 or so and the tuning was so sharp that as the tubes warmed the amp up it would go out of resonance and I'd end up arcing everywhere because the plate circuit was so far out of resonance. A life lesson learned... Same thing applies here, though all you see is no help from the choke you designed. It's all about spreading that RF around nice and evenly and giving up performance where performance is not needed, as is in this case. It's all about trade-offs. Trade performance for function. 73 AA5IT
So the swr meter in my cb radio shows all is good, no problems. So I went and bought me a external meter only to double check. I connected it, and what do I see... the meter says no good, swr way up there. Took a look at my radios meter and now it’s also showing a high swr all of a sudden. So I start to adjust the antenna a few times, checked if it did help and yes it did. My external meter shows a low swr. The radios swr agrees also, wow this is great! All I did was to plug the meter out and back with the antenna in the radio. But no, now after I disconnected and put the meter away, the radio says the swr is high! What the hell is going on? And how can I learn to fix this because I don’t have money now to go out and buy another meter. And since this is happening on my radio, I wouldn’t be able to make sure an older radio that I planned to buy would be sure. I could not even help if someone where to ask me to come with my swr meter and make sure there’s are not too high. What can I do to make the radios swr meter work together and show same good results like the did, until I discontent the external one. Radio is showing very high again.
I am not an engineer but have been measuring antennas SWR for about 50 years and my guess is your coax length without the SWR meter and patch is resonant at 11 meters. Probably a multiple of 108 inches (a quarter wave at 11 meters or 27.255 mHz) times the velocity factor of the coax and an example would be .7 (velocity factor of your coax may be slightly different) times 108 inches, or a multiple of 75.6 inches, or 151.2 inches or 226.8 inches, etc.. Try leaving your patch cord in the line by replacing your SWR meter with a double female barrel connector and then check your SWR at the radio.
@@geekthesteve6215 Thank you, that sounds reasonable. Would a shorter coax cable from the swr meter to the radio help get reliable readings you think? It would be great if I could trust the swr meter if I ever should get a radio that has no built in swr meter or high swr protection.
Maybe a dry soldered joint in the new meter? Have you got r.f. chokes? I was given a supposedly non working MFJ160-10 tuner by a friend but I used it for two years, we went to Wales and still good but connecting it at home it no longer worked. Looking inside revealed nothing, it all looked perfect until I poked around and found a loose wire. It had been touching but not soldered. G4GHB.
@@bill-2018 Ohh this could be, because I bought it on amazon like budget version. The package of the cables says “uhf” so I was thinking it might be the reason, but then again a cable is just a cable, so I’m not sure 🤔
Thanks a lot for this video, now it is clearer for me, except that you speak fast and for a non-english-speaker it's not easy to understand but i've seen it two times and it's ok now !😁
Well… shoot… I used to be concerned that my HF feedline was too long, & the extra coiled up behind the house was picking up noise. NOW I know what I’m going to do with a LOT of that extra length. I might wind up (intended) putting 3 or 4 or 5 of these in series, & see if the 40m hash from the washing machine can be quashed.
Dam it mark.....strong words followed by other strong words some bleeps and then some less strong words.....now I need to rebuild my main choke as I'm useing 4 ft140 31 and 4 ft 140 43 toroids in a kind of binocular shape with 6 turns ish of my coax then directly into the radio I measured about 25 ish db down across hf and now I'm wondering if I split it in 2 and do 2 of the 31 mix and 2 of the 43 with 3 turns then repeat again so I have the 3 turns theory satisfied and the chokes In series theory also satisfied... dam you ye bleepus maximus
🤣🤣🤣 I’ll point out that you can MEASURE these things using the kit I sell now if you’d like, so you know for sure which configuration is better. I’d also request that you take notes, take pictures of the graphs, and at least write up your results, or do a video showing your results. Somehow, share it with the world. I will absolutely link that here and boost it. I haven’t done anything with binocular chokes, or mixing .. mixes.
You need to relearn your theory. An inductor does NOT stop the flow of AC or RF current no matter what inductive reactance present. What stops and attenuates common mode is resistive parallel resistance that dissipates the common mode. An inductor does not consume power it just changes it's phase. Cores create this resistance within the core characteristics.
5:28 ... "Necessarily balanced, becuzz physics" Dude! It's the TEM wave. Not deep physics. Though, of course, real coax isn't perfect so there's a tiny CM.
Inductances don't provide a load for energy. Only resistance can load energy. Your 2 core balun does not provide better rejection because it increases inductance since inductance can't dissipate energy. You did not actually measure any currents any where. You would make a very poor example of an engineer. You can't just say something is true. You have to show it. And if inductances can't dissipate power, where is the resistance at?????
Does it make you feel better about yourself to tear into others like that? I'm seriously asking, because I don't understand why anyone would word that the way you did. You can challenge the information I present without making it a personal attack. I welcome discussions about the content presented, and if you can convince me I got something wrong, I'll happily own up to it. But when you start off with "You're a very poor engineer," I'm less inclined to even listen to the potentially valid point you're bringing up. (See below.) And this completely ignores the point that good engineers aren't always right. Good engineers can make mistakes and learn from them. If I am wrong here, that doesn't make me a "poor example of an engineer." It makes me human. To address the technical point you're trying to make: We aren't trying to dissipate the power, we're trying to block it and prevent it from flowing in the first place. The chokes provide a high impedance to the common mode currents, which actually reduces the amount of current flowing. When you compare the "about 1000 ohms" of inductive impedance in the choke against the "about 50 ohms" of impedance in the antenna or feed line, the current will predominantly use the lower 50 ohm impedance. The higher the impedance (whether resistive or reactive) of the choke, the better this effect. "You didn't measure any currents anywhere." I did not measure any currents DIRECTLY, that is true. But I did measure the signal strength loss by converting the differential signal out of the VNA into a common mode signal, passing it through the choke as a common mode current, then converting it back to differential mode so the VNA can measure the signal. This is analogous to what's ACTUALLY happening in real life. Any attentuation, whether from resistive effects, reactive effects of the chokes we're building, interactions with the environment, or anything else, will show up as S21 loss (negative "gain" values as the NanoVNA calls them). By keeping the other loss effects relatively constant and only changing the choke, we can measure the effectiveness of those changes on the choke's ability to prevent common mode signals. Is that measuring current? Fine. Not directly. But I am directly measuring what we actually care about, so I'm ok with that.
Early radio engineers said wavelengths below 200 metres were of little use and wavelengths above were required to make long distance communication so radio amateurs were given everything below 200m. Only when amateurs started making long distance contacts were they interested. The 'professionals' got it wrong! G4GHB.
FANTASTIC tutorial. 30+ year BSEE and Electromagnetic major here. You know your subject and can present it well. Thank you for all the work put into this.
Thank you! I appreciate the kind words. :-)
These are the answers about common mode that I have been looking for , for a long time. great lesson, thank you
I'm glad it helped. 🙂
Excellent presentation-accessible to the newcomer but with enough information to whet an Elmer's appetite. It spurred me to buy the Halibut Electronics kit, and now I need to build ALL THE CURRENT CHOKES!
First of all, nice video, one of the best I've seen on building a CM choke. Only one minor comment: "Building something and trying it" is NOT really an "engineering approach". It is actually an "experimental" approach. As an engineer for 43 years, I found that taking an "experimental approach" ended up taking up a lot more time in the "development phase" of a product than working the design out mathematically before laying any lines down on a drawing. Having said that, I frequently approach my amateur radio projects as experimental in nature simply because I like building a lot more than I like math. In amateur radio I am not bound by schedules, budgets, difficult test requirements, etc, so I can enjoy the build to "see what I end up with". So, this is not meant to be a negative comment about your video. I found your explanations were clear and accurate with nice visual aids, and more importantly easier to understand than wading through the math. Good overview on setting up and using a VNA, too.
--Mark, KE6BB 👍
That’s a fair point. I think what I was trying to say was “practical” vs “theoretical.” To me, that’s the difference between engineering and science. Engineering is more applied, science is more conceptual.
But you’re right, “experimental” is an even better term for what I’m doing here.
Thanks for the feedback! :-)
16:00 - outer conductors of the two NanoVNA ports (at least on the old 1.5G one, the one made out of three SA612A mixers) are shorted through the device's common ground, just FYI
True. That would make the plane of testing right at the edge of the VNA, including the cables between the VNA and the test rig, which is ... less than ideal... 🙂
Nice presentation and it makes sense to test any balun before use to actually know what you are getting. Thank you and now I need to test.
I know I'm kinda late (apparently not getting notifications for many of the comments on this video!), but I will point you to electronics.halibut.com/product/common-mode-current-choke-test-rig/ 🙂
far better explanation than any other ham video I have seen yet!!
Thank you. 🙂
A video about your Common mode test board would be helpful for the folks that purchased one
Excellent explanation and test work Mark. As another RF guy I am troubled by 2 things. One at 09:20, how can CM noise get inside a shielded radio when applied between coax shield and transceiver "ground"? I suspect all CM noise entry is either via poor coax or transceiver shielding. Two at 22:03 is the assumption that a CM choke should be X times the cable Zo. I doubt that the surge impedance on the outside of the coax is also 50 Ohms. Anybody wiser?
1: you’re probably right that RF gets in on bad coax too. Heck, it’ll get in on the antenna itself! But any non-perfect ground will also put that noise into the receiver.
2: The goal is to make the common mode impedance as high as possible at the target frequency. Common wisdom as proposed by people smarter than me is “about 1000 ohms.” That’s about 20x a 50 ohm system … and I think I’m realizing I did my dB math wrong in the video. That’s 20x impedance (linear with power), not 20x voltage (square with power). So it’s only 13dB, not 26dB. If I’m right, all the numbers I gave in the video are twice as good as they “need” to be. Huh. Well, we have some mighty fine CMCCs out there I guess.
@@SmittyHalibut Yup I agree with the math. Just the relation to 50 Ohms which is the internal cable Z not the outside "third wire". No matter, we need Z as high as possible, in most cases with a good Rloss to gobble up energy. As far as RX CM I think it is an "all cables" sneak in (line cord, USB, LAN...). CM chokes DO help. But as for TX induced RF chokes should be at a voltage maxima on the line where ever that it is. PS It is amazing how far the ham community has come in understanding lines, VNAs etc. Cheers!
Very good presentation and treatise of one of the most prevalent ham station basic problems....
Thank you. :-)
Dude, you need to make more videos ! Good Job !
:-) Thank you.
Forgive my late input . I note in a few of the comments that many would love to see a comparison against a Coax bottle type Choke and what the attenuation figures come out like . . Great video by the way . Well done 73. Ron G4DIY
Thanks for the positive feedback! I'm not sure what design you mean by "Coax bottle type choke". A quick google search turned up a design where you wrap coax around a pill bottle? Is that what you mean? If so, that's just an air-core choke. The bottle is only a form to wrap the coax around, but doesn't appreciably impact the magnetic permeability of the choke.
If you mean some other design, please point me to a page or article describing what you mean.
@@SmittyHalibut hi there again. Well my “bottle type “ is literally a plastic bottle about 8 inches high and about 4 inch diameter . I have coiled Rg 213 Coax around it with about 20 turns . So its a coaxial choke effectively.
@@ronbennett5591 Ok, yeah. So it's an air-core choke. I've gotten feedback elsewhere in this comments section that I never gave a properly-wound air-core choke a fair shake, and that's true, really. I need to make an update to this video with my CMCC Test Rig kit instead of the hand-built test rig in this video; when I do that, I'll give an air-wound choke a proper test.
Great talk. Thinking about getting my ham license, but I think I’d be in the “mostly building, not operating” group as well. What are POTA and SOTA?
“OTA” usually means “On The Air” so it’s just a matter of the first letter(s). POTA is Parks, and SOTA is Summits. Both their websites can be found by searching. (Read: I don’t remember the URLs off hand. :-)
I too have done some research and testing to determine the optimum ferrite materials and number of turns to get the most out of my RFI choke filters. I based my conclusions on ferrite manufacturers equations and my personal tests using signal generator, oscilloscope and nanoVNA to sweep the entire HF band (e.g. building a combined differential and common mode filter for the output of a home-brew DC switching power supply for my shack). I too found that, for a snap-on ferrite, 0.5 inch hole, type 31, 3 turns was optimal before capacitive coupling between turns reduced effectiveness. But for an FT-240-31 I think 5 to 7 was optimal, according to manufacturer's equations and testing (this was 3 years ago so I don't quite recall exactly).
However, I also built a 1:1 current balun for use as a wire dipole feed point and I put 11 turns (16 AWG PTFE insulated wire, split and wound in two direction) on that one, and it tested extremely well, -35 dB attenuation across most of the HF band. So it seems the geometry of the toroid makes a big difference in terms of how many turns is optimal. And your tests also showed better braud-band attenuation with the 13 turns around a FT-140. Do you have a general explanation for this? Why does the capacitive coupling add up so much with the snap-ons?
Its more about the turns of cable than the ferrite material. When on a torroid, the turns are spread out, more space between windings. When wrapped in a "Gordian knot", they're immediately adjacent to each other, maximizing the capacitance between turns.
Alright, you just put all the pieces of the puzzle together. Fantastic work and presentation. Following.. Thank you so much!
I’m glad I could help. :-)
I took issue with the description of a coax as a differential cable early on in the talk (around 4 minutes). Signal Integrity engineers refer to them as single ended for a reason or two - a differential cable, for example requires at least 4 vna ports set into 2 logical port groups; and the shield is not the return conductor, it is a shield. It serves the same purpose as the shield over a twisted pair like in an ethernet cable, which is a true differential system. One more comment is that shields of coax are almost always tied to the chassis ground on both sides, not the signaling elements of a transceiver so there is no push-pull of currents. Any differential currents created are simply inductive coupling between the center conductor and the shield.
That all said, I love the data collected on the experimental chokes the nanoVNA plots. I wish we could have seen data for 2 chokes placed on a single gordian knot to see how adding chokes improved the common mode rejection also.
I just re-watched that section to make sure I was thinking about the right thing.
You are absolutely correct that there is a world of difference between a Single Ended signal (like on coax) and a Balanced signal (like on twin lead, or twisted pair). But that's more about voltages, not currents.. (Yes, voltages which then drive currents though impedances, yes, so it's not entirely NOT about current... See below...)
But even a Single Ended signal will have opposite and equal current flows on the coax. My mistake was referring to those currents as "balanced", little b, the adjective. They are not Balanced, big B, the proper noun describing the signal type. That's why it's called Differential Mode for currents, because that's different than Balanced for voltages.
The push-pull of currents part, again, you are correct. That's the impact of a Balanced (big B) signal on the currents that we talked about earlier. But that still doesn't mean that there isn't an equal and opposite current flow on the feed line. If there's no other return path for current, it HAS go to on the feedline. And even if there is another return path for current, like you said, it will induce the return current in the coax and probably prefer that path over the other (assuming your coax is well connected and low impedance compared to the alternate path.)
So, you're not wrong in any of your response. But I think those points don't really impact the description of the examples presented.
Thank you. Very nicely made and easy to understand.
Great video on checking CM chokes. Thanks for doing all the work determining that 13T on a 31 Mix core is the best, fits right in with G3TXQ work that he did on chokes, baluns and ununs.
He actually said his favorite was 13 turns on a 43 mix 240 toroid not 31 mix.
Thanks for the video. As I was watching I was wandering about influences of possible phase shifts in the device. The influence on the working of the common mode choke. Signal coming in the choke (on one side) and signal comming in (on the opposite side and from the opposite direction) may have some phase shift. The load is probably not pure resistive....
Are you talking about the Differential Mode response, meaning how transparent it looks to the signals flowing through it? Or the Common Mode response, meaning how the phase affects the resistive/reactive balance of the impedance it imposes?
If the former, the Differential Mode view, by definition Differential Mode currents sum to zero through the inductor, so there is no impact at all. When I=0, it doesn't matter what the j component of X is because it will be multiplied by 0 to get your V vector. "What if phase differences result in I!=0?" Then those differences would become Common Mode. By definition.
If the latter, the Common Mode view, then yes you're absolutely right. The impedance imposed by a Common Mode Current Choke is not purely resistive across the whole band of interest. It's an RLC complex circuit. An IDEAL choke would be all R and very little LC over the whole band of interest, but that's nearly impossible.
In the video I show highly resonant chokes with a very deep attenuation "peak" (shown as a "valley" on the graph because it's a negative number). In those cases, if you measured the Complex Impedance of the CMCC, you would see a high R/low X at that peak, because that's the resonant point of the choke, and resonance at the frequency happens when f*L and f*C are equal but opposite, which cancels them out and leaves you with nothing but R. But outside that peak, the response becomes very reactive.
Awesome stuff!
When you talked about calculating or experimenting with Gordian Knots to make a choke for a specific band, I had to comment because I've done exactly that.
I fly paramotors, wanted to build an antenna for GMRS (462 MHz) out of a chunk of RG316 with half of the coax sheath removed, then string it up one of the glider lines.
2 problems:
1) "Center/core fed vertical" doesn't resonate well on any frequency. Experimenting showed that the angle between the two legs needs to be around 30° (aIrc) for resonance, and
2) Even if I got it resonant, Instill needed to decouple the antenna from the feed line, otherwise the radio body, the hand holding it, any metal or human near the coax would cause wild swings in the SWR.
So I added a Gordian Knot, and experimented until I reach both resonance and decoupled the antenna from the feed line.
In my case there was 2 loops at around 6cm diameter. Location of the knot along the coax was also critical.
Couple of us flew those antennas for a while and they worked well but wore out quickly.
Next version will be with more flexible coax, and the knot up higher on the line so it won't get beat up as badly during handling.
Also trying to figure out how to do multibands ...lol
That's awesome! (Sorry for the late reply; not sure why I didn't get a notification for this one.) Gordian knots can work well for mono-band operations, if you tune them. And a rig like described here can help you do that. But they are inherently mono-band.
built one as well, as per your istructions (the version with the switch). works as a charm! maaany thanks!!!
Yay! I'm glad to hear it! 🙂
What would happen if one were to place the common mode choke in an iron or aluminum tube? For instance, if someone were to insert a linear common mode choke into the support mast itself to protect the coaxial cable from weather elements like UV rays? What are the potential impacts, and have you ever measured them?
Maybe I missed this but why multiple common mode chokes are recommended (at feed point, at trx, in middle) and why single one at transceiver isn't enough for eliminating "bad things" entering trx?
Great presentation! Is there a way that a NANO VNA can output it's choke results in the magnitude of "Z" rather than dB?
Great presentation! I am just wondering , when it is about a current inducted in the shield , then probably this conductor is not anymore a feed line with 50 ohms characteristic impedance ?
Very true. At that point, it's a long wire antenna with some unknown impedance.
Loved the engineering approach Mark. I am playing with common mode chokes now and so I am going to build a test box, everything you say about it makes sense. I was hoping to see more detail on ferrite core testing, going from 1 turn to 20 turns or so. At some point putting more turns on the ferrite the interwinding capacitance increases and so more turns is no help, just the same reason that more turns on a knot is no help. It would have been good to see how you got to 13 turns being the optimum for the size core and size coax you have. I am sure that a different size and type of core and different size coax has different optimum turns. The NanoVNA has opened up so much testing to low budge engineers - and thank you for pointing out the calibration issues. This is gold
Hi, I found similar, that interwinding capacitance appeared to be an issue with too many turns of coax on the core. Controversially perhaps my best wideband common mode chokes are wound with twisted pairs from CAT6 cable. Each twisted pair has a characteristic impedance of nominally 100ohms so effectively I have two in parallel to get to 50 ohms. I am getting much better bandwidth and very low insertion SWR.
Hi....Am I right that you still have a "jumper" between the grounds of the sma connectors on the VNA side, just not visible in the picture?
I have in mind to build this one and a frankenstein version that uses the inner and outer across a chock to see what the sweeps compare like 🙂....then of course measure the CMC on coax installations before and after chokes...then get down some operating maybe....thanks for a great presentation
Soooo... Good catch. No, on the second version of the box, I didn't have that jumper between the shields on the two VNA ports. I should have.
I encourage you to make your own CMCC Test Rig if you want to. I will provide as a "Throw money at the problem" option if you'd prefer, I do now offer those test rigs as a kit from my company's website. I started Halibut Electronics _AFTER_ recording this video, so it didn't come up. If you're interested, it's available here: electronics.halibut.com/product/common-mode-current-choke-test-rig/
@@SmittyHalibut Thanks Mark...the kit looks like decent value given the connector costs alone, I would certainly purchase one if I was in the US, but tax and duty on top make it a little less attractive for me, that said, I will make up the V2 with a jumper and try to persuade my club we need a proper kit ;-) I do have a question....Most of the test arrangements I have seen connect the output of the VNA (centre conductor) to both inner and outer conductors on the choke to generate CMC....It is not clear to me how feeding just the inner (from VNA) to shield (on choke) creates CMC....the sweeps look the same. What am I missing here?
GM0EDJ
@@paultemple5566 good question! Remember from the video that current flows in interesting ways in coax. Differential mode currents, where there is an equal amount of return current, necessarily flows inside the coax between the center conductor and the inside of the shield. And common mode current, that does NOT have a return current, necessarily travels on the outside of the shield.
Because: Physics.
So the important part isn’t which conductor of the Device Under Test (DUT) the signal is put on, it’s more important that the return current is NOT sent through the DUT. If you use coax, even the center conductor, as a single wire without sending the return current through the coax, that single wire’s current will NECESSARILY travel on the outside of the coax.
It’s weird, I know. And I couldn’t explain to you now why that is, but it was explained to me back in college *mumble* decades ago, and it made sense then, so I’ve accepted it on faith ever since.
So, in my screw up of the v2 test rig, without jumpering the shields together, it just means the return currents are INSIDE the VNA. Which means the two jumpers connecting the VNA and the test rig are also common mode. It will add noise to the measurements, but it doesn’t invalidate them.
Does that make sense?
@@SmittyHalibut Thanks Mark....good enough for me and makes sense, hence the jumper to avoid a return current through the DUT....on with build and testing
@@paultemple5566 Be sure to let me know how it goes! 🙂
I haven't seen the entire video so I apologize but thought I should say that TRX (a channel) has a video where he builds and measures the effectiveness of a common mode choke with a NanoVNA. I will come back when and if I see a measurement like he did. If not, go and see what he shows us.
I'm back. I like what I'm seeing very much. Its much more elegant to make a test device box. I think though, you were both doing the same test. That said, what did you think of his initial test of toroid material ?
That said, I'm only seeing you show poorly wound coax baluns, and nothing properly wound. So far anyway. I'm going to wind a tested design balun with a 43 mix and good enameled copper wire 1 to 1 and see if it does what others have measured. A coax balun is not the way to go. But, I do like your testing set up, and have both a regular nanoVNA and a V2.
@@californiakayaker It depends on what problem you're trying to solve. If you have a perfect dipole antenna with identical shape and size no uneven obstructions in the near field a wire wound balun is better, I think we can all agree on that.
A wire wound balun won't solve my problem in my mobile between 160m ~ 40m where I am attempting to capacitively couple with the earth. I need coax balums right behind the unun or it all goes to Hell real fast. You get the antenna perfectly tuned then start applying about 500 watts and it all falls right on its face, as the amplifier kicks out because it sees 60 watts coming back. 🤦 G3TXQ and K0BG discuss this on their websites.
Excellent, excellent video…thanks for posting! 👍
I tried to make a 1:1 today but my toroid seem to be smaller than all those yt examples.
How did the 2 clamp on with one coil work. It would use less coax and may give more blocking. Or I am thinking it would. 4 clamp ons should, in my mind do good on one coil. But I am not sure if I am thinking the right way.
It would almost certainly help, yes. The question is, which "costs more?" In dollars, ferrite is more expensive than cable and only increases the inductance linearly with the number of clamp-ons, where as adding turns of cable increases the inductance to the SQUARE of the number of turns.
But if your source of cable is finite (eg: it's already in place and only so long), and you can afford to put more ferrites on it, then yes that sounds like the better option.
To answer your direct question: I don't have data for what the two-clamps-on-one-loop looks like. I'll take this opportunity to point you to electronics.halibut.com/product/common-mode-current-choke-test-rig/ and ask you to post the results, though! 🙂
Jim Brown (K9YC) had some useful graphs in A Ham's Guide to RFI and also in Understanding How Ferrites Can Prevent and Eliminate RF Interference. You should be able to search for those titles and his name.
One of the best explained presentation on baluns I found.
A question just came up in my mind: If you use two ferrites in series, does the separation between the two matter? IOW, does the flux from each one couple with that of the other one? I think I'll start saving up for a VNA.
Also, I'm curious about one other thing *(NOT* related to coax). If I buy a pre-made common mode choke, can I connect one of the sets of windings backwards and turn it into a *DIFFERENTIAL* mode choke? How would that compare with using two *separate* inductors (wound on *different* cores) for that purpose?
_73 de AF6AS_
That's a GREAT video explaining Common Mode Chokes - One thing, you never talk about power handling ability since these are usually used with TX as well as RX situations. The air core coax wound chokes would be limited only by power handling of coax, or close depending on how much capacitance affected power handling.
The ferrite cores would be my concern - I think even the clipon ferrites would handle 100 watts, but when 1000 watts is involved it would require more?
Well done mark. I enjoyed your presentation very much and look forward to more.in my opinion this is what ham radio is about.experimentation and sharing. Unfortunately you have to suffer the snide comments of others who know better (maybe).I hope you are not discouraged my freind.
Nah, it's all good. I considered it a teaching experience. Here's to hoping it was also a learning experience for him. Thank you. :-)
Thanks for this video, as a returner to the hobby after nearly 40 yrs, it was very useful to me. Forgive me if this is a dumb question but please would you confirm that when you say "outside" you mean outside edge of the shield and not the outside plastic sheath of the coax? I ask this because I know you can get RF burns from the very outside of the plastic sheath of a coax. Thanks again.
Not a dumb question at all!
The currents are traveling on the shield, that is true. But RF can capacitively couple through the insulation into your skin. If the currents are high enough, then yes you might get burned (more likely a tingle) from touching the coax.
Now, that’s a REALLY BAD common mode current. If that’s happening, something is seriously wrong with the antenna or feed system. Under normal circumstances, with normal levels of reflected common mode current, it’s very unlikely you’ll get an RF burn by touching insulated coax. But once that current hits the metal chassis of the radio, or the shielded-not-insulated microphone that’s in your hand, you’re much more likely to feel even small to moderate amounts of current.
Get thee a CMCC post haste! :-)
@@SmittyHalibut Thanks so much for your detailed reply, it all makes sense now. I had not thought about the plastic sheath becoming a capacitor but of course it will. I have a Nano VNA and will use it to make some chokes. Thanks again for your help! 😄😄👍👍👍
Wow, well taught! Thank you!
I'm new to the hobby . . I have made a choke balun which i want to use near my HF radio, It is made of 240 mix 43 toroid and RG-58, my question is: is the distance between the balun and HF rig critical ? jumper i made will give me 8 inches distance b/w choke and radio
Is Halibut Electronics going to be at Dayton this year?
What if the coax has two shields, one inner foil and an outer braid?? Does that mean 5 currents?? How to wire up an RF choke with toroid ferrite??????
If the foil and braid are touching, then no, consider that a single shield.
If there’s a layer of insulation between the inner and outer shield, then you have a special coax that’s designed specifically for keeping differential mode and common mode currents separate. I’m not an expert on these, but I think the goal is to ground the outer shield to the chassis at the transmitter and leave it disconnected at the antenna. It’s basically a shield for the feedline. Then the inner shield can be connected directly to the transmitter amplifier ground reference (not the chassis), and the antenna.
It’s a more complex transmitter arrangement. If you don’t have that, then just connect both shields to the transmitter, and only the inner shield at the antenna.
Top presentation, graphics and explanation!
I am yet to play with my VNA and just got some 31 mix torroids for this exact purpose.
I have not seen anything like that test box but would like to build one. I have bookmarked the video for playlist to build. Perfect timing, thank you!👍🤠🇦🇺73
de VK1DON
Funny you should mention wanting to build a CMCC Test Rig. When I recorded this, I hadn’t yet started my own company. I now sell those kits. electronics.halibut.com/
Thanks very much! I was looking up what S11, S12, S21, S22 means and ended up bonking my head with formulas that explain scatter parameters. I know how and why inductors work but that Snn was new to me..easier then i thought..now it makes sense and i can use the values my vna spits out. Thanks alot.
I'm glad I could help! 🙂
so on your Gordian resonant curve, (22.49) the left of the peak is capacitive and the right of the peak is inductive? Is the peak then resistive? Thanks
...I think so? I'll be honest, I didn't measure the complex impedance across the frequency range, but that would make sense. The peak is caused by resonance, and resonance is when the reactance of the L and C balance each other out and you're left with nothing but R.
So, I'm going to say "Yes!" 🙂
Electrically speaking, what's the difference between balanced and unbalanced?
Balanced means the voltages go up and down relative to ground on BOTH wires, with the same signal but inverse of each other. The average voltage is always 0, always ground. This is what you have on twin-lead feedlines like ladder line, or XLR mic cables.
Unbalanced means one wire is ground, a fixed voltage, and the other wire goes up and down relative to the ground reference. This is what you have with coax like UHF/BNC/N connectors, or RCA audio cables.
There are some special cases that I’ll ignore here (like “pseudo balanced,” and phono (as in record players) signals which can be treated as balanced, despite the use of RCA connectors.)
Balanced and unbalanced has to do with the voltages. This is not to be confused with differential mode and common mode, which has to do with currents.
Differential mode currents are when the current flowing in each wire of your transmission line are exactly balanced. The same amount of current flowing in both directions. The sum of the currents is zero.
Common mode current is any difference between the currents in the transmission line. If the sum of currents is NOT zero, the bit that is not balanced between the two wires is the common mode.
For example: if there’s a transmission line, two wires, whether balanced or unbalanced, and one wire has 2A flowing one direction and the other wire has 3A flowing the other direction, then there is 2A of differential mode current and 1A of common mode current.
@@SmittyHalibut so you're saying that the voltages go up and down relative to ground in both wires. So if I power my transceiver with a center tap transformer so that the ground is between the pl 259 shield and the other side, the shield and the center conductor will be equal voltages 180° apart and therefore balanced? How does that change the operation of my transmission system in the long run? What does it matter where my reference ground is electrically?
@@BusDriverRFI Don't do anything to the AC power input on your radio. That will cause all kinds of damage.
A given transmission line is either balanced or unbalanced. Coax is inherently unbalanced, twin-lead is inherently balanced. If you want to convert from one to the other, you use a (real) Balun (not the things a lot of people claim are baluns.)
There are a few reasons one would choose a balanced transmission line over unbalanced. Balanced transmission lines have less loss than coax; for VERY high powered systems (100kW and more), coax becomes large copper pipe and becomes incredibly expensive, where balanced transmission lines can still just be large gauge wire. For hams, a well deployed balanced signal can also be better at eliminating common mode noise than coax. This can be useful on receive.
The downsides to a balanced signal is that its fields extend past the boundary of the cable itself, whereas coax is entirely contained inside the cable. This means you have to be very careful how you deploy balanced cable: If you put anything close to the cable that interacts with the E and M fields (eg: running it close to a metallic tower), it will impact the signal. So you have to space it away from near by stuff, including other balanced transmission lines. You don't have to do this with coax: You can tie it immediately to the tower, next to other coax runs, power cables, whatever. It doesn't matter because all the RF is contained entirely inside the shield.
Weighing these pros and cons, most hams are running low enough power and good enough coax that the disadvantages of balanced feed lines outweigh the advantages. So most of us use coax. But you see balanced feed lines a lot more in much larger installations.
@@SmittyHalibut seems strange that a 1:1 could cause all kinds of damage. I wasn't going to do that anyway because I don't want high voltage on my radio case. It was a hypothetical question and not a practical question as to determining whether the "unbalanced" line would be considered "balanced" if it was fed with the "balanced voltage" input. Seems logical that since we are talking about the voltage being balanced, the thing keeping coax from being balanced is the input biasing.
Great work Mark! Now all you have to do to improve your test box is add a SPDT (3 position, center open) switch to the Port 0 wiring from the NanoVNA. Wire the "pole" pin as a short to ground on one "throw" of the switch and the same "pole" pin thru a precision 50 ohm resistor to ground on the other "throw" of the switch. Then add a SPST switch to bridge the center pins on the 2 output connectors for thru or "bypass" mode. VOILA,! You've now got a built-in calibrator too. Total of 3 switches: DPDT, SPDT (3-way), SPST.
That's not a bad idea. THe concern is, it doesn't take into account the connectors on the test side of the adapter. But, given all the other compromises we're doing, it's probably not really a lot worse. And, it still doesn't stop you from using your own SOL standards.
I've put a ticket in my backlog to add that to the next revision of electronics.halibut.com/product/common-mode-current-choke-test-rig/
*psst* Keep an eye on electronics.halibut.com/ .. The next batch of CMCCs will have this feature on them. 🙂
Neat work, you do need to have an understanding of the basic physics, but really, what better way than showing by practicality what effect any changes would have on your desired output. I have nothing to do with ham radios but even i understood the whole concept as presented. Well done.
Thank you for the kind words. 🙂 And, if you ever decide you want to get your ham radio license, RUclips (and the Internet as a whole) has a whole lot of resources to make that happen. Just let me/us know, and we'll point you in the right direction. Cheers!
Hi, I find that a 1:1 choke will degrade my reception by 1-2 S. Is this ok? I have a choke on a toroid with 121 turns of coax
121 turns! That seems excessive. Depending on what coax that is, you're probably just seeing the loss of so much added coax in your feedline. If that was supposed to be 12 turns, that's much more reasonable and at that point, I suspect something is wrong with the choke. A properly functioning choke should have zero impact on the normal reception of your antenna system, except the added loss of the additional feed line in the choke itself. But that should be around 1dB for moderate coax, and less than that for good coax. It'll be WAY less than 1 to 2 S units (an S unit is 6dB, so that's 6 to 12dB!)
Of course 12 threads. RG58 on FT240-43 toroid. Attenuation according to VNA 25-35db. There's nothing wrong with that. On ft8 lower report. When SSB calling kiwisdr, I can hear myself much more faintly. I don't understand why that is. Where can I go wrong?
@@SmittyHalibut
@@miranovak8098 attenuation, using my test rig to measure the common mode of the choke? Or just connected directly to the VNA?
If you’re using the adapter described in the video (that I sell now, electronics.halibut.com) then 25dB common mode loss is good, but that’s not what you need to measure for reception loss.
Connect your choke directly to the VNA, no adapter. And measure the S21 LOGMAG on that. That should be 0dB loss, or very close to it. If it’s higher than 0dB, something is very wrong with the choke.
The coax might be damaged by the winding process ?
It might be, depending on the cable. Thats why I use RG316, which has a very small minimum bend radius, for winding on FT140 cores. Larger coax on larger cores, you definitely should not abuse the minimum bend radius.
It’s possible that’s what’s happening here on @miranovak8098 ‘s choke.
An excellent presentation, I have been experimenting with common mode chokes for a few years. Further down the comments you will see the observations I have made. Greater than 25dB is easily possible from 160m to 10m, I will retest mine on 6m but I suspect it will be fine there too. Spoiler alert, I no longer believe that winding them with coax is a good wideband solution...
I searched for but couldn't find your other comment. I am interested in your findings and your choke design. Could you repost?
@@geekthesteve6215 I couldn't find them either! I have done some experiments using twisted pairs pulled out of CAT5/6 cable, each pair has a characteristic impedance of nominally 100 ohms so two pairs fed in parallel retains a 50 ohm characteristic impedance. I have made many chokes wound with coaxial cable and always found the bandwidth disappointing. I have pictures and plots of this solution, nearly 40dB of attenuation at 7MHz...
@@geekthesteve6215 This was my previous comment: Hi, I found similar, that interwinding capacitance appeared to be an issue with too many turns of coax on the core. Controversially perhaps my best wideband common mode chokes are wound with twisted pairs from CAT6 cable. Each twisted pair has a characteristic impedance of nominally 100ohms so effectively I have two in parallel to get to 50 ohms. I am getting much better bandwidth and very low insertion SWR.
@@g0fvt That's a really interesting idea. Do you know why the Cat6 results in wider bandwidth than coax?
@@SmittyHalibut I believe that a major factor is the reduced capacitance between turns.
I made a variety of common mode chokes wound with coax, but I found that putting too many turns on the cores caused the high frequency performance to deteriorate. The braid of even the small coaxes has a lot of cross section so adjacent closely spaced turns will couple significantly. The CAT5/6 conductors have a very small cross section, so should result in reduced capacitance between turns. Certainly the insertion SWR is surprisingly low and the common mode attenuation has a very broad characteristic.
I have not tested my first example at beyond 100w carrier...so far...
Great Video!
Hey Smitty thanks for your efforts . Have you done any efficiency tests ?
I’ve made a few baluns the latest being a 49 to 1 on a big ft243 core . It measured really good on the nano vna however it got quite warm using 100W and HF which lead me to question its efficiency ?
73 Chuck VK2SS
I haven't. That's a very different thing than what I'm trying to measure here. What you're measuring is an actual differential mode impedance transformer, not common mode choking. (This is why I hate the term "Balun" because it is commonly used to describe so many different things.)
I'm not sure I understand the magnetic fields in an impedance transforming un-un (the more accurate name of a 49:1 for an end-fed) to speculate on what's happening. But to measure the performance of the 49:1, you'd need a non-inductive resistor that's (49*50ohms=) 2450ohms, and capable of dissipating the full 100W of your transmitter. You'd put that on the antenna side of the un-un, and transmit into it. Then measure the heat in the TRANSFORMER, not the dummy load. You want to see all the heat being dissipated by the load, not the transformer.
If your transformer is cool and the dummy load is hot, then your transformer is good, and I'd look at the antenna for faults. (This is the part where I couldn't tell you what to look for.)
If, while transmitting into the dummy load, your transformer still gets hot, then look in the transformer for faults: Broken wires, cold solder joints, a-symmetries in the windings, etc.
I hope that helps. The 100W 2450ohm non-inductive load is likely to be a challenge to find. I don't have any pointers on that either. Sorry. You'll probably end up having to make it yourself.
I'm building a similar test jig... But have a question
Can you use a single 6 position On-On toggle switch instead of 2 separate 3position switches?
I have a 6 pos in the junc box but I'm worried the cables from the two channels on the VNA would be too close together when soldered on the 6pos switches terminals.
*ahem* electronics.halibut.com/product/common-mode-current-choke-test-rig/ ;-)
You can, but the risk is exactly what you mentioned: You want to keep the two sides as separate as possible. They will electrically couple, throwing off your readings.
@@halibutelectronics Awesome thanks so much for the help..
As for the link... I am building it myself ... "to save money" ...
Of course I'm already at least $60 deep into it ... and now I have to wait on some switches ..
BEST HOBBY EVER!!! ;-)
@@grumpycat_1 Do not ever let me, or anyone else, stop you from building it yourself. That's rad. 🙂
@@halibutelectronics Of course! ...
I think its actually just a matter of reminding myself that the value of DIY'ing isn't to "save money" b/c it rarely works out that way... and thats OK...
That being said sometimes it is smarter to just buy it (especially tools) ...
@@grumpycat_1 sometimes you want a project, and sometimes you just want the result. There’s no shame in sometimes wanting the result without the project, nor is there shame in wanting the project even though the result is cheaper/easier/quicker/whatever-er. Just, ya know, be aware of which you’re doing and why. :-)
Finally. A video that does a good job of explaining what’s going on with baluns. Question: did you try placing two DIFFERENT types of baluns in series? For example, two different baluns with complementary responses?
I hope you plan to do more videos. You do a great job and I’ve subscribed to your channel, so you might as well :)
I haven't! That's a really good idea. I think I spoke briefly about different mixes of cores, using a 31 and a 43 to combine their response curves, but I did not talk about different builds of chokes.
As for more videos: I do! I'm neck deep in trying to get SOAR into production right now, so I'm short on time for doing videos. Having said that, I am working on another QSO Today presentation for the September 2022 Expo. It's on a totally different topic than Common Mode Current Chokes, but I hope it'll be interesting too. :-)
Excellent video brother
You've probably figured this out by now, but when you keep saying you are making the circuit so that it is less resonant in that one place on the spectrum so much, what you are actually doing is lowering the "Q". Low "Q" equals wide bandwidth without the "performance" at any one frequency than you had before you started lowering said "Q". It's the same principal as the tank circuit in your vacuum tube linear amplifier. We shoot for about 12 "Q" because there we get a pretty wide bandwidth in which to work without making it so that when we go to tune it up on our frequency of choice within a band, we don't have so high a "Q" that if we breathe on the tuning capacitor we're not off into another band. I once built an amp with a "Q" of 40 or so and the tuning was so sharp that as the tubes warmed the amp up it would go out of resonance and I'd end up arcing everywhere because the plate circuit was so far out of resonance. A life lesson learned...
Same thing applies here, though all you see is no help from the choke you designed. It's all about spreading that RF around nice and evenly and giving up performance where performance is not needed, as is in this case. It's all about trade-offs. Trade performance for function.
73
AA5IT
So the swr meter in my cb radio shows all is good, no problems. So I went and bought me a external meter only to double check. I connected it, and what do I see... the meter says no good, swr way up there. Took a look at my radios meter and now it’s also showing a high swr all of a sudden.
So I start to adjust the antenna a few times, checked if it did help and yes it did. My external meter shows a low swr. The radios swr agrees also, wow this is great! All I did was to plug the meter out and back with the antenna in the radio.
But no, now after I disconnected and put the meter away, the radio says the swr is high!
What the hell is going on? And how can I learn to fix this because I don’t have money now to go out and buy another meter. And since this is happening on my radio, I wouldn’t be able to make sure an older radio that I planned to buy would be sure. I could not even help if someone where to ask me to come with my swr meter and make sure there’s are not too high.
What can I do to make the radios swr meter work together and show same good results like the did, until I discontent the external one. Radio is showing very high again.
I am not an engineer but have been measuring antennas SWR for about 50 years and my guess is your coax length without the SWR meter and patch is resonant at 11 meters. Probably a multiple of 108 inches (a quarter wave at 11 meters or 27.255 mHz) times the velocity factor of the coax and an example would be .7 (velocity factor of your coax may be slightly different) times 108 inches, or a multiple of 75.6 inches, or 151.2 inches or 226.8 inches, etc.. Try leaving your patch cord in the line by replacing your SWR meter with a double female barrel connector and then check your SWR at the radio.
@@geekthesteve6215 Thank you, that sounds reasonable. Would a shorter coax cable from the swr meter to the radio help get reliable readings you think? It would be great if I could trust the swr meter if I ever should get a radio that has no built in swr meter or high swr protection.
Maybe a dry soldered joint in the new meter? Have you got r.f. chokes?
I was given a supposedly non working MFJ160-10 tuner by a friend but I used it for two years, we went to Wales and still good but connecting it at home it no longer worked. Looking inside revealed nothing, it all looked perfect until I poked around and found a loose wire. It had been touching but not soldered.
G4GHB.
@@bill-2018 Ohh this could be, because I bought it on amazon like budget version. The package of the cables says “uhf” so I was thinking it might be the reason, but then again a cable is just a cable, so I’m not sure 🤔
Thanks a lot for this video, now it is clearer for me, except that you speak fast and for a non-english-speaker it's not easy to understand but i've seen it two times and it's ok now !😁
Very nice, thank you
_Cal Poly?_ Which one? I graduated from _Cal Poly, Pomona_ in 1976 with a BSEE.
_73 de AF6AS_
Well… shoot… I used to be concerned that my HF feedline was too long, & the extra coiled up behind the house was picking up noise.
NOW I know what I’m going to do with a LOT of that extra length. I might wind up (intended) putting 3 or 4 or 5 of these in series, & see if the 40m hash from the washing machine can be quashed.
Haha! FERRITES FOR THE WIN!
BTW: engineers do not build e.g. bridges by a "trial and error" approach 😅
Dam it mark.....strong words followed by other strong words some bleeps and then some less strong words.....now I need to rebuild my main choke as I'm useing 4 ft140 31 and 4 ft 140 43 toroids in a kind of binocular shape with 6 turns ish of my coax then directly into the radio I measured about 25 ish db down across hf and now I'm wondering if I split it in 2 and do 2 of the 31 mix and 2 of the 43 with 3 turns then repeat again so I have the 3 turns theory satisfied and the chokes In series theory also satisfied... dam you ye bleepus maximus
🤣🤣🤣
I’ll point out that you can MEASURE these things using the kit I sell now if you’d like, so you know for sure which configuration is better.
I’d also request that you take notes, take pictures of the graphs, and at least write up your results, or do a video showing your results. Somehow, share it with the world. I will absolutely link that here and boost it. I haven’t done anything with binocular chokes, or mixing .. mixes.
You need to relearn your theory. An inductor does NOT stop the flow of AC or RF current no matter what inductive reactance present. What stops and attenuates common mode is resistive parallel resistance that dissipates the common mode. An inductor does not consume power it just changes it's phase. Cores create this resistance within the core characteristics.
Thanks for the constructive input. I‘m sure Marc appreciates it 👍🏼
It helped me understand it better.
5:28 ... "Necessarily balanced, becuzz physics" Dude! It's the TEM wave. Not deep physics. Though, of course, real coax isn't perfect so there's a tiny CM.
Good info but dude, your soldering needs some work.....LOL, Thanks for the video..
Lewis Dorothy Williams Melissa Allen Kimberly
Inductances don't provide a load for energy. Only resistance can load energy. Your 2 core balun does not provide better rejection because it increases inductance since inductance can't dissipate energy. You did not actually measure any currents any where. You would make a very poor example of an engineer. You can't just say something is true. You have to show it. And if inductances can't dissipate power, where is the resistance at?????
Does it make you feel better about yourself to tear into others like that? I'm seriously asking, because I don't understand why anyone would word that the way you did. You can challenge the information I present without making it a personal attack. I welcome discussions about the content presented, and if you can convince me I got something wrong, I'll happily own up to it. But when you start off with "You're a very poor engineer," I'm less inclined to even listen to the potentially valid point you're bringing up. (See below.)
And this completely ignores the point that good engineers aren't always right. Good engineers can make mistakes and learn from them. If I am wrong here, that doesn't make me a "poor example of an engineer." It makes me human.
To address the technical point you're trying to make: We aren't trying to dissipate the power, we're trying to block it and prevent it from flowing in the first place. The chokes provide a high impedance to the common mode currents, which actually reduces the amount of current flowing. When you compare the "about 1000 ohms" of inductive impedance in the choke against the "about 50 ohms" of impedance in the antenna or feed line, the current will predominantly use the lower 50 ohm impedance. The higher the impedance (whether resistive or reactive) of the choke, the better this effect.
"You didn't measure any currents anywhere." I did not measure any currents DIRECTLY, that is true. But I did measure the signal strength loss by converting the differential signal out of the VNA into a common mode signal, passing it through the choke as a common mode current, then converting it back to differential mode so the VNA can measure the signal. This is analogous to what's ACTUALLY happening in real life. Any attentuation, whether from resistive effects, reactive effects of the chokes we're building, interactions with the environment, or anything else, will show up as S21 loss (negative "gain" values as the NanoVNA calls them). By keeping the other loss effects relatively constant and only changing the choke, we can measure the effectiveness of those changes on the choke's ability to prevent common mode signals.
Is that measuring current? Fine. Not directly. But I am directly measuring what we actually care about, so I'm ok with that.
Early radio engineers said wavelengths below 200 metres were of little use and wavelengths above were required to make long distance communication so radio amateurs were given everything below 200m.
Only when amateurs started making long distance contacts were they interested. The 'professionals' got it wrong!
G4GHB.
Moore Margaret Harris Kimberly Perez Susan
Excellent thank you very much