Announcement

Collapse
No announcement yet.

Low Impedance Pickup Research

Collapse
This is a sticky topic.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Originally posted by Arcadian View Post
    Thanks for that Guy, I'll give them a try.

    Hope youre having some success with your experiments. Is this for electric guitar?
    I still have a '64 Burns short-scale jazz which has three of the low-impedance 'split-sound' pickups (about 30 Ohms if I remember right, plus two matching transformers).The wiring for them is unbelievable (complex, that is) .I've also got copies of some of the original Jim Burns patents which I must dig-out.

    Richard
    Richard,

    One of my first summer jobs after freshman year in college, when I had just started driving my first car, was at the Ampeg factory in Linden NJ. At that time about 1964, Ampeg started distributing Burns Guitars and I had a chance to see the inside wiring and it was pretty complex compared to other U.S. models. If I remember correctly, it had a "wild dog" setting which was an out of phase sound. By using split coils, one coil for the lower 3 bass strings and another coil for the 3 higher strings, they could use the bass side from the neck pickup and the treble side from the bridge pickup to achieve a wide range of tonal variences. By using and mixing the six coils from three pickups through two transformers they could also maintain a humbucking arrangement to minimize noise. That is where I was first introduced to concept of low impedance pickups and have been fascinated by them ever since.

    Thanks for reminding me of that.

    Joseph Rogowski

    Comment


    • Thats it exactly! I bought mine in '66 (sounds like I'm probably a few of years younger than yourself), they werent that popular but I liked them. Its probably worth a few hundred dollars currently. Cost all I had, another £30 would have got me a '60s Strat!, -such is life!

      I've just been playing the latest iteration of the pickup, couldnt put the guitar down, I'm over the moon with it.The response is very good but what is most noticeable is clarity and power of the 'touch' harmonics, -I actually find I'm playing different as a result.

      The string balance was a real headache, after trying numerous things I ended-up with a bit of a left-field solution, but it works great. Round magnets, 5mm x 2mm(high) under strings 6,5,4, 3mm high under string 3, and 2mm exactly centred between strings 1 and 2 , so 5 magnets rather than 6.
      I think the problem is a bit worse than usual because of dropped tunings, ie slacker strings so I'm guessing more excursion.
      I've also made a 'bar' version to pickup 12 sympathetic strings on one instrument, not wired yet but looking forward to that.
      Getting late here but will post a couple of images when time allows.

      Richard

      Comment


      • Here is an image of the Mk4 or Mk5 version of the pickup (lost count), which I'll probably go with for now as I'm reasonably happy with it.Shows the slightly odd mag arrangement (the yellow stuff is just double-sided to fix the ebony cover slip). The other (which I hope is of interest to bbsailor) is the '64 Burns with the low z split sound pickups. The Burns museum in Sweden reckons the white finish (Albino) is pretty rare.

        RichardClick image for larger version

Name:	burns2.jpg
Views:	1
Size:	85.7 KB
ID:	830891Click image for larger version

Name:	Loz4.jpg
Views:	1
Size:	66.7 KB
ID:	830892

        Comment


        • Originally posted by Arcadian View Post
          Here is an image of the Mk4 or Mk5 version of the pickup (lost count), which I'll probably go with for now as I'm reasonably happy with it.Shows the slightly odd mag arrangement (the yellow stuff is just double-sided to fix the ebony cover slip). The other (which I hope is of interest to bbsailor) is the '64 Burns with the low z split sound pickups. The Burns museum in Sweden reckons the white finish (Albino) is pretty rare.

          Richard[ATTACH=CONFIG]25040[/ATTACH][ATTACH=CONFIG]25041[/ATTACH]
          Richard,

          How is the string balance with the modified location of the transformer mounting position and relocation of the B-E string magnet?

          Joseph Rogowski

          Comment


          • |Hi Joseph

            The re-location of the break (slot) in the string loop helped a little but unusually the top E string was over-powerful too, hence the need to reduce both further.
            I mainly play quiet solo stuff in DADGAD and other drop-tunings so this is probably the most demanding in terms of string balance, and the arrangement now works very well indeed (for my particular requirements at least). The balance is such that after a few moments playing I didnt give it another thought.
            I did try mixing it with the piezo undersaddle already installed and although it did perhaps add a little something, not enough to be worthwhile and I'll probably remove it (piezo) once I have the finalised version of the Lo-Z installed.

            One thing is noticeable though, the clarity of the pickup is such that it will show any faults in the guitar setup, -I have a bit of a buzz on 1st string after about fret 10 which isnt much of a problem acoustically but becomes very apparent through the pickup.
            I've just ordered some CSE-187L's and a few other types, and if this offers a further improvement I will be very happy indeed!

            What I like about this concept is the limitless design opportunities it affords,the ease of making (my last one took about 1/2hr) and the ridiculously small cost. I've played with pickups by Shadow, LR Baggs, Fishman and currently have EMG and B-Band installed and they all have significant shortcomings for my purposes, and it's a very expensive game too!

            One other query, if I may, is it the case that the string-loop (brass plate in my case) should be grounded? -I'm sure this was covered somewhere in the posts but its taking a while to assimilate it all!

            Thanks, Richard

            Comment


            • Originally posted by Arcadian View Post
              |Hi Joseph

              The re-location of the break (slot) in the string loop helped a little but unusually the top E string was over-powerful too, hence the need to reduce both further.
              I mainly play quiet solo stuff in DADGAD and other drop-tunings so this is probably the most demanding in terms of string balance, and the arrangement now works very well indeed (for my particular requirements at least). The balance is such that after a few moments playing I didnt give it another thought.
              I did try mixing it with the piezo undersaddle already installed and although it did perhaps add a little something, not enough to be worthwhile and I'll probably remove it (piezo) once I have the finalised version of the Lo-Z installed.

              One thing is noticeable though, the clarity of the pickup is such that it will show any faults in the guitar setup, -I have a bit of a buzz on 1st string after about fret 10 which isnt much of a problem acoustically but becomes very apparent through the pickup.
              I've just ordered some CSE-187L's and a few other types, and if this offers a further improvement I will be very happy indeed!

              What I like about this concept is the limitless design opportunities it affords,the ease of making (my last one took about 1/2hr) and the ridiculously small cost. I've played with pickups by Shadow, LR Baggs, Fishman and currently have EMG and B-Band installed and they all have significant shortcomings for my purposes, and it's a very expensive game too!

              One other query, if I may, is it the case that the string-loop (brass plate in my case) should be grounded? -I'm sure this was covered somewhere in the posts but its taking a while to assimilate it all!

              Thanks, Richard
              Richard,

              Once you obtain the CSE-187L current transformers try this modification. Move the CT mounting slit so it is directly under the B-string. This will cause the B-string to sound a little lower because the view of the current path from a side view takes a dip directly under the B-string as the width of the U-shaped CT primary is spaced .375 inches or the same as the string spacing. This should make the B-string less dominant while keeping the E-string a little more balanced. Solder a wire to the underside of the brass plate or the low impedance string loop. You can even attach a wire to the CT primary below the string loop. Make sure you attach this string loop ground to the metal frame of the CT and then continue it to the shield of the 2-conductor mic cable. Use a stereo or 3-conductor jack on your guitar and wire the two CT output wires to the the center 2 conductors of the mic wires or pins 2 and 3 of the XLR mic connection and the ground to pin 1. This should make it very low noise. Keep up posted about your progress.

              You made a nice creative use of the available space. What is the thickness of the brass plate? How will you attach the transformer to the underside of the brass plate? Do you have access to a copper plate, the same thickness, as it will have a lower resistance and will affect the tonal voicing for your experimentation.

              I hope this helps?

              Joseph Rogowski

              Comment


              • Sorry to have been a stranger for a couple of years now....

                Mental modeling of a single turn under-string pickup with a step-up transformer has led to rejection of off-the-shelf transformers such as the oft-mentioned Triad CSE187. Why? Core losses appear to dominate for frequencies greater than the 400 Hz upper limit in the Triad specifications. Not Triad's fault, of course; we're trying to use the thing for something it's not intended to be used for.

                Perhaps I'm more sensitive to the higher registers because of my specific application. Going two full octaves above the open tuning is not uncommon on a steel. Nevertheless, perhaps what follows may be useful to all.

                I believe the Triad CSE187 uses a silicon steel E-I laminated core. The laminations appear to be about 0.020" (0.5mm) thick. Fine for 60 Hz applications, or even the 400 Hz used in aircraft electrical systems. But to reduce core loss from eddy current to acceptable levels at 2 kHz, E-I laminations would have to be literally paper-thin (about one-fifth the thickness). Laminations would be fragile and the core would be tedious to assemble.

                I have been looking into powdered iron as a core material. I managed to get about 100 turns of AWG 26 enameled copper around a Micrometals T106-2 toroid. (The nominal OD, in hundredths of an inch, is the "106" in the part number, so the core is about the size of a stack of five quarters or Euro coins. The "-2" indicated that the core uses their material #2.) Based on their data (at Micrometals, Inc. - Power Cores), this has an expected inductance of 135 microhenries. Winding toroids by hand is not the most stimulating task!

                I recently obtained a smaller core (T68-26) produced from a higher permeability material (µ_r = 75, vs 10 for the type 2 material). While the higher permeability might not be an advantage here, there are 26-material E-cores available in a wider range of sizes than in the 2-material. Because of relative ease in winding, an E-core may be the way to go.

                Micrometals products do not seem to be distributed by the usual outlets willing to sell in unit quantities (e.g., Digikey, Mouser). Specialty outlets catering to amateur radio do stock the toroids, but I have been unable to locate a source for Micrometals E-cores in unit quantity (other than a rather large one at surplussales.com). Ferrite E-cores seem readily available, so I may need to investigate those.

                Comment


                • Toroidal transformers

                  Man of Steel,

                  The core laminations used in the ALUMITONE pickups are not THAT thin ("paper thin"). They consist of a high permeability material - I presume nickel - not silicon steel, and they are sorta toroidal. They have two really tiny bobbins of super thin wire and maybe 8 or 10 really elongated U- shaped core laminations that are interleaved to make an elongated rectangular toroid.

                  Toroidal cores are hard to wind if you don't use interleaved core laminations and plastic formers to wind coils on like Lace does, but toroids are inherently hum-rejecting and going to an E- core will lose you that. Of course, you can shield it, but I don't think Joe was recommending the transformer for a finished product - experimentation purposes, really.

                  You can buy toroidal that are meant for current sensing and make a brass screw through the hole part of your single turn sense coil, and you end up with a 1:500 turns ratio without having to wind anything at all. Some have frequency response way beyond the audio range.

                  Brass or copper sheet lets you put low temp solder in the threaded hole that the screw joins to in order to make a complete pickup loop, and to keep resistance, and therefore impedance really, really low.

                  FWIW

                  Comment


                  • Originally posted by Man Of Steel View Post
                    Sorry to have been a stranger for a couple of years now....

                    Mental modeling of a single turn under-string pickup with a step-up transformer has led to rejection of off-the-shelf transformers such as the oft-mentioned Triad CSE187. Why? Core losses appear to dominate for frequencies greater than the 400 Hz upper limit in the Triad specifications. Not Triad's fault, of course; we're trying to use the thing for something it's not intended to be used for.

                    Perhaps I'm more sensitive to the higher registers because of my specific application. Going two full octaves above the open tuning is not uncommon on a steel. Nevertheless, perhaps what follows may be useful to all.

                    I believe the Triad CSE187 uses a silicon steel E-I laminated core. The laminations appear to be about 0.020" (0.5mm) thick. Fine for 60 Hz applications, or even the 400 Hz used in aircraft electrical systems. But to reduce core loss from eddy current to acceptable levels at 2 kHz, E-I laminations would have to be literally paper-thin (about one-fifth the thickness). Laminations would be fragile and the core would be tedious to assemble.

                    I have been looking into powdered iron as a core material. I managed to get about 100 turns of AWG 26 enameled copper around a Micrometals T106-2 toroid. (The nominal OD, in hundredths of an inch, is the "106" in the part number, so the core is about the size of a stack of five quarters or Euro coins. The "-2" indicated that the core uses their material #2.) Based on their data (at Micrometals, Inc. - Power Cores), this has an expected inductance of 135 microhenries. Winding toroids by hand is not the most stimulating task!

                    I recently obtained a smaller core (T68-26) produced from a higher permeability material (µ_r = 75, vs 10 for the type 2 material). While the higher permeability might not be an advantage here, there are 26-material E-cores available in a wider range of sizes than in the 2-material. Because of relative ease in winding, an E-core may be the way to go.

                    Micrometals products do not seem to be distributed by the usual outlets willing to sell in unit quantities (e.g., Digikey, Mouser). Specialty outlets catering to amateur radio do stock the toroids, but I have been unable to locate a source for Micrometals E-cores in unit quantity (other than a rather large one at surplussales.com). Ferrite E-cores seem readily available, so I may need to investigate those.
                    Man Of Steel,

                    Edcor makes ribbon microphone transformers, with a 1 to 37 turns ratio (RMX1...wire leads and RMX1-PC... PC board mount). These transformers use a high permeability alloy with a thickness of 0.014" https://www.edcorusa.com/t/Mtr-Core-Ni.

                    True, that high perameability CT would be better but Triad CSE-187L has good potential to work beyond its published frequency and power limits because it is being used at a current between .05A to.1A depending of the wire gauge of the primary string loop. When you ground the primary string loop to the metal frame of the transformer or put a 2.4K ohm resistor from each of the CT output pin to the transformer metal shell or ground, you will have a very low noise pickup (compared to a higher impedance type) with a wider range of equalization potential. Always use 2 conductor shielded cable and treat this CT pickup like a low impedance microphone for the best results.

                    Another commercial CT is the Prem Magnetics SPCT-251 with 2000 turns on the output and an open space for the primary string loop. AWG 6 square copper wire works very well on this CT. Threading the AWG 6 square wire through the CT and making a string loop to accomodate about a 2" string width, only needs a magnet in the center to make this pickup work. Because the CT has 2000 turns vs 500 turns on the CSE-187L the output impedance will be about 16 times higher or in the 3,200 to 4,000 ohm range and work well into a 25K to 100K line input. Making a good low impedance string loop solder joint is very important for maintaining a stronger bass response and control the pickup voicing.

                    If you can find current transformers with better cores and still have low noise and equalization potential, please let us know.

                    Thanks

                    Joseph Rogowski

                    Comment


                    • Originally posted by MR COFFEE View Post
                      Man of Steel,

                      The core laminations used in the ALUMITONE pickups are not THAT thin ("paper thin"). They consist of a high permeability material - I presume nickel - not silicon steel, and they are sorta toroidal. They have two really tiny bobbins of super thin wire and maybe 8 or 10 really elongated U- shaped core laminations that are interleaved to make an elongated rectangular toroid.

                      Toroidal cores are hard to wind if you don't use interleaved core laminations and plastic formers to wind coils on like Lace does, but toroids are inherently hum-rejecting and going to an E- core will lose you that. Of course, you can shield it, but I don't think Joe was recommending the transformer for a finished product - experimentation purposes, really.

                      You can buy toroidal that are meant for current sensing and make a brass screw through the hole part of your single turn sense coil, and you end up with a 1:500 turns ratio without having to wind anything at all. Some have frequency response way beyond the audio range.

                      Brass or copper sheet lets you put low temp solder in the threaded hole that the screw joins to in order to make a complete pickup loop, and to keep resistance, and therefore impedance really, really low.

                      FWIW
                      Mr Coffee,

                      The reason why they use interlocking U-shaped laminations is because thay can punch out the pickup frame and bend it into the final shape. They don't need to solder a joint to make the string loop, they just use the alumium frame metal and slip the transformer lamination around the metal frame where both coils appear to be in series with the current from both loops going through the two small coils mounted on the U-shaped interlocking laminations. The two coils on the bottom are the current transformer secondary with the metal frame being the primary.

                      The amount of turns on the Alumitone need to provide enough output to drive a high impedance amp input while the 500 turn CSE-187L based pickup needs to be in the 150 to 250 ohms range to drive a standard XLR mic input.

                      As more people play with this stuff and share what they learn, we will evolve the design of these current transformer low impedance pickups.

                      Joseph Rogowski
                      Last edited by bbsailor; 09-13-2014, 04:18 PM.

                      Comment


                      • Typically, toroid cores having really thin laminations are tape-wound, not assembled from stacked punchings. Tape can be far thinner than E I punchings or ring punchings.

                        The high-permeability nickel alloys used to make tape-wound coils are very sensitive to mechanical strain (like mu-metal shields), so the tape-wound core is encased in a cover that keeps winding forces off the core. The core is floating in grease or wax or some kind of rubbery potting agent within the cover.

                        Silicon steel is not particularly strain-sensitive, and can be got very thin, down to 0.006", if memory serves. A common way to handle such thin laminations in small quantities is by photo etching, versus stamping.

                        Eddy current loss varies linearly with the material resistivity, and as the square of lamination thickness and of the frequency: The higher the resistivity, the lower the loss. Silicon steel has higher resistivity than typical nickel-based magnetic alloys. The thinner the laminations, the smaller the loss.

                        Eddy current - power-loss equation

                        If the laminations are thin enough, skin effect in the laminations will not be all that important, so one verifies that the laminations are thinner than the calculated skin depth.

                        Eddy current - skin depth equation

                        Comment


                        • Hi Joe,
                          I agree completely about the stacked laminations and eliminating the need to join ends of the primary loop to slip the transformer on. It also allows Lace to use very small linear coil forns wound with a normal winder, rather than the toroid core and stick (I know there are toroid-winding machines, but I doubt anybody here has THAT kind of winding equipment). Small cheap dynamic microphone transformers are nearly identical if one wants to experiment with the Lace design (yeah, I know you can't sell them because they're patented, but there is no law against experimenting). And winding a linear coil isn't such a pain for the pickup winders that hang out around here.

                          But what I was suggesting to Man of Steel was using a high frequency toroid current sense transformer and using a threaded brass rod to complete the primary loop through the toroid, so you can get very low resistance junctions, especially if you tap the holes with a tap that has been ground smaller so the threads jam into the metal sheet parts of the primary loop. They sell them encapsulated in a little tombstone-shaped package with a hole for the primary to pass through the center of the toroid. You can use two of them and get a semi-conventional looking pickup with one threaded rod through a transformer on both ends and use copper or brass sheet with threaded holes similar in geometry to the Alumitone design.

                          I'm not intending to knock the CSE-187L, but I figured he was interested in toroid construction.

                          FWIW.

                          Comment


                          • Originally posted by MR COFFEE View Post
                            Hi Joe,
                            Small cheap dynamic microphone transformers are nearly identical if one wants to experiment with the Lace design (yeah, I know you can't sell them because they're patented, but there is no law against experimenting).
                            FWIW.
                            Exactly, and experimenting is always good because you learn things. But one should realize that the Lace low impedance design does not offer any significantly better trade off for bandwidth and signal. If you transform the level up at the pickup to get the same output level as a normal pickup, you get essentially the same inductance looking into the cable and have the same frequency response. If you want flatter frequency response to higher frequencies, you can just wind fewer turns on a regular bobbin and use a transformer at the amp or a low noise preamp, if a simple effective design is what you want.

                            Comment


                            • Originally posted by Mike Sulzer View Post
                              Exactly, and experimenting is always good because you learn things. But one should realize that the Lace low impedance design does not offer any significantly better trade off for bandwidth and signal. If you transform the level up at the pickup to get the same output level as a normal pickup, you get essentially the same inductance looking into the cable and have the same frequency response. If you want flatter frequency response to higher frequencies, you can just wind fewer turns on a regular bobbin and use a transformer at the amp or a low noise preamp, if a simple effective design is what you want.
                              Mike,

                              For those who want to experience a little wider bandwidth on their guitar pickups, eliminating the transmission cable losses will offer an audible difference. If you target a pickup output impedance of about 150 ohms (nominal but could range from about 100 ohms to 400 ohms) then you can run your pickup output directly into an XLR mic input of a mixer or go directly into a mic matching transformer (Shure A95U series) before going into the final amplifier and eliminate the 300 pf to 500 pf of capacitance on the high impedance transmission line (10 ft straight coax cable and more for a curly-cue, stretchy cable).

                              Reduced pickup impedance aids in better transmission and lower potential noise.

                              Thanks for your thoughtful comments.

                              Joseph Rogowski

                              Comment


                              • Originally posted by MR COFFEE View Post
                                But what I was suggesting to Man of Steel was using a high frequency toroid current sense transformer and using a threaded brass rod to complete the primary loop through the toroid, so you can get very low resistance junctions, especially if you tap the holes with a tap that has been ground smaller so the threads jam into the metal sheet parts of the primary loop. They sell them encapsulated in a little tombstone-shaped package with a hole for the primary to pass through the center of the toroid. You can use two of them and get a semi-conventional looking pickup with one threaded rod through a transformer on both ends and use copper or brass sheet with threaded holes similar in geometry to the Alumitone design.
                                All this will work, except for that threaded brass rod. If a 5000-turn coil produces a few volts, one turn will produce maybe a millivolt. Such a low voltage will be unable to punch through any kind of surface film. Sufficiently good mechanical joints are very hard to arrange. All joints must instead be welded, brazed, or soldered.

                                Comment

                                Working...
                                X