Announcement

Collapse
No announcement yet.

Crosstalk measurements between inductors

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Crosstalk measurements between inductors


  • #2
    Re: Crosstalk measurements between inductors

    "Dennis arbitrarily used a 26-kHz signal"
    Why would anyone use 26 KHz? Did you mean 2.6 KHz?

    The crosstalk is a function of current
    and, obviously, inductor impedance rises with increasing frequency and therefore
    the current decreases. It would be more reasonable to build a "typical" tweeter
    crossover with one inductor. Something like 8 uF input, that .3 mH shunt and a 6
    ohm dummy resistive load. Short the 8uf to ground to simulate the low source
    impedance of the amp, and measure across the 6 ohm resistor. That would be the
    sense circuit.
    Some thought should be put into the driven circuit. Worst case is probably the
    inductor in a midrange circuit with the most current flowing, perhaps the HP section.

    Perhaps we should look to real world cases, such as the B&W 801 woofer to mid
    crossover where there have been reports of significant crosstalk.
    Last edited by Pete Basel; 03-02-2013, 09:49 PM.

    Comment


    • #3
      Re: Crosstalk measurements between inductors

      Thanks for the update Paul.

      Comment


      • #4
        Re: Crosstalk measurements between inductors

        Nope, I did not mean 2.6 kHz. I wrote what was printed in the article, and in his testing of the ferrite (1.7 mH) and steel laminate (5 mH) inductors, he used frequencies of 4 kHz and 100 Hz respectively, all frequencies chosen "simply for measurement convenience". My goal was to reasonably replicate the tests and results documented by Dennis in his article as explanation on why I keep telling people that a 2" spacing between air core inductors is quite adequate, not requiring the inductors to be stood up or rotated. I believe I accomplished my goal, and while I'm sure others will suggest a "better" way, I don't think the basic results will change in any significant manner. Perhaps the B&W crossover you referenced had some serious flaws regarding component placement/orientation?

        Originally posted by Pete Basel View Post
        "Dennis arbitrarily used a 26-kHz signal"
        Why would anyone use 26 KHz? Did you mean 2.6 KHz?

        The crosstalk is a function of current
        and the obviously inductor impedance rises with increasing frequency and therefore
        the current decreases. It would be more reasonable to build a "typical" tweeter
        crossover with one inductor. Something like 8 uF input, that .3 mH shunt and a 6
        ohm dummy resistive load. Short the 8uf to ground to simulate the low source
        impedance of the amp, and measure across the 6 ohm resistor. That would be the
        sense circuit.
        Some thought should be put into the driven circuit. Worst case is probably the
        inductor in a midrange circuit with the most current flowing, perhaps the HP section.

        Perhaps we should look to real world cases, such as the B&W 801 woofer to mid
        crossover where there have been reports of significant crosstalk.

        Comment


        • #5
          Re: Crosstalk measurements between inductors

          Possibly stupid question, does the geometry of the coils being tested play a role here. The blanket rule of thumb here is 2" but does that apply to coils 1" in diameter the same as those 3"? Just have to imagine the resulting magnetic field is dependent on the size of the circle of electrons driving it?

          Comment


          • #6
            Re: Crosstalk measurements between inductors

            You cannot make blanket generalizations. If the same inductor is used in
            a different application with 10X the current then the field strength is 10X.
            I'd say take the results as a rough guideline.
            Also, let's say that at some low frequency we have 40 dB of attenuation
            in a tweeter circuit, coupling in that range will cause more of an issue
            since a tweeter cannot handle much LF energy. It depends on many factors
            so, keep the guideline in mind but also think about the consequences of
            coupling. The simple thing is to place inductors in the corners of a board to
            provide the maximum distance; why do it any other way?


            Originally posted by Paul K. View Post
            Nope, I did not mean 2.6 kHz. I wrote what was printed in the article, and in his testing of the ferrite (1.7 mH) and steel laminate (5 mH) inductors, he used frequencies of 4 kHz and 100 Hz respectively, all frequencies chosen "simply for measurement convenience". My goal was to reasonably replicate the tests and results documented by Dennis in his article as explanation on why I keep telling people that a 2" spacing between air core inductors is quite adequate, not requiring the inductors to be stood up or rotated. I believe I accomplished my goal, and while I'm sure others will suggest a "better" way, I don't think the basic results will change in any significant manner. Perhaps the B&W crossover you referenced had some serious flaws regarding component placement/orientation?

            Comment


            • #7
              Re: Crosstalk measurements between inductors

              Originally posted by Paul K. View Post
              Perhaps the B&W crossover you referenced had some serious flaws regarding component placement/orientation?
              From memory, I believe that a revision of the system, probably for cost reduction, combined 2 crossover boards into one. The issue came about with the single board version. The problem was reported in VanA's newsletter many years ago.

              Comment


              • #8
                Re: Crosstalk measurements between inductors

                Originally posted by Pete Basel View Post
                The simple thing is to place inductors in the corners of a board to provide the maximum distance; why do it any other way?
                Yes, and placing them orthogonally to each other too.

                Comment


                • #9
                  Re: Crosstalk measurements between inductors

                  Go read the full article that Dennis wrote and you'll see that it covers a lot more than what I was replicating, including larger inductors, both physically and in inductance, and different frequencies. Obviously, as I've said at least twice (in this thread and my other thread about mutual inductance) if you have the room, then by all means spread out the inductors as far as you want. If room is limited, then you'll need to orient them appropriately to minimize both mutual inductance and voltage crosstalk. There's an in-between scenario that I suspect occurs frequently, and I have absolutely no problem with a 2" spacing for air-core inductors laid down flat.

                  In my testing, as I stated, I measured the crosstalk at 5 different current levels in the circuit with the two coils' spools touching and got the same amount of crosstalk in terms of dB or percentage. Obviously the amount of signal coupled increases with increased current, but the amount coupled still was only ~1.7% regardless of current level. With the two inductors spaced at 2", the amount coupled dropped to ~0.2%, which is insignificant IMO.
                  Paul

                  Originally posted by Pete Basel View Post
                  You cannot make blanket generalizations. If the same inductor is used in
                  a different application with 10X the current then the field strength is 10X.
                  I'd say take the results as a rough guideline.
                  Also, let's say that at some low frequency we have 40 dB of attenuation
                  in a tweeter circuit, coupling in that range will cause more of an issue
                  since a tweeter cannot handle much LF energy. It depends on many factors
                  so, keep the guideline in mind but also think about the consequences of
                  coupling. The simple thing is to place inductors in the corners of a board to
                  provide the maximum distance; why do it any other way?

                  Comment


                  • #10
                    Re: Crosstalk measurements between inductors

                    Yes okay I think we are pretty much in agreement. What you are saying
                    is that for air core inductors the coupling is linear by pointing out the constant
                    percentage. This makes sense since air cores are linear by nature.

                    My point about current level, is that the inductors with more turns and more
                    current, such as the woofer inductor may have different requirements due to
                    the stronger fields than those with significantly lower current. Or perhaps the
                    inductor in a poorly designed trap, where the system impedance dips low
                    indicating that the current is high as is seen in some commercial designs.
                    That inductor would have a much stronger field, not that anyone here would
                    make such a mistake.

                    Comment


                    • #11
                      Re: Crosstalk measurements between inductors

                      Yes, we now agree on the primary point I was trying to make. I can run another set of tests but using a pair or larger inductors (in size, number of turns and inductance), as well as a test pitting a large inductor against a small inductor. I have time tomorrow and might do it or maybe you could do it instead?
                      Paul

                      Originally posted by Pete Basel View Post
                      Yes okay I think we are pretty much in agreement. What you are saying
                      is that for air core inductors the coupling is linear by pointing out the constant
                      percentage. This makes sense since air cores are linear by nature.

                      My point about current level, is that the inductors with more turns and more
                      current, such as the woofer inductor may have different requirements due to
                      the stronger fields than those with significantly lower current. Or perhaps the
                      inductor in a poorly designed trap, where the system impedance dips low
                      indicating that the current is high as is seen in some commercial designs.
                      That inductor would have a much stronger field, not that anyone here would
                      make such a mistake.

                      Comment


                      • #12
                        Re: Crosstalk measurements between inductors

                        I ran additional tests today that reasonably simulate, I think, crosstalk effects between large and small inductors, like between a woofer crossover inductor and a midrange or tweeter crossover inductor.

                        Connected in series to the output of a power amp were a 4-mH inductor and an 8-ohm resistor, with the resistor representing a driver. The power amp input was driven by a sine wave at 250 Hz from my signal generator. For the "pickup" inductor to "receive" the field radiated from the driven inductor, I used two, one an identical 4-mH and the other with a 0.3 mH value. I measured the crosstalk in both pickup inductors at two power levels into the 8-ohm resistor, 1 watt and 10 watts, and with two spacings between the driven and pickup inductors, touching and 2". For each set of conditions, I measured the voltage across the driven inductor and the pickup inductor, then calculated how much the voltage in the pickup inductor was in terms of dB and percentage compared to the voltage in the driven inductor. Of course, all coils were air-core and both driven and pickup coils were laid down flat on one of their spools caps. Voltages were measured with my Fluke 8050A True RMS DMM.

                        Using a 4-mH pickup inductor, at 1-watt and zero distance, the crosstalk voltage in the pickup inductor was 2.3% or -33 dB of that in the driven inductor. Increasing the spacing between inductors to 2", the crosstalk voltage was 0.28% or -51 dB. Increasing the power to 10 watts, at zero distance the crosstalk voltage was 2.2% or -33 dB. Increasing the spacing to 2", the crosstalk voltage was 0.28% or -51 dB.

                        Using the 0.3-mH pickup inductor, at 1-watt and zero distance, the crosstalk voltage was 0.48% or -46 dB. Increasing the spacing to 2", the crosstalk voltage was 0.05% or -66 dB. Increasing the power to 10 watts, at zero distance the crosstalk voltage was 0.47% or -47 dB. Increasing the spacing to 2", the crosstalk voltage was 0.05% or -66 dB. Note that I rounded off all of my calculations to just two significant numbers simply for convenience in this report.

                        I'm not sure what you might conclude from these tests. I do note than when the pickup coil is much smaller than the driven coil, the crosstalk voltage is much smaller in it than when both coils are the same, whether both are 4 mH that I used in today's tests, or both are 0.3 mH from my previous tests that initiated this thread. Using a 10-watt input and a spacing of 2", with a pair of 0.3-mH coils the crosstalk measured -54 dB/0.2%, with a pair of 4-mH coils it measured -51 dB/0.28%, and for a 4-mH/0.3-mH coil combo, the crosstalk measured -66 dB/0.05%. The other interesting outcome is, that, the amount of crosstalk between similar inductors (identical in my case), whether large or small, is pretty consistent. With a 10-watt input (to the 8-ohm resistor) and a spacing of 2", a pair of 4-mH inductors had crosstalk that measured -51 dB/0.28%, and a pair of 0.3-mH inductors had measured crosstalk of -54 dB/0.2%.

                        Any thoughts or comments, Pete?
                        Paul

                        Originally posted by Pete Basel View Post
                        Yes okay I think we are pretty much in agreement. What you are saying
                        is that for air core inductors the coupling is linear by pointing out the constant
                        percentage. This makes sense since air cores are linear by nature.

                        My point about current level, is that the inductors with more turns and more
                        current, such as the woofer inductor may have different requirements due to
                        the stronger fields than those with significantly lower current. Or perhaps the
                        inductor in a poorly designed trap, where the system impedance dips low
                        indicating that the current is high as is seen in some commercial designs.
                        That inductor would have a much stronger field, not that anyone here would
                        make such a mistake.

                        Comment


                        • #13
                          Re: Crosstalk measurements between inductors

                          Okay- while you're at it-

                          I would like to see the effect on a cap using a pickup coil, and placing the cap inside the coil. I know caps have an affect on coils if the body is silver-mica or aluminum, as it should in theory, but what about vice-versa?
                          Can you do this while you're at it?

                          Thanks,
                          Wolf
                          "Wolf, you shall now be known as "King of the Zip ties." -Pete00t
                          "Wolf and speakers equivalent to Picasso and 'Blue'" -dantheman
                          "He is a true ambassador for this forum and speaker DIY in general." -Ed Froste
                          "We're all in this together, so keep your stick on the ice!" - Red Green aka Steve Smith

                          *InDIYana event website*

                          Photobucket pages:
                          https://app.photobucket.com/u/wolf_teeth_speaker

                          My blog/writeups/thoughts here at PE:
                          http://techtalk.parts-express.com/blog.php?u=4102

                          Comment


                          • #14
                            Re: Crosstalk measurements between inductors

                            I'm not sure exactly what you want to do and measure. So would you please elaborate and be as specific as possible (I apologize for seeming dense and no promises). I think you want a cap to be placed inside the air core of a "driven" inductor, then measure the cap to see if its value changes?

                            Edit: Or, do you want to see if there's any crosstalk voltage picked up by the cap?
                            Paul

                            Originally posted by Wolf View Post
                            Okay- while you're at it-

                            I would like to see the effect on a cap using a pickup coil, and placing the cap inside the coil. I know caps have an affect on coils if the body is silver-mica or aluminum, as it should in theory, but what about vice-versa?
                            Can you do this while you're at it?

                            Thanks,
                            Wolf
                            Last edited by Paul K.; 03-04-2013, 04:57 PM.

                            Comment


                            • #15
                              Re: Crosstalk measurements between inductors

                              Great work. I was curious what might happen with the two coils being different sizes (and this is the most likely scenario, IMO.) Sounds to me like it might be an impedance mismatch -- working in our favor in this case -- causing transmission losses from the imbalance. Cool stuff. I really appreciate your taking the time to test all this.

                              I'm inclined to agree with your results. Below -40dB, you're looking at less LF-to-HF bleed level than the highpass filter itself will allow. Less is always better, but that'll do.

                              Comment

                              Working...
                              X