Low Impedance Pickup Research

[jakez]

Bbsailor

Thanks a lot for all the great information you freely shared in this thread.
I was really impressed when I discovered your first post some months ago.
You really made me want to do it.
So, I did...and my new removable guitar pickup works just fine .
It is made with:

- A piece of 4mm2 insulated copper wire.
- One AS 104 1/500 transformer.
- 9 magnets NdFeB grade 25 20x10x1mm : 3 should have be enough for electroacoustic purpose, I think. But magnets are also used to fit the pickup on the guitar by pinching the soundboard between its two parts, so I needed some strength.
- Stereo jack socket and plug, stereo shielded cable and XLR.
- Some pieces of wood and cork. for a vintage look...

I measured the impedance : Z= 244 Ohm.
The pickup is connected to a mixing board with a standard 6 meters XLR balanced microphone cable. Impedance of a board input is 3kOhm so it's perfect.

I describe the microphone making, the tests and measurements on my website here.

http://http://violons.jsld.fr/index.php/electroacoustique/un-micro-guitare-basse-impedance

Thanks again Joseph.

[bbsailor]

Bbsailor

Thanks a lot for all the great information you freely shared in this thread.
I was really impressed when I discovered your first post some months ago.
You really made me want to do it.
So, I did...and my new removable guitar pickup works just fine .
It is made with:

- A piece of 4mm2 insulated copper wire.
- One AS 104 1/500 transformer.
- 9 magnets NdFeB grade 25 20x10x1mm : 3 should have be enough for electroacoustic purpose, I think. But magnets are also used to fit the pickup on the guitar by pinching the soundboard between its two parts, so I needed some strength.
- Stereo jack socket and plug, stereo shielded cable and XLR.
- Some pieces of wood and cork. for a vintage look...

I measured the impedance : Z= 244 Ohm.
The pickup is connected to a mixing board with a standard 6 meters XLR balanced microphone cable. Impedance of a board input is 3kOhm so it's perfect.

I describe the microphone making, the tests and measurements on my website here.

http://http://violons.jsld.fr/index.php/electroacoustique/un-micro-guitare-basse-impedance

Thanks again Joseph.

Jakez,

Thank you for your kind comments about this thread.

When making a low impedance for an acoustic guitar, obtain a good balance between the plain strings and wound strings with thinner magnetic cores can require some more experimentation. Here are some diffferent things to try.

1. Try using CT cores designed to operate at about 60 to 400 Hz.
2. Look up skin depth of the solid wire you are using and experiment with stranded wire bundles or multiple parallel string loops made from thinner strands of magnet wire like AWG 16 to conduct more high frequencies to the wire core since lower resistance to higher frequencies makes more current at these frequencies and produces more output. This is a balancing act to measure and listen to. Let your ears be the final judge but measure what sounds good to better understand all the variables involved.
3. Visit an electronics supplier called “Electronic Goldmine” and look at their part number G22583 5/32 inch diameter wire braid. Use this as the string loop going through the toroid CT and bound together with a copper tube and silver solder. This wire has many strands of AWG 44 and is equivalent to AWG 10 but is less sensitive to the skin effect. If using this wire on a Triad CSE 186L with the 3 turn primary removed, tape the two outer laminated cores next to the primary opening to prevent shorting out the raw wire primary and string loop wire against the transformer frame.
4. Low noise allows less output from lower turn CTs to be amplified more and sound clean. Make sure you ground the string loop to the CSE series CT metal frame and connect this to the two conductor shielded wire ground or pin 1 of the XLR.

If you use a 750 turn CT you will need to use a primary and string loop with about twice the wire diameter area of the 500 turn CT to maintain a good low impedance. If you use a 1000 turn CT you will need 4 times the wire diameter area of the 500 turn CT.

Getting you peak output near 10 millivolts is a good level to seek through your design process.

Do you have an English translation of your web site?

Thanks

Joseph J. Rogowski

[jakez]

Jakez,

Thank you for your kind comments about this thread.

When making a low impedance for an acoustic guitar, obtain a good balance between the plain strings and wound strings with thinner magnetic cores can require some more experimentation. Here are some diffferent things to try.

1. Try using CT cores designed to operate at about 60 to 400 Hz.
2. Look up skin depth of the solid wire you are using and experiment with stranded wire bundles or multiple parallel string loops made from thinner strands of magnet wire like AWG 16 to conduct more high frequencies to the wire core since lower resistance to higher frequencies makes more current at these frequencies and produces more output. This is a balancing act to measure and listen to. Let your ears be the final judge but measure what sounds good to better understand all the variables involved.
3. Visit an electronics supplier called “Electronic Goldmine” and look at their part number G22583 5/32 inch diameter wire braid. Use this as the string loop going through the toroid CT and bound together with a copper tube and silver solder. This wire has many strands of AWG 44 and is equivalent to AWG 10 but is less sensitive to the skin effect. If using this wire on a Triad CSE 186L with the 3 turn primary removed, tape the two outer laminated cores next to the primary opening to prevent shorting out the raw wire primary and string loop wire against the transformer frame.
4. Low noise allows less output from lower turn CTs to be amplified more and sound clean. Make sure you ground the string loop to the CSE series CT metal frame and connect this to the two conductor shielded wire ground or pin 1 of the XLR.

If you use a 750 turn CT you will need to use a primary and string loop with about twice the wire diameter area of the 500 turn CT to maintain a good low impedance. If you use a 1000 turn CT you will need 4 times the wire diameter area of the 500 turn CT.

Getting you peak output near 10 millivolts is a good level to seek through your design process.

Do you have an English translation of your web site?

Thanks

Joseph J. Rogowski

Thanks for your answer, Joseph.

I fully agree. To obtain a good balance between all the strings is the point.
As I understand it, the use of stranded wire bundles could have a result, due to skin effect. It lowers the resistance for high frequencies and so, increases the output for these frequencies.
Actually my homemade micro balance is more on the high frequencies side and I think I don't need more. What my ears tell me is confirmed by the measurements. I was not surprised as Lenz law naturally boosts high frequencies and harmonics. In addition when you listen an acoustic instrument unplugged you hear the soundboard and soundhole vibrations which are not very well captured by the magnetic sensor! You feel it as a lack when the guitar is plugged. Practically it remains possible to set and get the "good" balance with the mixing board tone control.

Thus, I'd rather see a possible improvement in a thicker primary loop. It would have an influence mainly on low frequency if skin effect cancel the influence of the modification on high frequencies.
But this modification needs another CT. It could be a good idea because AS 104 is not designed for low frequencies. But, doing so, you get also bulkier components . And AS 104 is easily available and particularly affordable ...
I will think about it. for the next pickup I will make.

Up to now I have not translated my website.
I will also think about it .

[bea]

And finally a prototype also from me:







CT is an AX-500, specified for 50/60 Hz. Wire 10 mm^2 solid copper. Magnet: ceramic, taken from a defective Gibson TBPlus pseudo-mudbucker.
Directly plugged into the input of my Allen&Heath ZED-10, Headphones Sennheiser HD-650. The output level of the pickup is just fine - as i hoped. No noice audible through the headphones.

The guitar is tuned in fifths - FCGDAE - and covers the whole range of a guitar and a bass guitar. Strings: Daddario Chromes. Scale 62.8 cm (Gibson-alike). Sound: just perfect. Really fine and clear treble range and good balance over the whole tonal range. A bit week in the deep bass range, which is actually good for this instrument but not necessarily for a bass guitar. Mid range a lot better balanced than that of the 2nd PU - a Tele Neck with 500 turns of 0.2 mm wire.


I might build a 2nd prototype using a 2nd CT in the hope to be able to control the bass range.

Finally a disadvantage: the AX-500 is really huge and fat and difficult to hide, especially on an archtop guitar (soundwise the thing would be ideal on a good acoustic archtop...)

[Joe Gwinn]

Finally a disadvantage: the AX-500 is really huge and fat and difficult to hide, especially on an archtop guitar (soundwise the thing would be ideal on a good acoustic archtop...)

Twist the end of the copper loop so the current transformer can lay flat?

[bbsailor]

And finally a prototype also from me:







CT is an AX-500, specified for 50/60 Hz. Wire 10 mm^2 solid copper. Magnet: ceramic, taken from a defective Gibson TBPlus pseudo-mudbucker.
Directly plugged into the input of my Allen&Heath ZED-10, Headphones Sennheiser HD-650. The output level of the pickup is just fine - as i hoped. No noice audible through the headphones.

The guitar is tuned in fifths - FCGDAE - and covers the whole range of a guitar and a bass guitar. Strings: Daddario Chromes. Scale 62.8 cm (Gibson-alike). Sound: just perfect. Really fine and clear treble range and good balance over the whole tonal range. A bit week in the deep bass range, which is actually good for this instrument but not necessarily for a bass guitar. Mid range a lot better balanced than that of the 2nd PU - a Tele Neck with 500 turns of 0.2 mm wire.


I might build a 2nd prototype using a 2nd CT in the hope to be able to control the bass range.

Finally a disadvantage: the AX-500 is really huge and fat and difficult to hide, especially on an archtop guitar (soundwise the thing would be ideal on a good acoustic archtop...)

Bea,

Nice prototype design.

Here are a few things to consider. If you extend the control area cutout a few CM toward the strings, you can then embed the Current Transformer in that hole and just extend the control plate to cover this extended area.

Here is a way to control tonal balance. Find magnet wire that will fit in the CT opening with the heavy wire you are using. Make 3 or 4 string loops that you can lay on top of the heavy wire connected as separate parallel loops. Since they are mounted on top of the heavy wire they are closer to the strings and you will hear a different tonal balance. Experiment with using double stick tape to hold these thinner parallel string loops directly on the heavy string loop, 1mm and 2mm above the heavy string loop and listen to the sound difference.

If you are planning to put passive volume and or tone controls use the rule of 50 to determine the minimum pot value of the volume control. Multiply the dc resistance of the CT by 50 to load the CT without too much loss. The Les Paul low z pickups use a 2.5 K volume pot. Look up the wiring for the Les Paul recording pickups to see how to wire your current transformer pickup.

What output level are you now getting from the neck low z current transformer pickup?

When you tinker with making these low z CT based pickups and start changing the variables, you will start hearing differences that are not published due to its more modern design to target low z XLR input impedance ranges without the traditional resonant hump of high z pickups.

Joseph J. Rogowski

[bea]

That prototype will not be finally mounted in this guitar... it is a true prototype. But maybe i'll try to modify it a bit: try to bend the loop and bring the secondary below the primary.

Output level, impedance: i did not measure anything (and DC resistance is meaningless). But he output level is just fine to drive a microphone input. On a similar level as the output of the other pickup, which is also LoZ (500 turns of 0.2 mm wire).

[jakez]

And finally a prototype also from me:




CT is an AX-500, specified for 50/60 Hz. Wire 10 mm^2 solid copper. Magnet: ceramic, taken from a defective Gibson TBPlus pseudo-mudbucker.
Directly plugged into the input of my Allen&Heath ZED-10, Headphones Sennheiser HD-650. The output level of the pickup is just fine - as i hoped. No noice audible through the headphones.

The guitar is tuned in fifths - FCGDAE - and covers the whole range of a guitar and a bass guitar. Strings: Daddario Chromes. Scale 62.8 cm (Gibson-alike). Sound: just perfect. Really fine and clear treble range and good balance over the whole tonal range. A bit week in the deep bass range, which is actually good for this instrument but not necessarily for a bass guitar. Mid range a lot better balanced than that of the 2nd PU - a Tele Neck with 500 turns of 0.2 mm wire.


I might build a 2nd prototype using a 2nd CT in the hope to be able to control the bass range.

Finally a disadvantage: the AX-500 is really huge and fat and difficult to hide, especially on an archtop guitar (soundwise the thing would be ideal on a good acoustic archtop...)

Hi Bea

I'm happy to see your microphone and to know that it works fine. It's very interesting to read all the details you gave us.

When I saw your AX-500 I found that it seemed very similar to my AS 104, bulky but actually no more than CSE 187L
The data sheets show indeed a lot of common features, if I don't mistake (and I was surprised by this):
- the same company: Talema,
- exactly the same sizing, 20,4x17,2x9,53
- the same turn ratio: 1/500
- same isolation 2,5kV, roughly same primary current range 15A, same color...
The only difference I see is winding resistance 25 Ohm for AS AS104 (I measured 22) and 32 Ohm for AX-500, which means thinner wire but not a lot.
The feature we don't know is the magnetic material used in each.

And they specify 50-60HZ for AS104 and 20kHz to 200kHZ for AX-500.
I'd be happy to understand how such huge difference is possible with such similar features.

Is somebody, in the forum, knowing the answer ?

Jakez.

[bea]

And they specify 50-60HZ for AS104 and 20kHz to 200kHZ for AX-500.

the other way around: the AX-500 is 50-60 Hz, and the AS104 20-200 kHz.

I'd be happy to understand how such huge difference is possible with such similar features.

A question i asked myself as well. Maybe the properties of the ferrite core could differ? Or they are both much broader and only labeled for the specific use?

[bea]

Some thoughts on extending the bass response:

the lower frequency limit of a transformer depends on L/R. So we can either increase L or decrease R if we want more bass.

Increasing L seems to be be done by Lace: at least in some of their Alumitones there is some laminated steel immediately below the frame, probably close enough to have an effect on L.

Now back to my construction. An easy way to increase L is to do a few more windings on the primary.

Let's assume 2 windings: L doubles, but also R doubles, even if we can keep the same cross section. No improvement; we would have to compensate for the increased length.

So let's assume 3 windings: L increases by a factor of 9. R would increase by a factor of 3 if the cross section of the wire could kept constant, and thus the lowest frequency will go down by more than an octave.
(I expect the sensitivity of the primary to increase and therefore to compensate for the loss due to the smaller windings ratio)

Main problem: the holes in the CTs have a cross section of slightly less than 20 mm^2. Feeding 3 loops of 10 mm^2 wire through it is impossible; the limit should be somewhere near 4 mm^2. Mechanically reducing the cross section near the CT with, e.g. a file? And how to put insulation on the wire? Some kind of laquer? Cellulose laquer? Nail polish?

EDIT: a 4 mm^2 wire has a diameter of 2.26 mm. That would definitely fit in three turns through the hole in the CT. (Four turns would be impossible.) And the ratio L/R is larger than that of my single loop, so the bass response will be stronger.

Next thinner wire: 2.5 mm^2. Diameter 1.6 mm. We would at least need 5 windings to obtain a larger ratio of L/R. Which should be possible, maybe even 6 turns.

So : maybe the optimum performance of such an approach is actually reached by a few loops instead of only one?

So what about iron in the core instead of magnetic ferrite? How would that affect the inductivity?
I could imagine something like a mix of the two LoZ approaches: a few windings of really thick wire, say, 1.5-6 mm^2, around a traditional pickup core and then use a suitable CT to obtain either microphone or line sensitivity. What Do You think?

[Mike Sulzer]

Two turns makes increases the L by four times if all the flux is in common.

If you are going to use multiple turns, you could also use enough so that you get sufficient voltage to feed a mic. level input without the transformer at the pickup. Not sure if this saves space our wastes space, but it is something to consider. Of course, Les Paul like this approach long ago, but that was a bigger pickup not suitable for your application.

Some thoughts on extending the bass response:

the lower frequency limit of a transformer depends on L/R. So we can either increase L or decrease R if we want more bass.

Increasing L seems to be be done by Lace: at least in some of their Alumitones there is some laminated steel immediately below the frame, probably close enough to have an effect on L.

Now back to my construction. An easy way to increase L is to do a few more windings on the primary.

Let's assume 2 windings: L doubles, but also R doubles, even if we can keep the same cross section. No improvement; we would have to compensate for the increased length.

So let's assume 3 windings: L increases by a factor of 9. R would increase by a factor of 3 if the cross section of the wire could kept constant, and thus the lowest frequency will go down by more than an octave.
(I expect the sensitivity of the primary to increase and therefore to compensate for the loss due to the smaller windings ratio)

Main problem: the holes in the CTs have a cross section of slightly less than 20 mm^2. Feeding 3 loops of 10 mm^2 wire through it is impossible; the limit should be somewhere near 4 mm^2. Mechanically reducing the cross section near the CT with, e.g. a file? And how to put insulation on the wire? Some kind of laquer? Cellulose laquer? Nail polish?

EDIT: a 4 mm^2 wire has a diameter of 2.26 mm. That would definitely fit in three turns through the hole in the CT. (Four turns would be impossible.) And the ratio L/R is larger than that of my single loop, so the bass response will be stronger.

Next thinner wire: 2.5 mm^2. Diameter 1.6 mm. We would at least need 5 windings to obtain a larger ratio of L/R. Which should be possible, maybe even 6 turns.

So : maybe the optimum performance of such an approach is actually reached by a few loops instead of only one?

So what about iron in the core instead of magnetic ferrite? How would that affect the inductivity?
I could imagine something like a mix of the two LoZ approaches: a few windings of really thick wire, say, 1.5-6 mm^2, around a traditional pickup core and then use a suitable CT to obtain either microphone or line sensitivity. What Do You think?

[Helmholtz]

Two turns makes increases the L by four times if all the flux is in common.

If you are going to use multiple turns, you could also use enough so that you get sufficient voltage to feed a mic. level input without the transformer at the pickup. Not sure if this saves space our wastes space, but it is something to consider. Of course, Les Paul like this approach long ago, but that was a bigger pickup not suitable for your application.

Agree and this solution would guarantee the best possible bass response. It requires around 500 turns to achieve the same signal level and impedance as the described prototype.

[Joe Gwinn]

And they specify 50-60HZ for [AX-400] and 20kHz to 200kHZ for [AS104].
I'd be happy to understand how such huge difference is possible with such similar features.

It will be a combination of the ferrite material used and how the multiturn secondary is wound.

[bbsailor]

The CTs with the higher permeability are the Kemet CT series specifically designed to sense low current. http://www.kemet.com/Lists/ProductCa..._SE0203_CT.pdf

The cores are Permalloy with no gap. I just measured the unloaded (no string loop installed) 1000 turn Kemet CT-06-100 using a Extech 380193 LCR meter. 3.02H at 120 Hz, 809mH at 1Khz, 32 ohms DCR. This series comes in 500, 750 and 1000 turn models. The key design point is to make a string loop thick enough to keep the output impedance low enough to not load down the upper frequency response that you want to pick up. The rule of thumb is to keep the source impedance, that is the CT output impedance about one tenth of the actual input impedance of the stage it is feeding. Currently, low Z XLR mic inputs are about 2400 ohms, so that means that the maximum output impedance of the CT should be about 240 ohms. Using stranded wire has a higher DCR than a solid copper wire of the same diameter, due to the air space between strands but the higher frequencies will not be as restricted as much as using a thick solid copper wire. I am just reflecting on the physics of what is happening but your ears should tell you what sounds good.

I can fit a single solid strand of AWG 4 wire into this Kemet CT-06-50 almost 6mm opening. Bending this very heavy with the CT in the center, without breaking it may require some design challenges to ensure that once you make the string loop and secure its single loop current generation connection that you have a low impedance pickup with an output impedance in the acceptable range of a low Z mic input. i suspect that you can use an AWG 4 wire and go up to a 750 turn CT or even an 1000 turn CT and still have something that may please your ear. An 8 inch piece of AWG 4 through a 1000 turn CT should produce an output impedance near 240 ohms depending on how well (low resistance connection) you make the string loop connection. Let your ear decide but understand the physics behind it. You can even use the effects of the skin effect to better balance the emphasis of lower frequencies using these known characteristics of the wire string loop size, current driven induction sources and current transformer theory.

I am so glad that folks are finally targeting the XLR input impedance and levels to seek a new sound for guitar pickups. Thanks to all of you experimenting and driving these concepts forward. The challenges now will be the mechanics of dealing with your chosen wire gauge size/style, CT turns ratio, CT physical style chosen (Toroid or E-I frame) and the way you apply these physics concepts to build something that is practical and easy to install on a guitar.

Thanks,

Joseph J. Rogowski

[bea]

If you are going to use multiple turns, you could also use enough so that you get sufficient voltage to feed a mic. level input without the transformer at the pickup. Not sure if this saves space our wastes space, but it is something to consider. Of course, Les Paul like this approach long ago, but that was a bigger pickup not suitable for your application.

Actually the other pickup of the guitar above - the tele neck - is such a pickup: 500 windings, 0.2 mm wire. It is not "neutral", "hifi" or however You may call it, but it emphasizes the mids - just as it did with its original high Z winding. And of course it feeds an XLR input nicely.