Will there ever be anything else?

[DrStrangelove]

I totally agree with you guys. I don't know how in hell anyone ever makes money by selling things to musicians, though I suspect like PRS, the business plan involves selling beautiful expensive things to lawyers and dentists.

Sam Kennison profanely remarked on the starvation in Ethiopia, "Go where the f*ckin' food is, ya monkies!"

Same thing with selling pickups -- go where the money is.

Just remember that smart monied people often turn into
vintage-fetishist silly people when confronted with a novel guitar feature.
I will not name examples since they include a lot of my friends, only mention
that I still get ribbed for a couple of unusual custom-made guitars I own.

So, the real question is whether or not an electric guitar is an intelligence attenuator.

-drh

[DrStrangelove]

New ideas take time to become adopted. They have historically needed to reach some sort of critical mass as driven by:
accepted technology trends
economical alternative to what is commonly used
clear performance benefits
easier/cheaper manufacturing processes
fads and irrationality
public policy
tax incentives
emulation of respected people
effective marketing

If put in that perspective, variable reluctance guitar pickups are a 70+ year old
buggy whip long overdue for cremation, burial, and a beer-soaked wake followed
by urination on the ashes.

The soaring prices of copper and cobalt coupled with world control of
neodymium sources by Chinese interests will make an unpleasant time
for pickup makers.

As I reckon, there's still time to get a grant for pickup-making as a faith-based
abstinence-only anti-terrorism initiative, but the next administration won't be
so ideologically blinkered.

That said, I find your current mode pup prototype rife with good possibilities.

-drh

[bbsailor]

Current Mode PUP Possibilities

If put in that perspective, variable reluctance guitar pickups are a 70+ year old
buggy whip long overdue for cremation, burial, and a beer-soaked wake followed
by urination on the ashes.

The soaring prices of copper and cobalt coupled with world control of
neodymium sources by Chinese interests will make an unpleasant time
for pickup makers.

As I reckon, there's still time to get a grant for pickup-making as a faith-based
abstinence-only anti-terrorism initiative, but the next administration won't be
so ideologically blinkered.

That said, I find your current mode pup prototype rife with good possibilities.

-drh

drh,

Currently manufactured voltage-based pickups use commonly available parts. 3/16" diameter to 1/4" diameter magnets of various vintages and wire in the AWG 28 (low Z) to AWG 43 range (high Z) magnet wire is commonly used to wind on bobbins either purchased as molderd plastic bobbins or crafed from fiberboard into pup bobbins.

When we start talking about current based pup designs, we can take two approaches.

1, We can take the Lace approach and attempt to make the secondary of the current sensining transformer as close to the high Z circuits using 250K to 500K pots and high Z amp inputs.

or

2. We could take the Les Paul recording guitar approach and plan to put out a 600 ohm to 1000 ohm balanced output with the final matching being done at the amp to minimize line capacitance and the consequences of high Z circuits in noisy environments.

If we take the approach of using commonly off-the-shelf parts, we now must look for an efficient way to couple the high current low impedance primary loop of the pickup primary coil under the string to a boosted output with a turns ratio between 1-to-500 to 1-to-1000 exiting the guitar at about 600 to 1000 ohms.

What is needed is a source of audio inductors that can operate in the 40 Hz for bass or 80 Hz for Guitar to 10 KHz. Some may argue that a more restricted frequency range could be acceptable.

What would be ideal for the current-based pickup builder would be access to a toroid inductor with a center core opening capable of accepting an AWG 8 or 6 wire that would allow one or two toroids to be placed on primary loop.

Does anyone know of any sources for audio toroid inductors that meet this criteria? I have been searching but have not found anything except the CSE 186L current sensor transformers which is not a toroid.

If the high Z output approach is taken, a toroid with 2.5 to 3.5K ohms is needed to match high Z pots and amp inputs. Does anyone know a source for an audio toroid inductor like this?

To have an output impedance of 600 to 1000 ohms would require something between 18 to 30 ohms of resistance depending on the nature of the core material. Once off-the-shelf parts are located at a reasonable cost, much experimentation could be easily done by forum members to optimize this current-based design. With the rather wide, flat bandwidth of this design, electronic EQ and optimization could mimic any traditional voltage based pickup with a classic resonant hump providing the classic tonal flavor of the Stratocaster, Gibson Humbucker, etc, or whatever with the turn of a midrange EQ knob.


Another design uses the strings as the active transducer elements and requires that the guitar designer integrate a few features to enable this design.

1. The nut is the common ground end of the string and must be brought back to the body in a low resistance way. Connecting a brass nut to the truss rod with a metal collar and extending the truss rod into the body allows this design to be implemented easily.

2. A tune-o-matic bridge with the nyon intonation inserts is required to keep the strings electrically insulated form each other.

3 The metal tailpiece needs to be modified by drilling out each string hole to insert in insulating collar that keeps the metal string from contacting the metal of the tailpiece. This allows the ball end electrical connection of each string to be led to the control compartment where the individual transformers are mounted and connected to the truss rod common ground return to convert the low Z of each of the 6 string loops to a more usable voltage level. Once this is done the balance of each string could be done by impedance matching or with a trim pot.

4. String mixing could be done insider the guitar or individual string outputs could be sent out via a 7-pin connector (one common ground and 6 hots) or an 8-pin connector where the mixed signal is combined with the individual string outputs. This is also a MIDI pickup design.

5. Potential noise could be reduced by alternating the phase of each adjacent string so any noise induced in the strings will neutralize out in the transformer connections (like humbucking).

I have been working on pickup designs for over 45 years. I am convinced that the time has come for some different design approaches. Optical designs may be another way to go? I am open to hear all comments.

Joseph Rogowski

[Old Tele man]

I am convinced that the time has come for some different design approaches.
Joseph Rogowski

...a feeling held by many people, before and after, you, unfortunately. Design isn't everything, marketing and "timing" are (unfortunately) often more important and critical to any success.

[DrStrangelove]

Design isn't everything, marketing and "timing" are (unfortunately) often more important and critical to any success.

That's why you ignore the electric guitarists --
they're late adopters in a crowded market, and you
have the added expense of drool-proofing the pickup.

Bowed instrument and hammer dulcimer players should get the first
whack at a new design because you seldom think of them as using one.

-drh

[Steve Conner]

bbsailor, sounds like a job for the impedance matching transformers made for ribbon mics. Just don't look at the price!

Why in the world does it need to be a toroid, though? Lace pickups used an E-I core transformer IIRC, and ribbon mic matching transformers are E-I too. No manufacturer wants to be bothered winding 5000 turns of #40 onto a toroid, unless there's promise of a serious market.

[DrStrangelove]

Sheep Shagging

bbsailor wrote: [a whole buncha stuff]

Posit the design goal of Low-Z output.

By audio transformer, I assume you mean one with a v.low-loss core
such as permalloy or supermalloy, preferably tape-wound but sintered
powder still beats most silicon steels (except nanocrystallines). This
core type is also used in pulse transformer/inductor for switching
power supplies and is commonly available. You can also get a few
hits from a "flux-gate sensor" search.

WRT the sensor loop: how closely must it match the transformer primary?
Can we be extravagant and use 5-6" of .125x.250" bar stock?

The transformer should be shielded from the sensor magnet.
Either it has its own enclosure or the sensor loop has a barrier
between them.

-drh

[bbsailor]

Transformer Quest

bbsailor, sounds like a job for the impedance matching transformers made for ribbon mics. Just don't look at the price!

Why in the world does it need to be a toroid, though? Lace pickups used an E-I core transformer IIRC, and ribbon mic matching transformers are E-I too. No manufacturer wants to be bothered winding 5000 turns of #40 onto a toroid, unless there's promise of a serious market.

This potential design is not limited to a toroid inductor, however it is much easier to slip the large single turn loop through the open core. If there were a E type core that had enough room to repurpose with a larger AWG 6 to AWG 8 wire, that could be used also. I suspect that a laminated E core would be more efficient at lower frequencies. The CSE187L has a single loop of AWG 12 that could be removed and replaced by integrating the single loop AWG 12 wire through it. I found that if I increased the wire diameter, I obtained slightly better output but I had to wrestle with copper tubing to make butt connections to the transformer AWG 12 primary wire. An E core with room for an AWG 6 to AWG 8 and a 25 to 30 ohm secondary resistance (about 600 ohm impedance) would be an ideal candidate.

I was hoping I could stumble upon some off the shelf parts that might work without too much expense. The CSE187L current sensing transformers are a good experimental find at about $2.50 each. However, ribbon microphone transformers are very expensive. I did some of my early research over 30 years ago using a ribbon microphone transformer alligator clipped across a guitar string using a hand held magnet and got a reasonibly good sound that had more acoustic string properties than electic guitar properties. That is what started me on the low Z pickup quest. It is interseting that I have not seen very much progress in 30 years except the Lace low Z designs, Les Paul low Z pickups and the Epiphone Jack Casady Bass.

If I could find a toroid inductor that could operate in the lower audio range with about 500 turns at about 600 ohms impedance, that would be a good start. Maybe no such animal exists? It might be an existing off the shelf product that can be repurposed for these experiments. Hopefully, tapping the minds of the pickup braintrust can produce some suggestions. I have done the leg work and set up some rough coil/transformer parameters that experimenters could use and obtain useful results.

All you need to do is ensure that the transformer/inductor has an open center or an E core space that can accomodate an AWG 6 to AWG 8 solid copper wire that is the single loop going around the magnet being placed under the stings. Even if the output is not as strong as a traditional pickups with 6 to 8 thousand turns of AWG 42, it does have a potentially better signal to noise ratio, being a current-based balanced output and therefore could directly feed a balanced microphone input to obtain the extra amplification while remaining quiet.

Further, if multiple toroid coils or E-core transformers could be used, some interesting tonal variations could be obtained by switching transformer outputs in series or parallel, or using one transformer to connect a variable capacitance tonal network.

As drh suggested, this design could satisfy instrument players where there is limited space under the strings or unusual string angles that make conventional pickups impractical to use.

Joseph Rogowski

[DrStrangelove]

Steve,

The toroid has lower magnetic loss than the EI core, still less
if you make it out of fancy alloys. It's easier to get a toroid
that you can slip AWG 6 wire through, too.
A more efficient coil core means we can use weaker magnets
for lower string pull.

OTOH, premature optimization is risky.

Joseph:

Triad Magnetics' spec sheet says the CSE187 primary is only
large enough for AWG 12 = .08" dia. In that light, the 250 micro ohm
max resistance is overstated, perhaps to account for solder tinning resistance.

The CSE187's 500 turn/21ohm secondary suggests 26-28AWG if you
guesstimate the initial core as .25" square.

The 'worst' case, a single layer wound .25" square toroid,
corresponds to a 2" inside diameter.
Two layers #28 on a 1" ID core is a more useful size, I suspect.

The sensor loop could be made from sheet copper The DC resistance
of 6inches of 1/8" wide 16 gauge copper is close to that for
AWG 10 wire, ~500 micro ohms.

-drh

[bbsailor]

I found a candidate sensing coil

Steve,

The toroid has lower magnetic loss than the EI core, still less
if you make it out of fancy alloys. It's easier to get a toroid
that you can slip AWG 6 wire through, too.
A more efficient coil core means we can use weaker magnets
for lower string pull.

OTOH, premature optimization is risky.

Joseph:

Triad Magnetics' spec sheet says the CSE187 primary is only
large enough for AWG 12 = .08" dia. In that light, the 250 micro ohm
max resistance is overstated, perhaps to account for solder tinning resistance.

The CSE187's 500 turn/21ohm secondary suggests 26-28AWG if you
guesstimate the initial core as .25" square.

The 'worst' case, a single layer wound .25" square toroid,
corresponds to a 2" inside diameter.
Two layers #28 on a 1" ID core is a more useful size, I suspect.

The sensor loop could be made from sheet copper The DC resistance
of 6inches of 1/8" wide 16 gauge copper is close to that for
AWG 10 wire, ~500 micro ohms.

-drh

The CSE187L looks more like it has AWG 34 to AWG 36 wire by eyeballing it next to a piece of AWG 30. The transformer secondary looks about half the diameter of AWG 30.

I found a toroid current sensor by "Falco Electronics" in particular the CST013-01 for 50/60 Hz that has a 1000:1 turns radio with 81 ohms in the secondary. The physical size is 0.670" wide by 0.8" tall and 0.390" thick with a hole that has a 0.180" diameter. I need to order some samples and see how they work.

The inductance of the secondary is 1800 mH measured at 1KHz. Low frequency coupling will be the real challenge. Rather than a U-shaped loop made through the primary, I may need to make one complete turn for a little better coupling. I will probably even try putting one toroid on each end of the main pickup single loop to see how I can maximize coupling the output.

Joseph Rogowski

[bbsailor]

Please explain why the transformer should be magnetically shielded from the sensor magnet? It is easy to do, but I never thought it might be an issue.

You minimize DC bias on a transformer primary to avoid saturating the core prematurely.
At the other end of the power spectrum, a significant static magnetic bias in the core can boost the noise level undesirably. Done cleverly, it's called a flux gate magnetometer, but I digress.

The Khaganov patents (5767431,5831196) put a magnetic field barrier directly under the loop+magnet assembly "to minimize secondary coil sensitivity thereof to extraneous environmental electromagnetic flux" (i.e., it hums less).

Lace leaves it in his Transensors, to judge from the vivisections I performed.

My point is that the magnetic shield is worth keeping in mind, but may be unnecessary if we are clever/lucky.

-drh

[bbsailor]

I will probably even try putting one toroid on each end of the main pickup single loop to see how I can maximize coupling the output.

Joseph Rogowski

I tried two CSE187L low frequency current sense transformers (1:500 turns ratrio) at each end of the low impedance string sensing loop. I posed a photo and description in a new thread "More Low Impedance Pickup Research". Placing the output of the two transformers in series did increase the output to a level that would be useable by directly feeding it into a 500 ohm to 50K ohm matching transformer (1:10 turns ratio). It seems that the total equivalent turns ratio is 500 + 500 X 10 or 10000, not counting transformer or coupling losses. But, by keeping the single string sensing loop close to the magnet, more of the string induction is put into the single loop coil compared to having thousands of AWG 42 wire turns being farther from the strong magnetic field.

The ease of connecting these parts should make experiments very fast to try.

Joseph Rogowski

[bbsailor]

The transformer should be shielded from the sensor magnet.
Either it has its own enclosure or the sensor loop has a barrier
between them.

-drh

drh

I made a magnetic barrier per your suggestion and by looking at the Khaganov patents. I cut a piece of metal 1/32" thick, 1" wide and 2 1/16" long, using a Dremel tool and cut-off wheel. I shaped this into a U-shaped channel with the side walls being 1/8" high and being 5/8" between the parallel short walls. When I put this under the magnet, it improved the output as it focused the magnet field in one direction toward the strings. I kept the top of the channel walls at the same height as the top of the magnet and string sensing wire loop.

If the transformers are placed under the string sensing coil, this magnetic shield will also prevent the transformers cores from being magnetized.

Thanks for your help.

Joseph Rogowski