Low Impedance Pickup Research

[Mike Sulzer]

As bbsailor wrote, 1H is the secondary inductance (measured with open primary loop). It is given by 500² times the AL value of the core. If the primary loop/circuit is closed, the impedance measured at the secondary is given by the secondary open loop inductance in parallel with the stepped-up/reflected primary loop/circuit impedance + the DCR of the secondary coil.

Measurements taken with my LCR meter on a different "CT" with and without a shorted primary loop in principle confirmed my theoretical considerations (based on transformer equivalent circuit, which is the same for "current" and "voltage" transformers). For instance, the inductance value at the secondary increases with the area of the shorted primary loop.

Thanks, I see. So you have a "pickup" with an inductance of about 100 nanoH looking into a transformer primary of about 4 microH, and so the inductance of the transformer is large enough so that the voltage of the pickup is not significantly attenuated.

In the case you mentioned you have about 25 mH looking back into the transformer secondary with the pickup loop connected. So if the pickup is loaded by the actual input impedance of the mic preamp (assume 2500 ohms), then the high frequency roll off of the system is at

2500/25e-3/2/pi ~ 16 KHz

But there also must be some leakage inductance that we can think of as in series with the secondary of the transformer, and so this lowers the high frequency response some more.

[Helmholtz]

Yes, that's the model, the way I see it. The 100nH outer uncoupled primary loop inductance is just a rough guess, though. Could be off by a factor of 2 or so.

The 4µH primary inductance shunts the reflected load as seen from the primary, and its magnetizing current limits bass transfer. The most effective way to improve bass response is to lower the resistance of the primary loop. The reason is that the increasing magnetizing current at lower frequencies causes a voltage drop across the loop resistance proportional to 1/f and thus reduces the available transformer primary voltage. Lower resistance decreases this effect and shifts the roll-off to lower frequencies.

Any leakage inductance just adds to the 25mH at the secondary and further lowers the upper corner frequency. A good LCR meter should be able to read the total secondary serial inductance Ls at 1kHz.

Bass response could maybe be improved a little by a lower load resistance but at the cost of worse high frequency response.


Are there any frequency response measurements of such LoZ "CT" PUs?

[bbsailor]

If you use a Triad CSE-186L current transformer with the primary three turns of AWG-16 wire removed you can insert an AWG 8 solid wire as both the primary transformer turn with an extension to become the string loop at about 53 micro ohms per inch. This method only requires a single joint to make the string loop, unlike two joints required to connect a string loop to connect to a pre-installed CT primary connection..
We all know that AWG 42 has much less area than AWG 8. AWG 8 has about between 2614 to 2662 less area (depending on whether you use inches or mm) to compare how many turns of AWG 42 would occupy the same area as AWG 8.

Packing the equivalent of about 2650 turns of very thin wire (which is the same area as AWG 8) compacted in a pickup very close to the string inducing the voltage/current in the string loop is more efficient than trying to induce the same voltage/current in a high impedance pickup with AWG 42 wire distributed about .75 inches from the top wingdings with having a higher induced voltage than the lower induced voltage in the lower winding.

The efficiency of the Current Transformer (CT) pickup depends on a new set of criteria. These now include:
1. Transformer turns ratio
2. Permeability of the CT core
3. Resistance of the CT primary with the string loop being in series
4. Diameter of the string loop wire with the skin effect taken into consideration
5. Creative use of wire and wire sizes to affect the audio effect of CT pickup design variables.

The difference between the ideal transformer theory and real transformer theory creates a filter defined by the variables discussed this forum thread.

As more of the MEF members tinker with this stuff and present their results, we will all benefit from a worldwide design evolution.

Joseph J. Rogowski

[Mike Sulzer]

Packing the equivalent of about 2650 turns of very thin wire (which is the same area as AWG 8) compacted in a pickup very close to the string inducing the voltage/current in the string loop is more efficient than trying to induce the same voltage/current in a high impedance pickup with AWG 42 wire distributed about .75 inches from the top wingdings with having a higher induced voltage than the lower induced voltage in the lower winding.

2650 turns of #42 occupies a lot less than .75 inches height if the coil has the width of a standard humbucker bobbin, and you can use a wider bobbin if you want, further reducing the height of the coil. But I think a better comparison is to a coil of 500 turns of larger wire, which is easier to work with. This would have a mic. preamp level/impedance similar to the output of the CT.

[bea]

Yes, that's the model, the way I see it. The 100nH outer uncoupled primary loop inductance is just a rough guess, though. Could be off by a factor of 2 or so.

The 4µH primary inductance shunts the reflected load as seen from the primary, and its magnetizing current limits bass transfer. The most effective way to improve bass response is to lower the resistance of the primary loop. The reason is that the increasing magnetizing current at lower frequencies causes a voltage drop across the loop resistance proportional to 1/f and thus reduces the available transformer primary voltage. Lower resistance decreases this effect and shifts the roll-off to lower frequencies.

Any leakage inductance just adds to the 25mH at the secondary and further lowers the upper corner frequency. A good LCR meter should be able to read the total secondary serial inductance Ls at 1kHz.

Bass response could maybe be improved a little by a lower load resistance but at the cost of worse high frequency response.


Are there any frequency response measurements of such LoZ "CT" PUs?

You might know that a guy in Germany (GBB forum) is currently doing such pickups - and he is trying to measure the responses.

[Helmholtz]

You might know that a guy in Germany (GBB forum) is currently doing such pickups - and he is trying to measure the responses.

Haven't seen a frequency plot yet.

Regarding bass response:

A 1:500 CT showing an open loop secondary inductance of 1H produces a 4µH inductance in the single primary loop, as mentioned above. If the wire loop has a resistance of 1mOhm, the lower corner frequency calculates to about 40Hz, using the formula in post #323.

If the copper wire has a cross section of 10mm², the loop resistance will be around 0.25mOhm. This results in a corner frequency of 10Hz, which isn't bad at all.

[bea]

resistance will be around 0.25mOhm. This results in a corner frequency of 10Hz, which isn't bad at all.

Which
a) is consistent with the impression i got from playing it with bass strings through my headphones and
b) means that we have plenty enough of tolerance in the low frequency range which will allow us to play with different (slimmer!) cross sections or material with larger specific resistivity (even brass at 6-10 mm^2 ...)

[bbsailor]

Which
a) is consistent with the impression i got from playing it with bass strings through my headphones and
b) means that we have plenty enough of tolerance in the low frequency range which will allow us to play with different (slimmer!) cross sections or material with larger specific resistivity (even brass at 6-10 mm^2 ...)

bea and all MEF members trying this,

I used a CSE-187L with a pre-wound AWG 12 primary wire and about 6.0 inches of AWG 10 making a soldered on string loop in two places and got the following results using the Extech LCR meter:

216 ohms and about 104mh at 120 Hz: 243 ohms and about 23.8 mH at 1 KHz. Output peak is about 5 mv

Then I used a CSE-186L current transformer with the pre-installed primary removed using a total of 8 inches of AWG 8 going through the transformer primary and making the string loop with only one solder/crimping sleeve joint making a low Z pickup with these results:

104 ohms and 35.6 mH at 120 Hz and 111 ohms and 20.6 mH at 1KHz. Output peak is about 10 mv.

Then I used both pickups on my two string test jig with a Low E and High E strings 25.5 inches long. There was a definite output increase of 2X using the AWG 8 wire loop observed on my oscilloscope.

A lower string loop impedance means that more current is being developed in the string loop and thus there is more output as the voltage output is a factor of current strength.

Higher output levels improve the signal to noise ratio. Just try to keep the string loop and wire going through the transformer primary as thick (lowest resistance) as possible and the wire joint as low resistance as possible to minimize the string loop impedance and maximize your output level. Try to use copper and clean any oxide off the wire joint to ensure the lowest resistance connection.

Joseph J. Rogowski

[Singer15]

I am trying to make a very thin lo impedance pickup to fit at the end of a fingerboard. The problem is fitting what is required into a form that is 5mm thick and about 40/45 x 80mm. I have attempted a variety of things, using various magnets of 3mm thickness, various plates top and bottom, and various wire gauges. I can’t get the hang of 44 gauge wire, so I gave up on the hi impedance route, and am attempting low impedance.

My parameters leave no room for an on board CT like the CSE 187 or anything that big, so I have been running through a Shure 85 onto my amp, or directly into an XLR mike input. I determined that I could make a coil with 250 loops of 32 gauge which in theory would have an inductance of 192 ohms at 4khz. I thought this might work directly into the Shure 85 xfromer, but the output was still quite low.

My form/bobbin is made of two alum plates 40 x 80 x 0.8 mm, with magnets glued between. I tried circular magnets of various types and sizes; neo 12mm x 3mm and plain craft store fridge mags 10mm x 3mm. I made the mistake of lining them up in a daisy chain so they stuck together, but later realized this meant each one was reversed polarity from the adjacent one. Nevertheless I did get some that worked, but poorly. One major difficulty is finding glue that works well. I tried super glue but that failed. JB Kwik works but it too doesn’t hold up to a screwdriver; which is good for opening up failed pickups but not promising for permanent ones.

There is a lot of theory in this thread, but much of it is beyond my electronic knowledge, so I am looking for some practical tips on how I might best make a simple loZ pickup as described here. I would like it to be 5 x 45 x 80 mm, and able to run into a standard guitar amp through a microphone matching transformer (Shure 85 etc), or possibly into an XLR mike input.

- Is this form a problem? Would plastic be better? I tried thin steel plates but found the results very microphonic. Aluminum was much less so. Maybe brass? Should the case be grounded?
- What sort of magnet is necessary? Would a bar be better? It has to be 3mm thick.
- Would I get better results with more loops of thinner gauge, or would this be defeating the purpose, even staying within lo impedance?

Thanks so much everyone.

Don Nathan

[Helmholtz]

The aluminum plates will cause extreme eddy current losses resulting in low, muffled output. Good conductors like copper,aluminum or brass are not suitable as PU construction materials.
Can we see a picture?

[bbsailor]

I am trying to make a very thin lo impedance pickup to fit at the end of a fingerboard. The problem is fitting what is required into a form that is 5mm thick and about 40/45 x 80mm. I have attempted a variety of things, using various magnets of 3mm thickness, various plates top and bottom, and various wire gauges. I can’t get the hang of 44 gauge wire, so I gave up on the hi impedance route, and am attempting low impedance.

My parameters leave no room for an on board CT like the CSE 187 or anything that big, so I have been running through a Shure 85 onto my amp, or directly into an XLR mike input. I determined that I could make a coil with 250 loops of 32 gauge which in theory would have an inductance of 192 ohms at 4khz. I thought this might work directly into the Shure 85 xfromer, but the output was still quite low.

My form/bobbin is made of two alum plates 40 x 80 x 0.8 mm, with magnets glued between. I tried circular magnets of various types and sizes; neo 12mm x 3mm and plain craft store fridge mags 10mm x 3mm. I made the mistake of lining them up in a daisy chain so they stuck together, but later realized this meant each one was reversed polarity from the adjacent one. Nevertheless I did get some that worked, but poorly. One major difficulty is finding glue that works well. I tried super glue but that failed. JB Kwik works but it too doesn’t hold up to a screwdriver; which is good for opening up failed pickups but not promising for permanent ones.

There is a lot of theory in this thread, but much of it is beyond my electronic knowledge, so I am looking for some practical tips on how I might best make a simple loZ pickup as described here. I would like it to be 5 x 45 x 80 mm, and able to run into a standard guitar amp through a microphone matching transformer (Shure 85 etc), or possibly into an XLR mike input.

- Is this form a problem? Would plastic be better? I tried thin steel plates but found the results very microphonic. Aluminum was much less so. Maybe brass? Should the case be grounded?
- What sort of magnet is necessary? Would a bar be better? It has to be 3mm thick.
- Would I get better results with more loops of thinner gauge, or would this be defeating the purpose, even staying within lo impedance?

Thanks so much everyone.

Don Nathan

Don,

Look up Shure A95U matching transformer. It has two primary settings but the default setting of 150 Ohms needs to be changed to the 75ohm setting getting about twice the output voltage. Try this 75ohm setting using 250 turns.

If if you use AWG 34 to AWG 36, you can put more turns, about 500 to 600 turns, on the pickup to better match the 150 ohm input. The 150 ohm input gets you an output level about 12 times the input level to better approximate the signal level of a single coil pickup with about 6000 turns.

Post a photo of your guitar so we can all clearly see any further restrictions or opportunities to help you solve your problem.

Building pickups is a great way to learn!

Joseph J. Rogowski

[Singer15]

Making a 5mm thick pickup

Don,

Look up Shure A95U matching transformer. It has two primary settings but the default setting of 150 Ohms needs to be changed to the 75ohm setting getting about twice the output voltage. Try this 75ohm setting using 250 turns.

If if you use AWG 34 to AWG 36, you can put more turns, about 500 to 600 turns, on the pickup to better match the 150 ohm input. The 150 ohm input gets you an output level about 12 times the input level to better approximate the signal level of a single coil pickup with about 6000 turns.

Post a photo of your guitar so we can all clearly see any further restrictions or opportunities to help you solve your problem.

Building pickups is a great way to learn!

Joseph J. Rogowski

Thanks - more thoughts on that: I bought the Shure A85 thinking it was similar to the A95U, but it isn't! That siad, it certainly get you into the ballpark.
I've made 2 versions that work reasonably well. Both use aluminum plates with 6 button magnets (from a craft store) of 10mm diameter. This setup fits perfectly under the strings.

Version 1: 230 winds of 32 ga wire (about 110 ft). Into XLR plug the output is low but usable. Through the A85 it's higher. Sound is clean, except for some hum. E and B strings much louder. To reduce this I placed a small piece of thing steel to partially screen the coil on one side beneath the two high strings. This brought the outputs into better balance.

Version 2: I first wound this using steel plates, but they were very microphonic and created some odd noises, so I removed the plates and sandwiched the coil between 2 thin sheets of aluminum flashing stock, then taped it all up. It has 430 winds of 32 ga wire and a resistance of 35k. Similar results to version 1, but louder.

If I switch to 38 gauge, how many windings would it take to increase the output 50%? Is there a dependable way to calculate this? I've been using a website that calculates rectangular coil impedance here: https://www.eeweb.com/tools/rectangle-loop-inductance. I use 4khz to calculate Z. Is this the standard?

If I wind two coils, using 3 magnets per coil and offset them, would this be a humbucker if wound in opposite directions? I think getting rid of the hum would be desirable!

[bbsailor]

Don,

A few more facts about your situation will be helpful.

1. Please post a photo of your pickup area without a pickup so we can better see the hole and fingerboard end area more clearly.
2. What is the string width at the end of the fingerboard and right over the round hole where the end strings just pass over the hole?
3. When you press the end strings on the last fret, what is the distance of the lower part of the string to the guitar top?
4. What type of strings, acoustic or electric, and string gauge are you using?
5. What gauge wire do you have in stock?
6, Are your strings grounded?
7. Can you access electronic mail order companies?

Thanks

Joseph J. Rogowski

[Singer15]

Don,

A few more facts about your situation will be helpful.

1. Please post a photo of your pickup area without a pickup so we can better see the hole and fingerboard end area more clearly.
2. What is the string width at the end of the fingerboard and right over the round hole where the end strings just pass over the hole?
3. When you press the end strings on the last fret, what is the distance of the lower part of the string to the guitar top?
4. What type of strings, acoustic or electric, and string gauge are you using?
5. What gauge wire do you have in stock?
6, Are your strings grounded?
7. Can you access electronic mail order companies?

Thanks

Joseph J. Rogowski

Hi Joseph,

In response to your questions:

1. The problem regarding soundholes is that I have some guitars without them, so this pickup will have to be thin enough to fit between the strings and the top.
2. String width at end of fingerboard = 50mm
3. String ht = 12mm, down to 9mm when using the higher frets.
4. I am using Daddario gypsy strings which have silver plated copper windings, gauges from 10 to 44 thou.
5. Not much left except a roll of 44ga which I have had no luck with!
6. No grounding on acoustic guitars generally. Could be done however, since these all have metal tailpieces.
7. Yes, in fact that is how I will have to get my wire in future, having exhausted the supply in town here.

To add some more experimental results to this, today I wound 2 split coil pickups with 44 ga wire, with mixed results, generally poor. The first pickup had 135 windings per coil, the second had 500. I used plastic for the plates between which I glued the magnets.

The first one was definitely an improvement regarding hum cancelling, compared to the previous single coils, but it was very low output. I used the same 10mm x 3mm ceramic magnets from a dollar store, one set of 3 for the 456 and one set for the 123 strings, reversed windings, and opposite polarity.

The second try was not so good, and I couldn't get a good DC resistance reading, although the pickup actually produced a signal. I though I might have it wired out of phase but even switching the coil ends did nothing to improve the signal.

I suspect that heavier wire is the way to go here, the question is what gauge next? Also, would more powerful magnets improve the signal? These little ceramic magnets seem about as strong as an alnico bar magnet, judging by how well they stick to things.

Cheers,

Don Nathan

[bea]

Don, unfortunately the current transformer pickups are difficult to do that thin. It might be possible under the strings but outside where the CT would need to be in this case, the thickness would be close to 10 mm.

I would suggest doing sidewinders: a thin horizontal blade of steel, a layer of insulation, a central row of magnets, and two coils windings of relatively thick wire adjacent to the magnets and around the halves of the blade.

AFAIK, someone has successfully done that and shown it here in the forum. Even thinner than 5 mm.