Monday, October 12, 2020

CM choke, RF choke, RF isolator, line isolator, RF current filter

 Years ago Hans PE1BVQ was so kind to give me a common mode filter. It was at the time I had issues with my VDSL2 connection. I tried everything to avoid the coax from radiating and receiving unwanted signals. However, the CM choke didn't solve the problem. It was solved by someone else placing software  with amateurradio frequency notches in our VDSL2 modem.

Personally I never had any good results with CM chokes, not the results showing in this videos:

Of course it is all a matter of how much common mode there actually is present on your coax lines. If you haven't got it the difference will not be so obvious as shown in the video. 

I got 2 common mode chokes, the one from PE1BVQ with a unknown ferrite but probabely it is mix #43 material? The other one is built by me, it uses a 4C65 ferrite. I always left the balun from Hans in my coax line to the vertical but does it really help, I can't say. I tried to test and see any differences in RX and TX but it isn't detectable. I guess I have not enough common mode (lucky me!). 

The question now is, do both common mode chokes work? I haven't got expensive VNA (Vector Network Analyser). I know the nanoVNA is cheap but haven't got that as well. I got a digital multimeter which worked, but went defective just when I wanted to use it for this experiment. So what is left is a cheap 6 euro multimeter. I came across a nice diagram from a CM sensor, it looked easy so I build one. However due to other distractions I didn't read well, I assumed I had to measure current so I choose mA to do the test. Imagine it gave some strange readings, I was confused. Discussed it on a dutch HAM forum and then I saw my fault...I should choose mV. The current is rectified and you measure voltage over the resistor which is in parallel with the meter (not drawn in diagram). Anyway, it gives better readings now.

But what to measure now. What should I measure? I asked again in the dutch HAM forum and PA3DJS was so kind to tell me that you should measure without the choke first to see any difference. When connected both in/output should read almost no difference, but of course lower readings as without the choke. 

Well, this is what I measured in the end:

Input 10W FM on 5365KHz

Without CM choke: 450-455 mV on several places on the coax

With PE1BVQ choke: 230 mV on several places on the coax. Near the choke input (TX side) 3 mV and near the choke output (antenna side) 6 mV

With PE4BAS choke: 240 mV on several places on the coax. Near the choke input (TX side) 94mV and near the choke output (antenna side) 47 mV

Interesting to see that the chokes obviously work. Would be nice the see how much dB they actually surpress. I haven't got the equipment and knowledge for that (yet). I guess most average HAMs don't have that. But the CM sensor is easy to build and a nice experiment to see if a choke actually works.

There are many ways to build a CM choke. Interesting to see so many designs, some working well, some working just good enough. It looks like there isn't a design that is really the best. To have proof that it really works and has enough CM surpression you really need a VNA.

Videos I really like to see, if only it was for the funny accent, are those from TRX bench on this subject. The way Peter tells about the CM choke or 1:1 balun is really understandable, you learn a lot about this subject for shure.

This really was a nice experiment and wish I had more time to experiment with these kind of homemade filters & transformers. No doubt I will learn a lot from it...

Recent article on PA3HHO Pleun's blog:


  1. Hi Bas
    I got into making chokes last year after reading an article.I made a few for the 30m vertical i had and was impressed with the results, well worth the effort. It was a 240/type43 with 12 turns of rg412.
    good luck, interesting subject
    73 stu

    1. Hello Stu, yes the 43 type seems to be the best overall type of ferrite. It is indeed a interesting subject, however I never had the results they advertise. Not on RX and not on TX. Of course it could be I never had the right CM choke? To know what is best you really need a VNA is my conclusion. Even commercial CM chokes could be not that good, some are working some are not. 73, Bas

  2. Hi, Bas,
    When you measure, all depends on where you place the common mode choke "balun". If you place it near the TRX the coax is still a part of the antenna, and you will measure RF on the outside of the coax. If near the antenna, it limits the coax's function as antenna, and the cable should radiate/receive less noise.
    73, Jan

  3. Niemand in de buurt met een VNA? Helpt je idd echt om te *zien* wat er gebeurt. Je woont helaas net ff te ver om de mijne te kunnen inzetten.

    1. Jammer Lars, ik denk dat ik een nanoVNA ga aanschaffen. 73, Bas

  4. I had an RF issue in the shack at one point, where transmissions even at 10-15W would switch on/off a lamp in the shack (one of those touch-sensitive lamps). I tried all sorts to solve it but after I had a shack rip-out to install new radios, the problem disappeared completely (even if I tested at higher power with my higher licensed friend). I think there must have been a problem of some sort with all the connections or routing at the back of my radios which was causing me grief :-)

    73, Tom, M7MCQ.

    1. Tom, RFI can be caused by many things. I once had a defective jumper coax from radio to SWR meter/Tuner. I can tell you strange things happened. Till I changed the cable for a new one. Everything was allright again. 73, Bas

  5. I put up with 40 minutes of the German's video. What seems like a careful and technical analysis in fact is a load of rubbish.

    First point is the argument as to the 'first job' of a balun.

    In all cases here, I have used a balun with the single purpose of impedance transformation. The reduction in common mode, if any, was always considered a possibility, but without any expectation it would be particularly effective.

    The second point is that the video considers only a voltage balun. He doesn't consider a current balun at all. This is a serious omission.

    Lastly, he seems to think that a single balun and ferrite type can produce the same effective attenuation of CM over the entire HF spectrum. Any careful radio text will tell you this is entirely unreasonable, and that different mixes are needed at different frequencies.

    So the only 'a-ha!' moment is nt to do with any failure in the concept of a balun, but in the German's improper understanding and expectation.

    1. The problem is that everyone knows best. Confusing me. I tried to make a current balun from coax but I don't trust it at all. Will look into other possebilities.Yes, there is no ferrite till now that covers the whole HF with the same CM surpression results. But I've learned you can put several different chokes in line to get a broadband surpression. Well, I've a lot to learn about this subject that is for shure. 73, Bas

    2. Coax should be fine. But this is maybe better:

  6. The late G3TXQ had some very interesting info on his website about CM chokes. I just found out his website is down - which should not come as a surprise I guess. However, part of the info he shared popped up here:
    I used his measurements as a starting point to build an effective CM choke for my 40m c-pole. As stated above you can clearly see that a choke is effective in a particular frequency range based on the design (ferrite or air core, type of ferrite, coax turns, ...).

    1. Yes, G3TXQ has interesting info regarding type of ferrites and CM surpression. PA3HHO has published some as well. However if I build one I would love to measure it to know what it is actually doing. That is the only way to determine if I built it right. 73, Bas


Thanks for your comment. Bedankt voor je reactie. 73, Bas