Low Impedance Pickup Research

[bbsailor]

I want to share some recent low impedance pickup research with the foum members. The photo was taken on top of my research notebook where I placed all the parts that I used in this experiment.

The Low Z current sensing transformer is CSE187L and has a 1:500 turns ratio.
The primary side is 250 micro ohms or 1/4000th of an ohm.
The secondary is measured at 18 ohms DC.
The secondary impedance is about 400 to 500 ohms when the primary is loaded by a single loop around a magnet as shown in the photo. This impedance is very sensitive to the low Z impedance side as it is reflected back into the secondary.

I used copper tubing with an ID the same as the wire diameter to make good connections between the low Z primary and the transformer connection. This technique makes changing coils much easier than soldering. The resistance of these connections is very critical that they be very good or low.

I soldered a two conductor shielded audio cable to a 3-pin mic connector that goes to the 500 ohm to 50000 ohm line or mic matching transformer (Shure 95 series). This provided another 1:10 turns ratio boost. This makes the total turns ratio boost 1:5000 for a single coil pickup. I connected the cable shield to the transformer case. You may also want to try connecting one end of the low-Z primary loop to ground also.

On my notebook, along with the parts, I skeched out how a humbucking pickup could be made using this same technique. The output would have twice the output but only if you maintain the low impedance of the primary loop windings. Look to use .25" X .125" copper bars that are configured to accomodate your magnet size. Position the magnets N and S as shown. Make the low-Z transformer connection at the two dots as shown on the photo. This will give you the effect of two humbucking coils in series.

If the magnets are metal or conductive you need to keep the insulation on the wire to keep from shorting out the primary. You can wrap the magnet in electrical tape also to insulate it from the raw copper winding.

In the photo, the green wire is AWG12 with a 2" X .25" X .125" thick (N/S through the thickness), Neodymium from K&J Magnetics (BY042). The other magnets on the left side of the photo are ceramics I had laying around. Although much weaker than the BY042, they worked nearly as loud as I could can get them closer to the strings without sucking in the string.

This is not a refined design but is a crude experiment to demonstrate, for my own benefit, how Lace Alumitones work.

Here is what I learned.

The matching of the low Z primary single loop is critical to have an optimum energy transfer from the string to the transformers.
I could hear a level difference between using AWG 12 and the thicker AWG10. Thicker wire is better in this design.
The Humbucking design should produce a higher output. (My next experiment) It does. I get about 50% more output with the humbucking design using 2 magnets. The resistance of the primary loops needs to be very low (the lower the better) to improve the efficiency of this design.

I hope this answers questions from those who are curious about low impedance pickups. This technique eliminates coil winding and transfers the design to the mechanics of the pickup assembly.

If anyone want to make a quick pickup for their acoustic guitar, this pickup technique can be easily mounted at the end of the neck in the sound hole for a cost of $2.50 for the current transformer, $15.00 of the matching transformer (Radio Shack), $6.00 3-pin mic connector, $1.00 short piece of AWG10 solid wire, $5.00 10 ft of 2-conductor mic cable, $2.00 magnet.

If anyone tries this, I would like to hear about your results.

Joseph J. Rogowski

[David Schwab]

Very cool. Thanks for sharing that.

[DrStrangelove]

Joseph, yours is an extraordinary bit of work that takes time to ponder so I hope you'll forgive my late response.

The current mode loop's simplicity lends it to broad application.
The loop can be prototyped with 1/4" copper tape.
It can be produced as a printed circuit board.

The copper could be inexpensively silver plated for minute impedance corrections when matching with a particular current sense transformer.

The xformr itself could be optimized; toroidal cores tend to have better efficiency and extended high frequency response (-3dB@100kHz? Heh!).

Buffering the current sense xformr with an active device running on phantom power is another possibility.

The loop can be on flexible media or arched to follow fingerboard curves on bowed instruments.

Cello and viola players are still looking for an alternative to piezo sensors and would gladly embrace this for the better sound quality.

Thanks, Joseph.

-drh

[bbsailor]

The xformr itself could be optimized; toroidal cores tend to have better efficiency and extended high frequency response (-3dB@100kHz? Heh!).

I have attempted to do this research using commonly available parts to help educate myself about the technical parameters necessary to optimize a current-based vice voltage-based pickup. In the process, I wanted to share this information with the forum members as the minds of many are much better than the mind of one.

Look at the transbucker sketch and note the transformer core wrapping around one thick conductor of the pickup. See that this is what is also done in the alumitone sketch (single coil shown) and the humbucking form factor (that I recently purchased) works the same way.

I have also provided a photo of the CSE187L low frequency current sensing transformer. It is commonly available from many electronic supply houses. Do a web search on the part number and check out the pdf specification sheet.

Here are the key specifications: Primary resistance: 250 micro ohms or 1/4000ohm. Secondary resistance:21 ohms although some of mine are around 18 ohms. Turns ratio: 1:500. Frequency: 50 to 400 Hz although as an input transformer not being saturated, it will go higher into the audio spectrum. Cost: $2.50 to $3.00 depending on source and quantity ordered.

The single transformer loop primary measures out at the same diameter as AWG 12 copper wire used in houshold electrical cable. However, I found that the optimum current tarnsfer from a guitar string to a single coil loop is highly dependent on the resistance of that single loop. Circuit boards may be too thin to optimize this, but could work for experimentation. The thought to use a toroidal core might work as long as the coupling between the single loop to the toroid is good enough at low frequencies where the largest losses would occur. To match the Lace design, the current sensing transformer should have a secondary in the 2000 ohm plus range with a turns ratio of 1 to several thousand (possibly 5000 or more).

Since I had a few of the CSE187L at hand, I chose to view the single AWG 12 primary transformer loop as an extension of one loop of the pickup winding that is looped around the magnet. I measured the secondary impedance of the CSE187L at 1000 Hz and is is about 400 to 600 ohms dependent on the pickup loop wire thickness and length that is connected to the transformer primary. Since this is close to the impedance of commercial audio matching transformers (Shure 95 series), I just wired a 3-pin mic connector to the transformer and went though another 500 to 50000 ohm transformer to get some more voltage boost. This is not the most efficient way to go as I now have two transformer losses and optimizations to contend with. This design, however, does give a reasonable output that is very quiet, even near a computer.

When I increased the single loop wire diameter around the magnet to AWG10, the output improved somewhat. I suspect that using copper bars in the range of .25" to .375" wide by .125" thick would improve it even more. I even made a crude humbucking design using two magnets configured like the sketch on my first post using AWG10 wire soldered together with the transformer connected to the center common loop. This provided about 50% more voltage output compared to the single coil design using the same transformer.

With this design, the construction of a pickup shifts from winding thousands of turns of wire on a core (voltaged-based design) with the attendent capacitance issues and resonance issues, to the mechanics of current-based design to match the form factor commonly evolved single coil or humbucking sizes to match standardized guitar manufacturing practices.

Once off-the-shelf parts (transformers, coils, laminated cores) are located, this design opens new possibilities to pickup makers to make new pickup designs that have an extended frequency response that can be electronically shaped to sound like any traditional voltage-based pickup. It is always easier to shape responses that are present than try to create responses that are not, with the consequence of adding noise.

Drstrangelove and David Schwab fully comprehend why I posted this new design. I hope a stimulating discussion developes and that some folks try using various transformers in a current-based pickup design to educate themselves as well.

Joseph Rogowski

[DrStrangelove]

Running some numbers, the 10ga. loop is an estimated 5.5 inches long.
and has a DCR of ~460 micro ohms.

A .25" wide 5.5" long 1.25mil copper tape loop is 120 MILLI ohms,
perhaps a bit too much for this application.

One footnote: I was an early tester for some of the transensors and found their
humbucking versions were overloaded by 500k pots, unusable without 1M pots
and even then suffered from an audibly compressed dynamic range.

Lace's second iteration produced the much-improved California Specials which,
in addition to handling convention volume/tone pots, have the overall chime of
Fender single coils albeit with some of the Lace Sensor's lack of detail due to
comb filtering from it's wider magnetic aperture.

An active design is worth thinking about, IMO.

-drh

[David King]

Hey I'm all for this design, I've been meaning to design a decent bass pickup that's thin enough (1/8") to slide around under the strings without needing a routed cavity. Cinemag makes some really nice sounding audio transformers that aren't too expensive. Here's their DI box xformer which might be appropriate as a secondary step-up.
http://cinemag.biz/direct_box/direct_box.html but they have lots of others, I just don't know what to look for exactly to fit this application. I'm pretty sure they can wind anything you want.

[bbsailor]

Originally Posted by DrStrangelove

Running some numbers, the 10ga. loop is an estimated 5.5 inches long.
and has a DCR of ~460 micro ohms.

A .25" wide 5.5" long 1.25mil copper tape loop is 120 MILLI ohms,
perhaps a bit too much for this application.

One footnote: I was an early tester for some of the transensors and found their
humbucking versions were overloaded by 500k pots, unusable without 1M pots
and even then suffered from an audibly compressed dynamic range.

Lace's second iteration produced the much-improved California Specials which,
in addition to handling convention volume/tone pots, have the overall chime of
Iender single coils albeit with some of the Lace Sensor's lack of detail due to
comb filtering from it's wider magnetic aperture.

An active design is worth thinking about, IMO.

-drh

drh

Funny you should mention the optimum load for Lace low impedance pickups. I contacted Jeff Lace via email and he said that the optimum load for the Alumitone is a 250K Volume pot with either a .02 or .05 uf cap for the tone on a 250K pot. My black Alumitone humbuckers measures out at 3.4 K ohms and 10.5 H, not the 16H that is listed on the Lace web site. It seems my Alumitone humbuckers are closer to the specifications of the Transbucker. I wish Lace would get the technical information about their low impedance pickups closer to reality.

I plan to feed these Lace Alumitones into an EMG active tone network (treble and bass boost/cut and mid shift boost/cut concentric staked tone network) that has a 200K input impedance. Based on what you say, I may need to build a Tillman active buffer FET right after my six position selector switch to prevent loading them down? These pickups were a little edgy with a 500K volume pot and calmed down some when I reduced the load to 242K by putting a 470K resistor across the selector switch output. Next week I will find out how they sound connected to my active tone network.

About your active design comment, I have one observation. It is all about signal to noise ratio. We tend to want hotter pickups to get a nice, creamy overdrive sound but must suffer the consequences of the hot pickups also acting like sensitive noise antennas and the consequences of cable capacitance. With the low Z, single loop, current based design, the noise is lower as well as the gain. This lower gain can be activly boosted because the signal to noise ratio is better and the extra active boost does not add any more noise and isolates the pickup coils/transformer from the cable capacitance and amp input circuit loading.

Thanks for your comments.

Joseph Rogowski

[David Schwab]

Originally Posted by David King

Hey I'm all for this design, I've been meaning to design a decent bass pickup that's thin enough (1/8") to slide around under the strings without needing a routed cavity.

And this after you kept telling me no one needs low Z bass pickups...

[R.G.]

Cool stuff indeed.

Random thoughts:
- a good way to set this up is to put the current-sense stepup transfer right by the string sensing loop to minimize the distance for which heavy wire is needed.
- square-section or rectangular-section copper bar or a formed rectangular-section aluminum ring would make a good primary loop. Aluminum should work fine and be much cheaper if you used a formed or cast ring so you don't have to make a joint in the primary loop. Aluminum is only slightly less conductive than copper, and large sections are cheap. Could even be cast in place or cast to shape.
- but because of the difficulty soldering, copper's what the hacker will use
- if you did the rectangular-section small loop, you could clamp ferrite stepup cores over a section of it at one end of the primary loop, something like a pair of C-cores each with their own half-secondary. A split toroid might work, but this is what C's were made for.
-if you have to move the high current loop over to the current sense stepup trannie, use a wide, flat conductor pair separated by as little distance as possible; this produces the lowest inductance path to the transformer.
- if you're going to all this trouble, you want to get to 600 ohms, balanced out of the thing and over to the console. The other end is where to put the stepup. The stepup at the end raises the impedance and voltage up to match the amp it's running into.
- there are highly linear low noise differential preamps designed for moving coil phono stages that can probably take the place of the second transformer.
- high-ratio transformers get almost as picky to wind as pickups themselves; it's good to separate the voltage gain into a couple of stages to avoid the high-ratio issues; you'll eat more losses, but have less picky transformers to wind.
- you'll have to watch what you terminate the stepup chain with. Current transformers get really pickup about the termination impedance if you want wide linear response.
- if you don't care about wide linear response, you'll be back in the insulation-color-matters issues but on the transformers, not the pickup.

[DrStrangelove]

Originally Posted by bbsailor

Funny you should mention the optimum load for Lace low impedance pickups.

That comment was about the first humbucking Transensors of ~2000.
They required more than a few detail changes to make them a drop-in replacement for ordinary pickups.

-drh

[bbsailor]

R.G.

I had to reflect on your comments for a while, but here goes!!!

Originally Posted by R.G.

Cool stuff indeed.

Random thoughts:
- a good way to set this up is to put the current-sense stepup transfer right by the string sensing loop to minimize the distance for which heavy wire is needed.

Yes! This is the way that Lace does with the Alumitones and Transbuckers. My plan is to put the current sensing transformers right under the sensing coils under the magnets. Drsrtangelove suggested that there be magnetic shielding between the sensing coil magnet and the current sensing transformer.

- square-section or rectangular-section copper bar or a formed rectangular-section aluminum ring would make a good primary loop. Aluminum should work fine and be much cheaper if you used a formed or cast ring so you don't have to make a joint in the primary loop. Aluminum is only slightly less conductive than copper, and large sections are cheap. Could even be cast in place or cast to shape.

The optimum match between the surrent sensing transformer is based on the resistane value of the primary loop matching the resistance value of teh current sensing primary loop. Beyond that there are diminishing returns.

- but because of the difficulty soldering, copper's what the hacker will use

Yes! Copper is solderable while aluminum is not. Aluminum is a good commercial alternative while copper is a great DIY material.

- if you did the rectangular-section small loop, you could clamp ferrite stepup cores over a section of it at one end of the primary loop, something like a pair of C-cores each with their own half-secondary. A split toroid might work, but this is what C's were made for.

In my next design, I plan to put one CSE187L on each end of a sensor loop that is around the magnet. This way I can see how the coupling between these transformers can be either between single, series or parallel transformer connections to the 600 ohm microphone matching transformer and monotor the output level and tonal change.

-if you have to move the high current loop over to the current sense stepup trannie, use a wide, flat conductor pair separated by as little distance as possible; this produces the lowest inductance path to the transformer.

I plan to put the the first transformer right under the pickup sensor coil

- if you're going to all this trouble, you want to get to 600 ohms, balanced out of the thing and over to the console. The other end is where to put the stepup. The stepup at the end raises the impedance and voltage up to match the amp it's running into.

Agree. This minimizes cable capacitance effects on tonal rsponse.

- there are highly linear low noise differential preamps designed for moving coil phono stages that can probably take the place of the second transformer.

Yes. I believe that the first stage of impedance conversion should be done inductivly to minimize noise. Later stages could be active.

- high-ratio transformers get almost as picky to wind as pickups themselves; it's good to separate the voltage gain into a couple of stages to avoid the high-ratio issues; you'll eat more losses, but have less picky transformers to wind.

This all gets down to signal-to-noise-ratio. With low Z pickups and the potential to provide a balanced line output. This design could take advantage of the extra gain in microphone circuits vice guitar amp input circuits.

- you'll have to watch what you terminate the stepup chain with. Current transformers get really pickup about the termination impedance if you want wide linear response.

Agree. But, unlike guitar pickups that have a 12 db per octave response drop off beyond their resonance point, low Z pickups have an extended frequency range making them ideal for electronic EQ and mimicking of traditional single coil and humbucking pickups by just rotating a knob.

- if you don't care about wide linear response, you'll be back in the insulation-color-matters issues but on the transformers, not the pickup.

Irrational talk drives scientists crazy.

I have replied to your observations below each of your comments above.

Optimum design is based on connecting important dots. What dots are important are not a scientific issues but are marketing and social issues. What sounds good is not a scientific question but a social issue. When scientists are told what sounds good and asked why particular sounds sound good, they can do a scientific analysis and provide answers. Guitar pickups fall into this same categogy. Low impedance pickups connect a particular group of dots better than traditional pickups. The real issue is: how do we connect enough of the important dots to produce a relevant alternative to traditional Hi Z pickups? Should we seek to replace them or simply offer another alternative? These fundamental questions will eventually be answered.

Thanks or your comments.

Joseph Rogowski

[R.G.]

The real issue is: how do we connect enough of the important dots to produce a relevant alternative to traditional Hi Z pickups? Should we seek to replace them or simply offer another alternative? These fundamental questions will eventually be answered.

As a practical matter, it will be crucial that you produce something that plugs directly into existing guitar amps. Otherwise, uninformed musicians simply won't buy it. That's the social reality you were talking about, I think.

Otherwise you not only have to innovate pickups, you have to innovate wiring and amps, and educate the public to unthinkingly accept your solution when '59 Les Pauls didn't do that.

Ugly, but the truth.

[bbsailor]

Originally Posted by David King

Hey I'm all for this design, I've been meaning to design a decent bass pickup that's thin enough (1/8") to slide around under the strings without needing a routed cavity. Cinemag makes some really nice sounding audio transformers that aren't too expensive. Here's their DI box xformer which might be appropriate as a secondary step-up.
http://cinemag.biz/direct_box/direct_box.html but they have lots of others, I just don't know what to look for exactly to fit this application. I'm pretty sure they can wind anything you want.

David,

I checked out the Cinemag transformers and thay are pretty expensive. My plan was to simply use the Shure A95U 500 ohm to 50K ohm transformer and a Tip-Ring-Sleve .250" stereo jack on the guitar. Then, a standard guitar amp can be used with a balance line 2 conductor microphone cable.

The output of two CSE187L current sense transformers in series is about 375 ohms. I may be able to switch the wiring in the A95U to the lower input impedance wiring setting and obtain a little higher turns ratio and a potentially higher output. That is my next experiment.

I am only interested in doing this design with inexpensive off-the-shelf parts so anyone can build it once the design alternatives are fully explored. My progress will be posted here and the results will be in the public domain so others cannot attempt to patent this work without giving full disclosure or enough of a change to make it unique or novel.

Hopefully, others will follow along and offer ideas as sort of a group design project.

Joseph Rogowski

[salvarsan]

Originally Posted by R.G.

As a practical matter, it will be crucial that you produce something that plugs directly into existing guitar amps. Otherwise, uninformed musicians simply won't buy it.

"Uninformed musicians" is a proportionately smaller group when you look beyond guitar players.

Bowed instruments, stringed harps, and the like are waiting for something like this.

-drh