Guitar pickup theory #2a: Properties of Alnico magnet guitar pickup pole pieces

Properties of AlNiCo magnet grades.

I have compiled a list of attributes of alnico magnets since I could not find this information all in one place anywhere… sigh. This info is derived from the data sheets of several magnetic-material manufacturers, as well as hand math, and data culled form dozens of test-data papers.

I will give the charts first, explain them second, and cite the material third.

This first chart is mainly to reference the field strength in gauss at various distance from the poles. Field strength is dependent on several factors like residual inductance, distance, and geometry, so you can’t simply look at a data sheet and know what the field strength will be.  So I calculated the axial field strength at 3 distances from the magnet face.  The geometry was a cylindrical pole piece .187″ wide by .650″ long, which is fairly standard.

1mm is essentially surface gauss, 3mm represents typical string distance, and 6.333mm is to represent moving the pickup 1/4″ down from the strings. I calculated these by hand and checked my math against empirical data… for example: gauss for a vintage Strat pole magnet made of cast alnico 5 measures about 950g on the surface, and my math puts the field at 928g just above the surface, so that checks out.  The data used for my calculations was all from a single manufacturer, so they are consistent with each other, although in reality they will vary from one manufacturer to another since they all use slightly different alloy compositions.

Other info provided is just for reference, and Samarium-Cobalt/ Neodymium/ ceramic magnet types are included for comparison.

Skin Depth Coefficient information is culled from a post by Joe Gwinn (1) on the Ampage forum.  It gives a general idea of the tendency of the material to be subject to eddy currents, thus reducing the amplitude at higher frequencies (lower number = more susceptible to high-frequency loss).  I will calculate the missing fields when I have time.

Relative permeability is given for the sake of calculating the coil inductance with respect to the core as well as flux-carrying capability.

Any field marked “N/A” means that data is not available.

MAGNET FIELD STRENGTH AND PROPERTIES:

pole piece chart

So the bottom line in terms of just raw field strength vs grade is as follows:

Neo(hi) > SmCo(hi) > Ceramic(hi) > 5cast > 9cast=5sint=6cast > 6sint > 8cast > 8HCcast=2cast > 2sint=3cast > 8HCsint

This can also be applied to calculate the pull force on the string if we consider:

force = 0.577 * B^2 * A (where B is the flux density in the gap and A is magnet area) (note:0.577 is for rare earth magnets, I am not sure if there is a different coefficient for Alnico.  If I find one, I’ll revise the formula)

The “HC” grade of Alnico 8 means ‘high coercivity’ meaning that it is resistant to demagnetization, but this is at the expense of strength.

Before you send me angry messages contradicting this chart… BEAR IN MIND that this is just for field strength and not pull, and this particular shape,and only at a finite point at a set distance directly center-axis.  Off-axis, and for a huge block of material, the order of the list would change… also, I had to estimate a fairly linear Q2 curve for the sake of efficiency,  also:

  1. there are many sub-grades and compositions within each major grade (so this is just a guide).  I found about a dozen grades of alnico8, and half a dozen of alnico9 (see the following chart and text to see how radically that can change the magnet’s characteristics)

2. as the geometry and distance changes, different grades will swap places on the gauss chart, etc.  ALSO, alnico is nonlinear, with some grades being more linear than the others, and there is an extremely-wide variation from one manufacturer to another.  So while it looks like alnico 5 is 30% stronger than alnico 8, that does not always hold true.  Here is a quote from a manufacturer:

“Alnico 8 magnets have a lower remanence, but a higher coercivity than Alnico 5 magnets. This means that Alnico 5, though stronger in the right circumstances, is easier to demagnetize.”

3. This chart is field strength, NOT Holding force (against a ferromagnetic material) which does not *necessarily* equate to stronger field at a distance, so a statement like this is utterly meaningless without more details on geometry, specific grade and distance: “DUDE, the math says grade-X is stronger than grade-Y, but grade-Y totally sticks to my fridge harder so the math is wrong and grade-Y will totally make the hotter pickup!”  So don’t make that mistake please.  An example would be this: take a weak magnet and a strong magnet and place them both on steel… obviously the stronger magnet will be harder to pull off.  NOW place a steel u-channel around the weaker magnet and it will be harder to pull off.  Did we alter physics and invalidate our math?  No of course not, we changed the geometry and flux path.

Another example is: take 2 magnets: #1 is 3x weaker than #2, but is 4x wider than it is all. #2 is the opposite: stronger and 4x taller than it is wide.  #1 is the weaker magnet, and has a surface gauss of 2500, #2 is the stronger magnet and has a surface gauss of 7500.  The pull strength yanking it away from the fridge of the weaker magnet will be FOUR TIMES stronger than the stronger magnet.  ALSO, the strength of the weaker magnet (surface gauss 2500) will be 2200 at 1/8″, while the strength of the stronger magnet (gauss 7500) will be 1600 at 1/8″ because of the geometry.

4. this behavior is very non-linear, and that will vary according to the actual composition of each grade and shape:

See this note from Dexter Magnetics:

“Low coercivity materials, like Alnico 5, have a different external field shape because Hc is much lower than Br. Hc for Alnico 5 is about 5% of Br, so a magnet’s own external field, Bd, affects internal domain alignment toward the polar ends. Thus, domains at the ends and corners of an open circuit Alnico rod magnet do not remain aligned after magnetization in an air core coil (unless keepered), and plotted external field lines appear to have a focal point below the polar surfaces of the magnet. A length factor of 0.7 is often used in calculations to account for this effect (poles offset 15% from each polar end). However, 0.85 is a more realistic length factor for Alnico magnets with a geometry that causes them to operate above the knee in their second quadrant curve.”

Here is a graph of the demagnetization curve of an alnico magnet with load lines and recoil slope plotted to give a visual representation of what we are talking about (pardon the poor drawing, sorry):

Alnico curve with load lines

~~~~~

This second chart is the composition of the actual alloy grades.  Although alnico gets it’s name from aluminum/nickel/cobalt, it is actually a blend of 8 main materials, plus some trace materials.

ALLOY COMPOSITION:

alnico composition
Alnico magnets also contain amounts of niobium, silicon, Hafnium, and zirconium, but I could not find definitive data on these amounts.  However, generally cast magnets will contain several percent niobium and Hafnium, and sintered magnets contain a bit less: for example — cast magnets contain more niobium to enhance magnetic properties, while sintered magnets contain less because it degrades the mechanical properties and has diminishing returns and even detrimental effect in terms of added magnetic strength beyond a small amount, so typically 0.0-0.7% is used in sintered, and 0.0-1-3% is used in cast.

The closest I could find for percentages is the following list for grades 1-5 cast:

1 0.0%
2 0.5%
3 1.0%
4 1.5%
5 2.0%

(2) cited: DESIGN OPTIMIZATION OF MAGNETIC ALLOYS AND NICKEL-BASED SUPERALLOYS FOR HIGH TEMPERATURE APPLICATIONS Rajesh Jha George S. Dulikravich Department of Mechanical and Materials Engineering, MAIDROC Lab.
“1. Cobalt: It is a γ stabilizer. Hence, a solutionization anneal is needed to homogenize it to a single α phase. Cobalt increases coercivity and Curie temperature.

2. Nickel: It is also a γ stabilizer. It increases Hc (less than Cobalt), but at expense of Br.

3. Aluminum: It is a α stabilizer. Hence, it will help in reducing the solutionization temperature. It affects Hc positively.

4. Copper: It is also a α stabilizer. It affects Hc positively and increases it. Additionally, it also increases Br. In Alnico 8 and 9, Cu precipitates out of the α2 phase into particles and increases the phase separation between α1 and α2 phases. It results in an increase in Hc. In Alnico 5-7, Cu remains in α2 phases and it leads to an increase in Hc while decrease in Curie temperature.

5. Titanium: It is also a α stabilizer. It is one of the most reactive elements. It reacts with impurities such as S and N and precipitates out or purifies the magnet. It also eliminates Carbon. Carbon is a strong γ stabilizer and hence needs to be eliminated at any cost. Titanium helps in grain refining and assists columnar grain growth. It is to be noted that, due to columnar grain growth, majority of grains are aligned perpendicular to the chill plate. Large shape anisotropy can be achieved if spinodal decomposition occurs in this direction. Titanium increases Hc while it decreases Br.

6. Niobium: It is also a α stabilizer. It helps in neutralizing the affects of Carbon. Like Titanium, Nb assists in columnar grain growth. Nb increases Hc, while it decreases Br. But decrease in Br due to Nb is less than that observed due to Ti.

7. Hafnium: It is used for enhancing high-temperature properties. Hf usually precipitates at the grain boundary and helps in improving creep properties…”

(8. Iron makes up the remainder, along with other crystal trace such as silicon and zirconium)

~~~~~

WHAT DOES THIS MEAN?

Well, we can use the chart to make several assumptions, the most important being:

-field strength (low-amplitude sensitivity, output, higher order harmonic amplitude)

-eddy currents (high-frequency attenuation–and output to an extend depending how low into the frequencies the slope extends)

-permeability (coil inductance, ability to concentrate flux through the coil)

-coercivity (likeliness to lose charge in the presence of other poles of like-polarity)

So (very generally speaking) higher strength = higher output and brighter tone(due to more energy for upper harmonics), higher eddy currents = darker tone, higher permeability = more output-per-turn and lower freq/Q peak, etc.

So although the winding and coil geometry will have a much greater effect, if ‘all else is equal’, we can expect:

hot&bright: highest field strength, lowest eddy currents, moderate permeability

hot&balanced: high field strength, moderate eddy currents, moderate permeability

hot&dark: higher field strength, higher eddy currents, higher permeability

cool&bright: lower field strength, lower eddy currents

cool&balanced: moderate field strength, moderate eddy currents, moderate permeability

cool&dark: lowest field strength, highest eddy currents

BUT… we know that ‘all else’ is not always equal, so the assumptions we made about tone based on geometry and coil structure, etc.

These assumptions are for a Strat magnet-pole type.  If we take this information and apply it to a hum bucker, we can see where (ignoring frequency cancellation, wider aperture and electronic parameters) this can apply to humbuckers in that hum buckers use steel as pole pieces, which has a very high permeability and higher eddy loss, so we can take the logic above and assume that means a wider peak with a lower Q amplitude and lower resonant frequency (as well as hotter coil) from the additional inductance, as well as a sharper rolloff of higher frequencies from eddy current loss… so in other words: a hotter warmer sound with less attack and presence, which we know to be true.

This data also suggests that interesting hybrids can be used, like combining a strong&small rare-earth magnet with a steel pole to get higher sensitivity and output, but lower resonant frequency and Q, as well as more higher-frequency attenuation.  Single-coil-users tend to be highly resistant to change, so I fear that we have not seen much of this since words like ‘alnico’ and ‘vintage’ seem to be very strong selling-points:  ‘WE DON’T NEED ANY OF THAT NEW-FANGLED…. um… GOOD TONE.”

As an example, here are self-resonance sweeps of a coil with four different pole materials:

Blue = no poles (air core)

Orange = alnico 5 poles

Green = mild steel humbucker slugs

Yellow = steel humbucker skinny screws

pole resonance

The difference is quite obvious.  Here is a second graph.  Yellow is still the skinny hum bucker screws, and blue is still the air core, BUT Purple is the same skinny hum bucker screws, but this time they are saturated by a vary strong magnet:

pole saturated

You can see that the resonance has changed quite a bit by the magnet saturating the poles.

Hope this helps out.  I am fascinated by the subject, so I will continue to write about it, and I will correct this information as I go.

-Alex

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Appendix:

citation: (1) SDC cited Joe Gwinn from Ampage forum 2005:

“…The issue for pickup design is that recoil permeability affects two parameters of interest, inductance and eddy currents.  Inductance is directly affected, increasing by more or less the recoil-permeability factor compared to the same coil with air instead of alnico.  Eddy currents are also increased, as the increased permeability focuses the field somewhat, thus increasing the currents, but this is balanced by the bulk resistivity of the material.

The skin depth is the distance into the material at which a magnetic field varying at the specified frequency will have attenuated to about one third (1/2.7182) of the amplitude at the surface. At two skin depths, the amplitude will be about one nineth, and so on…

…The formula for the skin depth coefficient (SDC) is as follows: Sqrt[(rho/10^6)/mu]*10^4.5/(2pi ). Divide the SDC by the square root of the frequency in hertz to yield the skin depth in centimeters, where rho is the bulk resistivity in microhm-centimeters and mu is the recoil permeability (a pure ratio). This formula is from section 10(a) (on page 30) of “The Theory and Design of Inductance Coils”, V.G. Welsby, Macdonald & Co, (Publishers), Ltd., 180 pages, London 1950…”

~~~~~

common industry name examples:
Alnico Name/Grade equivalents (alnico-info.com):
Alnico 5 (Alnico5_ACA34), Alnico 5 (Alnico5_ACA37), Alnico 5 (Alnico5_ACA40), Alnico 5 (Alnico5_ACA44), Alnico 5 (Alnico5_ASA34),
Alnico 6 (Alnico6_ACA28), Alnico 6 (Alnico6_ASA28),
Alnico 5DG (Alnico5DG_ACA52),
Alnico 5-7 (Alnico5-7_ACA60),
Alnico 8 (Alnico8_ACA38), Alnico 8 (Alnico8_ACA40), Alnico 8 (Alnico8_ACA44), Alnico 8 (Alnico8_ASA38), Alnico 8 (Alnico8_ASA44), Alnico 8 (Alnico8_ASA48), Alnico 8 (Alnico8_ICA18), Alnico 8 (Alnico8_ISA18), Alnico 8 (Alnico8_ISA20),
Alnico 8HC (Alnico8HC_ACA36),
Alnico 9 (Alnico9_ACAT60), Alnico 9 (Alnico9_ACAT72), Alnico 9 (Alnico9_ACAT80),
Alnico 8HC (Alnico8HC_ASA36),
Alnico 3 (Alnico3_ICA10), Alnico 3 (Alnico3_ISA10),
Alnico 2 (Alnico2_ICA12), Alnico 2 (Alnico2_ISA12),
Alnico_BA7,
Alnico_BA8

Advertisements

About alexkenis

Guitarist, philosopher, tinkerer

31 comments

  1. I thought I’d maybe adjust the sensor output with an opamp, get rid of the offset and scale the voltage so it’s, say, 1V for 1000 Gauss… Would make it easier to read. The markings on my sensor are also useless, I forgot what it is, will have to test a bit to check the sensitivity.

  2. alexirae

    Hi guys, hope you can find this link useful:
    https://www.kjmagnetics.com/blog.asp?p=gaussmeter

  3. alexirae

    And this small circuit I made in DIYLC:

  4. alexirae

    Final result:

    (sorry for the multiple posts, I wanted to edit my original post but I wasn’t able to do it!)

  5. Oh nice job alexirae! That looks pretty close to what I made, but I have a 5v regulator on there and used set values instead of a pot for the ref (I probably should have used a pot). Thanks for posting that.

    Good to see bancika still has that layout software up. I had forgotten about it.

  6. Thanks for the circuit, it handles zero offset and output range with a minimum of components. However, I tested out my unidentified sensor and it maxes out on an Alnico2 polepiece and the weakest neo magnet I have (3200 Gauss). I have to find a sensor that can handle stronger fields. There’s not much selection available locally, I’ll probably have to order something.

  7. The least sensitive analog linear sensor I’ve found so far is Allegro A1308LLHLX-05-T. The sensitivity is 0.5mV/G, so at 5V supply voltage it can (theoretically) handle +/-0.5T. The real range is probably smaller. This is still not enough for neo, I’ll keep on looking.

    • I wrote a reply, but then realized that my math was off by a decimal place, so I scrapped it. Looks like for a pre-set sensitivity IC with built in electronics, the one you found is about the best you can hope to do. I looked at the data sheet from the Analog devices sensor (they seem to only make the one that will work), and that is set at 0.4mV/G, so that could work too. Run either at max Vcc to get a bit more range, but we are still looking at maybe 0.6-0.7t max.

      Other options appear to be high voltage sensors, but finding them is next to impossible. Another possibility is GaAs sensor elements, which will require some more supporting electronics and an instrumentation opamp. It looks like AST HE144, Siemens/Infineon KSY14, CYSJ302C are examples of high-strength field sensors tested out to up to 10T like this one maybe http://www.sonnecy-shop.com/en/linear-hall-effect-sensors-elements-cysj302c-max.-sensitivity-1.8-2.5-mv/mt-measuring-range-3t.html . They may work, but I have not looked into it too much though.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: