Guitar pickups with just 1.15 Ohms DC resistance!

[peskywinnets]

Talk about low impedance pickups!

This guy is experimenting winding pickups that have an end to end coil DC resistance reading of just 2.3 Ohms, but he's centre tapping them & feeding them into a balanced preamplifier (which is how I found his site - balanced pickups is something I'm upto right now), which actually means each half of the coil is actually just 1.15 ohms....

More thoughts on home-made guitar pickups joebrown.org.uk


Quite remarkable, (which is why I remarked)

(here's the preamplifier he's mating his pickups with Pre-Amplifier For Balanced Magnetic Guitar Pickup joebrown.org.uk)

[Steve Conner]

Joseph Rogowski (bbsailor) on this forum has been touting an arrangement for years that uses the string itself as a single-turn pickup, and has a DCR of something like 0.1 ohm.

We were just discussing how in the world to make an active preamp for it, or if that would even be possible.

You don't need a centre tap on the coil. Balanced microphones don't have one.

[Mike Sulzer]

At such low impedance, I doubt that you even need it balanced at all!

Joseph Rogowski (bbsailor) on this forum has been touting an arrangement for years that uses the string itself as a single-turn pickup, and has a DCR of something like 0.1 ohm.

We were just discussing how in the world to make an active preamp for it, or if that would even be possible.

You don't need a centre tap on the coil. Balanced microphones don't have one.

[bbsailor]

At such low impedance, I doubt that you even need it balanced at all!

Mike,

This is quite true as the strings have a very low impedance. I measures the source impedance of each string on a Gibson Scale using the Extech LCR meter. The string set I used had a .011 for the high E string.

E: .679 ohms
A: .892
D: 1.095
G: .774 (plain)
B: 1.008
E: 2.014

Typically, you want a return path to have at least a resistance 10X lower than the lowest source impedance string. Lower is better.

Note that the resistance/impedance is governed by the plain string diameter or the core diameter for wound strings.

For maximum power transfer you want the string impedance to match the input impedance of either a transformer or solid state device, if you can find a solid state device that is low enough. However, in microphones or pickups the goal is to have maximum voltage transfer to the load and is typically idealized at 10 times the source impedance. In high noise evironments I suspect that the lower power matching impedance would offer less noise than the 10X load impedance at the cost of a 6db loss.

Mike, can you verify this assertion with some model?

Noise can be mostly eliminated by reversing the phase of alternate strings or string pairs that have near the same source impedance. I used a Weller soldering gun to induce noise and played with alternate phasing of 8 ohm to 20K miniature output transformers (used with the 8 ohm side being across the string).

With this arrangement I can get about 150 mv (peak) out of each transformer. A simple six-input microphone mixer using one IC stage with a gain of 10 provides enough signal to drive any load with individually adjustable gain on each string all mixed into a single output.

The guitar string is loosely coupled to the magnetic field and the subsequent load on the vibrating string, unlike the ribbon motor in a microphone that is highly coupled to the stronger magnetic field and load offered by the matching transformer. Here is where some limitations that apply to ribbon microphone motor matching differ from guitar string loading. Also, because each string has a different bandwidth, modeling an optimum individual string match can be considered.

Winding individual coils under each string is limited by the size of the bobbin that can line up under each string, the amount of turns, the wire diameter, the transformer input impedance and the transformer tuns ratio to provide a usable output. To be humbuckung, each string would need two coils of the opposite magnet polarity with the coils connected in series with the signals adding.

I found that AWG 32 wire works pretty well on individual string bobbins using .125" diameter magnets with about 300 turns.

My latest device uses two coils, one under the bass E string and A string to send individual signals to a foot switch where the matching transformer is located along with two momemtary foot-operated switches to select which of the two bass strings are sent to the transformer and then to an off the shelf (Boss) octave divider. Since octave dividers only like one note at a time this set up works very well for that purpose.

As you can see, there are some fun things that can be done when the signals from individual strings can be captured and the isolation maintained with adequate low noise performance. Finding a hex opamp with low impedance inputs would be ideal but six miniature transformers do not take up too much space and are fairly inexpensive.

Joseph Rogowski

[Joe Gwinn]

This is quite true as the strings have a very low impedance. I measures the source impedance of each string on a Gibson Scale using the Extech LCR meter. The string set I used had a 0.011 [ohms] for the high E string.

E: 0.679 ohms
A: 0.892
D: 1.095
G: 0.774 (plain)
B: 1.008
E: 2.014

What was the test frequency? I would guess that it was 1 KHz.

Quoted for the high E string is 0.011 ohms in the text and 0.679 or 2.014 ohms in the table, a very large ratio.

The resistances in the table seem a bit high for a piece of steel wire that's maybe 22" long. How accurate is the Extech at such low resistances?

Do you have the equipment to do a 4-wire DC resistance measurement? I do, but currently have only a base guitar with four huge strings at hand.

[Steve Conner]

Interesting stuff!

I imagine the strings will have lots of inductive reactance, since they're made of steel and the wound ones are wrapped in yet more steel. Not to mention high AC resistance because of the skin effect. What does the Extech say for inductance?

[Mike Sulzer]

In high noise evironments I suspect that the lower power matching impedance would offer less noise than the 10X load impedance at the cost of a 6db loss.

Joseph,

This might depend upon what kind of noise it is. I suspect that electric field noise is very small at such a low impedance and is not the problem. If it is magnetic, as one would suspect, then I think the signal and hum should vary together when the input impedance is changed. So it would not matter about the input impedance for the signal to hum ratio.

Steve,

The inductance depends on the size of the loop, and so involves the return path. So the material of the string probably has little effect on the inductance since it effectively fills only a very small fraction of the loop.

Joe,

.011 is probably the diameter of the string in inches. It would be interesting to know more about the measurement.

[Scott S.]

Re: the preamplifier shown in the link above...

What advantage would his circuit have over an instrumentation amplifier (available as an 8 pin DIP chip with few external components)?

[tedmich]

Re: the preamplifier shown in the link above...

What advantage would his circuit have over an instrumentation amplifier (available as an 8 pin DIP chip with few external components)?

might be cheaper, it looks like a std. balanced receiver and balanced/unbalanced converter, similar to what THAT Corp (and Ti) make as single chips.

[bbsailor]

Interesting stuff!

I imagine the strings will have lots of inductive reactance, since they're made of steel and the wound ones are wrapped in yet more steel. Not to mention high AC resistance because of the skin effect. What does the Extech say for inductance?

Joe and Steve

I do not have a 4-wire, low resistance bridge but I have remeasured the strings on my Fender Tele using 0.011" diameter high E string set. My Gibson is currently apart for some modifications.

I will list the Extech raw results in comma separated values below. Please note that the leads of the Extech LCR meter reach to just inside the nut and the bridge providing a reading of about 25" of string length. Anyone else is welcome to verify these results.

String Name, R@1KHz, R@120Hz, Inductance@1KHz, Inductance @120Hz
E wound, 0.3, 0.3, 4uH, 0.009mH
A wound, 0.32, 0.32 3.7uH, 0.009mH
D wound, 0.56, 0.56, 2.3uH, 0.008mH
G plain, 0.71, 0.71, 0.1uH, 0.005mH
B plain, 1.08, 1.08, 1.3uH, 0.004mH
E plain, 0.85, 0.81, 0.7uH, 0.003mH

The results from the plain E and B strings are counter to logic but that is what I measured several times while making sure I had a good connection between the test lead clip and the string.

Measurements can vary based on the following variables.

String temperature
String metal composition
String coatings
Dirt
Rust
String age and amount of stretch since new
Test lead resistance
Clip lead spring pressure on string connection
String core diameter for wound strings

The plain strings range from 0.71 to 1.08 ohms while the wound strings range from 0.3 to 0.56 ohms.

If you wanted the maximum voltage transfer to a transformer you would want an input impedance at least 10X the string source impedance. However, each string may not produce a musically balanced output with that 10X matching so some way to adjust the individual string output level should be provided. A commonly available transformer is the Mouser Electronics 42KM019 which is 20K ohm (CT) to 4 ohm (CT). The DC resistance is Primary:1000 ohms, and Secondary: 0.5 ohms. The key to maximum voltage transfer from the string to the transformer is to limit the resistance in series with the low impedance 4 ohms input side of the transformer (used backwards with the 4 ohm side being the input). A few mV generated or induced on the string will give you over 100mv at the output of the transformer. By using the full winding or the CT winding on the 4 ohm side can provide some matching alternatives between the wound strings with lower impedance and the plain strings with a higher impedance.

One key point to the maximizing the output level is the length of string that is in the magnetic field; longer string length in a magnetic field produces more output, with wide variation in output between short (1 inch) and long (6 inch) field length. The long magnetic field should taper closer to the string as it gets closer to the bridge as the string vibration amplitude is reduced near the bridge.

I hope this stimulates some good technical discussions.

Joseph Rogowski

[Mike Sulzer]

String Name, R@1KHz, R@120Hz, Inductance@1KHz, Inductance @120Hz
E wound, 0.3, 0.3, 4uH, 0.009mH
A wound, 0.32, 0.32 3.7uH, 0.009mH
D wound, 0.56, 0.56, 2.3uH, 0.008mH
G plain, 0.71, 0.71, 0.1uH, 0.005mH
B plain, 1.08, 1.08, 1.3uH, 0.004mH
E plain, 0.85, 0.81, 0.7uH, 0.003mH

The impedance of a 1 micro-henry inductor at 1000 Hz is 2*pi*1000*1e-6 = .00628 ohms. I suspect that the Extech can not accurately measure this when the resistance is so much higher? Could this explain why the inductance measurements are so different at the two frequencies?

[Joe Gwinn]

The impedance of a 1 micro-henry inductor at 1000 Hz is 2*pi*1000*1e-6 = 0.00628 ohms. I suspect that the Extech can not accurately measure this when the resistance is so much higher? Could this explain why the inductance measurements are so different at the two frequencies?

The Extech needs the Q to exceed about 2 to achieve stated accuracy, and we have a Q that's far less than 1, so the inductance readings are not going to be accurate, as you suspect.

I don't know that I believe the resistance readings either.

[Joe Gwinn]

Four-wire resistance measurement of a base guitar

Well, I have only one guitar at hand, a beater of a 4-string base guitar, so I measured the DC resistance of its strings with a 6.5-digit digital volt-ohm meter.

I wasn't totally careful to prevent stray paths (in parallel with the string being tested), although I did insert pieces of paper in strategic locations to prevent shorting. All strings are wire-wound wire, and appear to be made entirely of steel.

Diameter = 0.046", Rdc= 0.9545 ohms.

Diameter = 0.066", Rdc = 0.7851 ohms.

Diameter = 0.081", Rdc = 0.4869 ohms.

Diameter = 0.099", Rdc = 0.5566 ohms.

Now, with more care I would get better answers. The above pattern doesn't quite make sense. One thing is clear, that string resistances of order one or two ohms are certainly possible. This is a factor higher than I thought likely.

[bbsailor]

Well, I have only one guitar at hand, a beater of a 4-string base guitar, so I measured the DC resistance of its strings with a 6.5-digit digital volt-ohm meter.

I wasn't totally careful to prevent stray paths (in parallel with the string being tested), although I did insert pieces of paper in strategic locations to prevent shorting. All strings are wire-wound wire, and appear to be made entirely of steel.

Diameter = 0.046", Rdc= 0.9545 ohms.

Diameter = 0.066", Rdc = 0.7851 ohms.

Diameter = 0.081", Rdc = 0.4869 ohms.

Diameter = 0.099", Rdc = 0.5566 ohms.

Now, with more care I would get better answers. The above pattern doesn't quite make sense. One thing is clear, that string resistances of order one or two ohms are certainly possible. This is a factor higher than I thought likely.

Joe,

Another way to look at it is to list the core size for each string and look at the low to high resistance ranking from that perspective. I think core size determines the overall conductivity.

Joseph Rogowski

[David Schwab]

I got this on my 5 string bass:

G-0.045" = 0.5 Ω

D-0.065" = 0.5 Ω

A-0.080" = 0.6 Ω

E-0.100" = 0.7 Ω

B-0.130" = 0.4 Ω

The strings are D'Addario XL Nickel Round Wounds. I did the measurements touching one probe right behind the bridge saddle and one right behind the nut using my Radio Shack digital VOM. The scale length is 34".