ad: UR5CDX-1

Snap on Ferrites

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by K6GBW, Jan 17, 2021.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: abrind-2
ad: Left-2
ad: l-BCInc
ad: l-gcopper
ad: Left-3
ad: Subscribe
  1. K6GBW

    K6GBW Ham Member QRZ Page

    I knew I'd find the answers I needed here. Thanks for the advice and the links. I'm sure I'll have this all tweaked in no time!
    WZ7U likes this.
  2. W9KEY

    W9KEY XML Subscriber QRZ Page

    Don't remember ever seeing this comparison of winding technique versus impedance. The slide is credited to K3LR.
    Have others actually verified this is true and repeatable??

    I'm skeptical that a 10-15% (or more) improvement in choking impedance is possible by simply using a crossover. Had assumed the crossover was just a convenient option to simplify mechanical winding - which can be challenging with stiff cables like RG-400. Given the wide core tolerances, tightness of the windings, variations of interwinding capacitance, and general difficulty in making repeatable measurements of chokes - I wonder if the above difference is just normal "tolerance" - and has nothing to due with the crossover winding technique??

    Said another way - if you hand built two "identical" chokes (wound exactly the same), perhaps using cores from different manufacturing batches - would you see similar impedance vs freq differences from one sample to the next??

    But if not - and the crossover actually does makes such a big difference:
    1. Why does the impedance increase so significantly with a simple crossover?
    2. Why don't recent published articles (including K9YC's Choke Cookbook) ALWAYS recommend such a crossover?

    ..... W9KEY
    WB2UAQ and AK5B like this.
  3. WA7ARK

    WA7ARK Ham Member QRZ Page

    K9YC and others have published this. It might predate the 2018 cookbook. I have confirmed it with my Vna.

    The difference is due to minimizing stray capacitance from the input side to the output side.
    WB2UAQ likes this.
  4. W9KEY

    W9KEY XML Subscriber QRZ Page

    Hum - guess I'll have to read through K9YC's website a bit harder for related publications:
    Given the volume of material- I'll report back in a year or two!!

    But it still seems strange (given the significant difference in K3LR's graph), that the 2018 Choke Cookbook didn't mention cross winding, nor include any such photos. He did specifically mention the repeatability problem, and highlighted several mechanical fixtures used to minimize mechanical variation between samples - as well as having built and measured hundreds of models.

    Mike - out of curiosity - do you recall if the VNA differences you noted were for chokes constructed from two different cores, or were measurements made using the exact same core - first wound straight, then changed to a cross configuration?

    I had past experience in the electro-ceramic business (but not inductors) and can certainly empathize with this quote from the 2018 Cookbook page 11, "This study revealed extremely wide variations in the fundamental properties of the 2.4-in cores purchased over a period of 12 years, and variations of +/-10% for the newly purchased cores. These variations have profound effects on the characteristics of chokes wound on these cores."

    From my experience, electronic components "can" exhibit quite tight tolerances within a given manufacturing batch (or not), but holding tight tolerances over multiple lots/weeks/years is another matter entirely. Manufacturers never want to scrap any viable production parts - and tolerances & prices are set accordingly. I have no idea what type or size sintering kilns are used in the ferrite business, but given the large physical size of the cores, I'd not be surprised if some tolerance variability was simply due to the individual core's position in the kiln during its firing process (versus its neighboring cores). And of course, there are many additional process variables requiring careful control ......
    WB2UAQ and PU2OZT like this.
  5. WB2UAQ

    WB2UAQ Ham Member QRZ Page

    I have experienced the same issues with ferrite over the past 18 years especially. +/-25% is very possible when it comes to the variation in some of the electrical characteristics. Recently, now that I'm retired, I ordered from Mouser and these last 5943003801's (FT-240-43 Amidon's p/n change) have an AsubL of about 25% lower. AsubL is nominally about 1000 nH/T^2 and these are coming in at only 750 nH/T^2. I check this at 1 kHz and vary the current supplied by the GR bridge to see how the flux density impacts the results. I followed material coming from about 3 core suppliers since about 2003 and glad I kept samples thru the years. Most of the cores we use are used for EMI suppression and one of the mfrs improved a particular core's EMI suppression characteristics and this screwed things up quite a lot for a transformer application (in this case we had amplifier instability issues over temperature and sometimes under certain load impedances). Another interesting issue, probably not for amateur radio concerns, is magnetostriction for high perm magnetic materials. Magnetostriction shows as a variation in perm with pressure on the core. I observed this firsthand for one magnetic application where the initial perm is spec'd to be 5000. I applied the winding tighter and then looser and there was a significant change in the measured inductance. These parts were wound with a hook-winder and this resulted in tighter or looser windings depending on the machine operator / settings.

    After re-visiting K9YC's material please let me know if he covers flux density in his discussion. Flux density is a big consideration that many ham radio treatises don't address. Often the core heating is attributed solely to the measured ESR only. I mentioned this on QRZ a few times and no one seems to give a rat's butt and the flux density doesn't have to get anywhere near saturation for this to be an issue:) Just get over 100 gauss and the impact will be discovered. 73
  6. WB2UAQ

    WB2UAQ Ham Member QRZ Page

    I second this observation:) Having wound many inductors at a given freq, more inductance can be realized and higher Q.
    AK5B likes this.
  7. W7HV

    W7HV Ham Member QRZ Page

    Where the ferrites are placed on the line can make a big diff depending on the situation. Generally, depending the the frequency and length of the feed line, there can be nodes in the common mode current and placing ferrites there will have little effect. They're best placed at common mode current maxima.
    AK5B and PU2OZT like this.
  8. WA7ARK

    WA7ARK Ham Member QRZ Page

    I agree. Simulation, or a "slide-along-the-coax" ferrite RF current-transformer will show where the standing-wave Common Mode current peaks are. In multi-band antennas, you have to check the other bands, too. Adding a CM choke may create CM current peaks where there were none before..., especially on the "other" bands of a multiband antenna.

    Here is my home-brew CM current meter:
    It was originally described in this thread:
    PU2OZT and VE4AJM like this.

Share This Page

ad: wmr-1