Pickup info

[ed packard]

Is this kind of "stuff" of interest?

Greetings chaps…My first post will be in the PICKUPS section. Pickups have long been made with windings of wire around either the magnets, or around a return path in the magnetic circuit. The vibrating string causes a change in flux (think something invisible like fluid), and the change in flux (means flow) causes a resulting change in voltage at the ends of the winding(s).

The winding may be made with different diameter wires. Measure the resistance (not the impedance) of the winding with an OHM meter and the value means nothing unless the size of the wire used is known.

The Tone and volume shaping parameters of the pickup are:
The resistance (R) of the coil.
The capacitance (C) of the coil.
The inductance (L) of the coil.
The magnetic field strength and shape.
The proximity of the pickup poles to the string.
The “load” placed across the pickups windings.

The RLC above make a frequency filter (band pass) applied to the vibrations received from the vibrating string(s). Which harmonics in the string (vibrations) are emphasized depend upon where the pickup is positioned, and the shape of the magnetic field as seen by the string. The shape of the field is a function of magnet shape, material, and any flux return path included. The harmonic content controls what is commonly called “tone”.

The output voltage of the coil is a linear function of the number of windings (up to a point). Double the turns (N) => gives double the output voltage (E) => double the resistance (R).

Double the turns (N) => squares the inductance (L).

Double the turns (N) => increases the capacitance (C) by an amount that depends upon the shape of the winding. Windings may be scramble wound, random wound, layer wound etc.. Layer wound has the most capacitance. Capacitance is lower per turn with larger wire diameter.

The C and L have a frequency dependant property called REACTANCE =>Xc and Xl.
The frequency at which Xc and Xl are equal is the “resonant” frequency of the pickup.
The inductive reactance (Xl) => varies as the frequency (f) times 2 times 3.1416 commonly written as (w)… pronounced as ohmega, w is “ohmega” = 2*pi*f, or Xl = wL. You can see that the reactance increases rapidly with increasing frequency.

The capacitive reactance (Xc) decreases with frequency as per 1/w => the reciprocal of w => or 1/((2* pi *f)*C), or Xc = 1/wC.

The summation of the RLC (combined reactances) gives the resonant frequency of the pickup. The Q => the ratio of reactance to resistance…the higher the resistance, the lower the Q. The lower the Q, the wider the frequency bandwidth.

That is the basics for the windings and magnets types of pickup.

Let’s look at the wire parameters => Resistivity => usually given in OHMS PER THOUSAND FEET The OHMS per KFT (Thousand ft) for #36 gauge wire is approx 470 ohms. The diameter is approx 0.0047 inches.
. This varies with wire size. If you know the wire size and the length of a turn in the coil, you can figure the number of turns in the coil.

The OHMS per KFT (Thousand ft) for #30 gauge wire is approx 115 ohms. The diameter is approx 0.0095 inches.

The OHMS per KFT (Thousand ft) for #33 gauge wire is approx 231 ohms. The diameter is approx 0.0074 inches.

The OHMS per KFT (Thousand ft) for #36 gauge wire is approx 470 ohms. The diameter is approx 0.0047 inches.

The OHMS per KFT (Thousand ft) for #39 gauge wire is approx 950 ohms. The diameter is approx 0.0033 inches.

From the above 4 wire size data, you can see the futility of trying to correlate pickup performance with ohm meter readings of the pickups winding without knowing the wire size…even then it is very dicey.

These days, computer modeling, and instrumentation like FSA (Frequency Spectrum Analysis) can save a lot of prototype time and give “objective and repeatable” performance data. The next two links will show examples of both. Watch out for the all semiconductor pickup in the links!

http://s75.photobucket.com/albums/i287/ ... 20SENSORS/

[richard37066]

Ah, Ed - yer at it again.

This is your old buddy Richard from the SGF - back to haunt you with an interminable list of questions. I haven't posted there for personal reasons and you've been suspiciously absent for whatever the reason.

Sure wish that you'd get that all-electronic pickup patented - or whatever - such that we can see precisely what's in the innards. This "teasing" has gotta go! I figger that if I pester you enough that you'll cave in and reveal the magic just to get me off of your back!

Your finger schlupped when typing the formula for capactive reactance. The reciprocal should have been written 1/omegaC or 1/(2 pi f) C. I think ya left out the "2" in the formula.

Ain't it nice that I get to aggravate you again?

Respectfully,

Richard

(Recognize that closing?)

[Bent]

Ed, wow! Not bad for a first post. The first part was fairly simple - something that I can understand with a bit of reading.
The second, more difficult, part is something that takes study and analysis. Might be over my head but I will try. This is a very fine "paper" you presented and we should try and do it justice by studying and learning from it.

My big problem is noise form my single coils. I love them for tone and hate them for all the noise they produce. You must have the knowledge - is there ANYTHING at all that I can do to further reduce the noise when making single coils?
What I am doing is: After winding with AWG 42 wire, I wrap a strip of copper foil around the taped windings and ground it to the ground wire. Then I wind another layer of tape around the foil. I make sure the pup is grounded good. I believe that the lower the resistance, the less hum. True or false? I wind the pups to 13.5K to 17.5K

Thank you for your fine posting!

[ed packard]

Hello Richard & Bent!

Richard…thanks for keeping an eye on the text…I did not know what my finger ..and mind (an assumption) were doing. I edited to try to fix it and add Xl XC clarification.

Bent…the info for your #48 wire is:

The OHMS per KFT (Thousand ft) for #48 gauge wire is approx 7576 ohms. The diameter is approx 0.00117 inches.

If you can blow the PHOTOBUCKET standard type pickup plots up enough to read the text, you will find some interesting conclusions available. First and most impoetant is that when the string crosses the magnet in vibration, it disturbs much less flux tha when it vibrates toward and away from the magnet by the same amount…the result is that because the string rotates there is a change in amplitude for the harmonic(s) as they vibrate. This can be controlled by shaping the magnetic circuit. It is why some pickups sound lively and others flat and dead.

How do you shape the magnetic field?...the type and shape of the magnets, the method in the return path, and the proximity of the pickup to the strings.

Bent, by noise do you mean AC hum, high frequency noise, or microphonics? Minimal microphonics is obtained by freezing the motion of the coil and magnets with respect to each other. Some folk use wax or silicon sealant, and fewer use wire whose insulation is chemically bondable polyurethane.

Hum from 60 HZ/50 Hz mains is best attacked via common mode rejection methods like opposing coils (humbuckers). The best that can be done with “single coils” would be to use the Copper tape wrap (3M makes a tape) as it appears that you have done. This should also help with RF noise like radio stations and computer interference.

What the goal would be is a high signal to noise ratio.

RE the semiconductor pickup:

No patent will be applied for by me…the data will be made public for use by whoever and for whatever reason….I have enough “intellectual” property” to my name anyway.
This is my retirement fun.

The key component in the semiconductor pickup is a HONEYWELL magnetis chip in the form of a Wheatstone bridge that allows measuring the angle of a magnetic field. As the string vibrates across the chip it changes the angle of the magnetic field, hence the output voltage. Follow this with an “in-amp” to set the output voltage level and you have an individual string pickup if you want, or an all strings combined device…or both at the same time. At this point we are still having an analog output.

Add a MUXed ADC circuit and we go digital. Add a DSP device and start manipulating the sounds.

The boards that you see on PHOTOBUCKET are built for 11/32” string spacing (MOST PSGs). Another batch have been released with the standard guitar string spacing of 3/8”.

There are two ways to populate the boards the first way is to DC couple the Mag chip to the in-amp…this allows using the channel by itself as a non contact position sensor for pedals and levers. The second way is to AC couple the mag chip to the in-amp…this is better for the vibrating string pickup function…same board, different Rs and Cs.

The mag chip of choice is an HMC1501 and associated family chip.

Back to the standard pickup: Next/or soon, we will cover the effect of the pedal/amp load upon the frequency response of the pickup= how to lose highs and attack in a hurry.

Edp

[Farmer]

Ed, you are going to be a fine addition to this forum. Go to the introductions, and tell us about yourself. Welcome Ed.......Mike

[Ozarkbrazz]

Good stuff here, this may cause me to try rebuilding the pickup in the used pedal steel I just purchased. My issue at the moment is the pickup is noisy, not the 60 cycle hum but picking up the pedals and mechanical noise. I'm just relearning to play after 30 yrs. so part of it may be technique but I dont think so. How critical is the mounting to the instrument? Russ B.

[Bent]

Russ,
Your pickup has gone microphonic. Only a new re-wind will fix that.When done winding, you "pot" it with wax

[ed packard]

Back to the common winding and magnets pickup:

Bent has it…after you wind it, you should immobilize the magnets and wires with respect to themselves and each other. Potting with wax is one way. Using a wire with insulation coated with a substance that can be bonded by a chemical is another. This kind of wire is called polyurethane bondable. It is used to make free standing coils and such.

Some folk have used silicon sealing compound…careful as the acidic sealant can eat the magnet wire with time.

Here is the table for wire properties vs. wire gauge number and turn length as a function of string count and string spacing.


Here are the tables for XL (Inductive reactance) as a function of frequency and Inductance value, and the table for XC (capacitive reactance) as a function of frequency and capacitance value.


In the last post we introduced some alphabet soup = R,L,C,Q,X etc. Now let’s add Z = impedance…the most misused word in the pickup world, used as a substitute for resistance re pickup windings.

If we combine XC,XL, and R in the correct manner you get Impedance = Z.

Just considering the L by itself for a moment, we can see from the table that as the frequency goes up, the XL increases…and as the L increases the XL increases.; so high freq and high L have a very high impedance just from the L value.

If this Source Impedance from the L is high with respect to a 500K ohm volume pedal, the high frequencies will be lost. If the L is low, then less highs will be lost. The C in the winding also tend to cost us highs, and the R costs us output across the board.

What we need is a low L, a low C, and a low R with lots of output…good luck, this is a world of compromise.

Forgetting the wire issue for a while, let’s beat up on the magnets. There are three types of magnets available to us = alnico, barium ferrite , and the rare earth magnets. The alnico type is the most common from the past pickups. They come in various diameters and lengths. The length to dia. ratio is important as it determines the field shape from the magnet and hence determines the string across the magnet to string towards and away from the magnet output differences. A further issue is that short fat alnicos will lose strength just from a ride across the California desert in a closed trunk!!

The Barium Ferrite mags don’t have that problem. They are generally used in bar form across the strings. Their field shape is different in general from the alnicos.

Then there is the rare earth type. In general they need a magnetically conductive shell to route their field to the strings and back….what flows out must flow back. We want to use as much of the flow as possible in the vibration to volts conversion. What is not used is called leakage.

One way is to use few turns and a step up transformer to convert the low signal to a useable one. The turn(s) do not have to be wire, but can be any non ferrous metal.

OK, so the string vibrates in the mag field which is dependant on the mag circuit and what does not get lost by the coil’s R,L,C,X,Z and the load properties goes to the Amp.

We could spend the rest of the day on winding types/styles re hum and other noises, but let’s look at a winding type that I have had much fun with.

The winding is made with bifilar wire, and the connection is called differential. Bifillar means the two strands of wire are bonded together side by side, and wound that way into the coil. Let’s say that one wire is d = red, and the other one green. When we are thru making the coil, we connect the start red and the finish green together. We now have three leads. If we tie the center lead (start red and finish green ) to ground, andeach of the other two leads to the input of a differential op amp, we have high common mode rejection (bye bye hum), and if the op amp is an instrumentation amp, we have high input impedance and little high frequency loss.

Complete the scheme by using a voltage controlled gain op amp. Now use your VP or? To control the voltage that controls the gain…or wait for the semiconductor pickups. I have just released the ones with the standard guitar 3/8” string spacing for the evaluation build…the 11/32” spacing for PSGs works fine.

[Georg]

Staying analog, I envision a VCA controlled by the signal-amplitude as well as DC, both control voltages adjusted via the VP, to get a rising compression-effect with VP towards fully on and normal-tapered regulation from mid-point down.

Plan to use a regular, 10 strings, low-Z coil into a built-in dif-amp stage that provides variable load, as main PU, but see a place for your semiconductor PU circuit for splitting out individual strings for further processing before mixing them back in.

As long as I can get the entire electronics inside the PSG and power it up there, I'm happy. Want nothing but power supply for the built-in electronics and a power-amp that takes line-level input (1.4 V), in addition to the PSG itself.

[ed packard]

Georg;

Sounds like an approach. I used a comparator on the analog signal to extract the peak voltage of the pickup signal to control the amplitude via the peak follower voltage…a bit of curve shaping involved to give flexibility to the amplitude control etc..

The DC control using the semiconductor pickup (or other) is obtained by using one channel (snipped off) of the pickup DC coupled. The channel is used for position sensing of the pedals and levers…also can be used as the DC for the VCA control (replace pots etc. in the VP) as desired = one channel layout does it all.

Onward and upward; At present I use TONE+ software in the computer to shape/modify the pickup signal (digitized by the computer). It has a large number of effects (30+?) that can be assembled into complex channels.

Edp