Solid-Body 12-Pickup DIY Autoharp

[janusjones]

Hi there - just joined. Did so, I confess, mostly looking for an answer to this question!

I am in the process of building a 36-string, 40-inch long autoharp. The strings are all electric bass and heavy electric guitar strings, and the machine is going to mount 2 full chord-bar covers for a total of 42 playable chords.

I'm going to have to mount 12 single-coil pickups to make the thing work.

The project is going to be so expensive (for me - I'm a high school English teacher) that I just went ahead and got the pickups on e-bay. I'm thinking of 6 tone controls and a master volume, but am unsure of how to swing a switch - there's going to be a pair of pickups for every 6 strings.

I have a friend who's a decent solderer and has a bit of experience modifying instruments, but we thought this was something we ought to get some advice on.

Any thoughts?

Thanks so much for any and all help! I'm hoping that, with advice, I'll be able to avoid frying myself building this thing.

[ChrisK]

Welcome a'board.

I have a few questions.

Why do you need two pickups per 6 strings?

Why do you need a switch?

Are the 12 pickups identical? What pickups did you buy?

Are the ones that you are pairing together RWRP (Reverse Wound Reverse Polarity) with respect to each other? For a combination of two pickups to be hum canceling, one has to be wound the opposite direction as compared to the other. The magnets also have to be of opposite polarity. The former is trivial to change (just reverse where you connect the leads into your circuit), the latter is not, especially in individual pole magnet pickups).

Testing for opposite magnetic polarity is easy; magnets of opposite polarity attract.

If the pickups are not RWRP in pairs, they will not be hum canceling.

If the pickups are not identical, you will need to have level adjustments. This means a volume pot for each set. Now, you could use trimming potentiometers to do this and still have a master volume.


What is the lowest frequency/note that will be produced by this instrument? If it is lower that 80Hz, you might want to consider using bass pickups for those notes. A standard 34" scale bass guitar has its lowest string tuned to about 40 Hz. For a 40" string length, you could go no higher than four steps lower and maintain the same string tension.

Speaking of tension, you should do a total string tension calculation for all 36 strings. You will be surprised by how much it is.

Since the step ratio in the equally tempered scale is 2^^1/12 or 1.059463, one can determine the optimum tone for a given gauge of string by converting the length ratio difference against that of a standard string length (34" for bass, 25" for guitar).

For instance, if the length is about 19" for a guitar string, you would get a note 5 steps higher.


You mention using 6 tone controls and a master volume.

Why do you want 6 tone controls? Presuming that this instrument will have a tuning interval of a string for every half-step, you will have three octaves of notes possible. A 24 fret electric guitar has a range of four octaves and does quite well with one tone control.

[newey]

JJ-

Hello and welcome!

OK, we've had some strange queries around here, but this is fresh!

I don't really have any solid answers for you, just a couple of random observations. . . FWIW

First, that's a lot of pickups, whether you pick them up on Ebay or not. I'm thinking noise issues may be a factor. I'm wondering if you might get some humbucking effect if each row alternated a RWRP pickup with a regular SC. Mind you, I don't know if that would have any effect given this mega-array, but it ought to do so.

Second, I've never tried winding my own pickups, but it can be done, and it might be more efficient, as well as cost effective, than 12 SCs. You could just wind one (or 2) realllyyy looonnnggg pickups.

Third, you might want to consider using steel guitar pickups. There are those for 10-string pedal steels. These are bar-style pickups, so positioning the strings over pole pieces is not an issue. Since there's a bit of "free play" in the magnetic fields, you might be able to space 3 of these across and get full coverage of 36 strings.

The only electric autoharp I've ever heard was the Electric Prunes (Listen to "I Had Too Much To Dream Last Night"). Don't know how they amplified that, but I suspect it was a piezo or a contact microphone, not 12 SCs . . .

Is there some sort of autoharp forum you could consult? Maybe this is old science to someone.

Sorry, that's all I've got. Further input will be forthcoming.

EDIT: And ChrisK's post and mine were like 2 ships passing in the night . . .

Edit 2.0: And I glossed over the part where you said you'd already bought the pickups! Oh well, what Ebay bringeth EBay can taketh away.

[ChrisK]

You could just wind one (or 2) realllyyy looonnnggg pickups.

Pickups are wound with #42 to #44 gauge wire. #42 has a resistance of about 1.5 Ohms per foot.

A Fender single coil pickup has a resistance around 6K (for the sake of this argument). That's 4,000 feet of wire that is 0.003" in diameter....................................

[ashcatlt]

Not to say you're crazy,but...

Why not a couple clavinet pickups? Scroll down a little.

I saw a guy play electric slide zither once. He got a tone that reminded me of Thoroughgood...

[ChrisK]

That's an excellent idea.


This infers another question; what is the string spacing on the harp?

[D2o]

Nov 17, 2008 23:26:11 GMT -5 newey said:

OK, we've had some strange queries around here, but this is fresh!

Indeed it is!

We have had loosely related inquiries, like this zither - I wonder if some of the wiring discussion there may also help?

D2o

[janusjones]

Wow - thanks for all the replies, everybody!

First, yes, the pickups are all identical. As for polarity, I don't know - I suppose I can test it out. The only reason I was thinking of 12 (instead of just 6) is that traditional electric guitars have two to three per set of strings - presumably for a fuller sound from the strings. I was going on the assumption that I'd need two per set to get a really rich effect (which I want).

String spacing will be the same as on a normal autoharp - all in all, the strings will take up about 12 inches of space. I've done the measurements and it worked out.

As for switches, I'm not positive I need them - just thought if I were going for more, I could use the switches to play with options - shutting off upper registers for a bass-boosted sound, shutting down bass for a more treble sound, etc.

I should get at least one bass pickup, honestly; the last four strings are bass strings. The low F is a bridged bass E string; the high C is a bridged upper register guitar C. Each string will be bridged to span the range of the three octaves; just like on a regular autoharp, there will be a few notes missing from the bottom octave. I've figured out exactly which strings to use and where to bridge them, so it won't be a problem.

About string tension, I checked with a luthier (for the Old Town School of Folk Music here in Chicago!) who said my plan (to use two planks of 3/4 inch birch ply glued together) would work for the string tension - no problems there.

One tone control? One master volume? What do you think?

As far as amped autoharps, I have one. Accoustic pickups tend to be too sensitive and pick up vocals; the electric ones tend to come installed. I am building this as an art project; also because I want something that plays like an autoharp, but is better suited to rhythm - has a fuller, bass-heavy sound and more chord options (42 will let me play just about anything I want to!).

Steel guitar pickups sound kind of sweet. I'll look into them. Same with clavinet pickups . . . though they look EXPENSIVE . . .

Thanks for all the suggestions! I'm hoping to figure out the wiring and such soon . . . any thoughts?

[ashcatlt]

Nov 18, 2008 11:58:04 GMT -5 janusjones said:

Same with clavinet pickups . . . though they look EXPENSIVE . . .

Yes, but you only need one or two. How much did you pay for each of these 12 identical pickups? If it averaged $20 a piece, that's right around one clavinet, no?

The multiple pickups on a guitar are there for tonal options. It's not so much that running more than one at a time makes the sound fuller. In fact, I think most would agree that the mid+ positions on a strat are generally thinner sounding than either of the pickups by themselves. The thing is that you pick up a different mix of overtones at different positions along the length of the string. In general, closer to the bridge will be brighter - with more of the higher harmonics and significantly less of the fundamental when compared to a pickup at the neck position. This is something I'd suggest you look into. It's going to take some math, or a lot of experimentation. You're starting from scratch here, so you don't have to worry about pre-routed holes or anything. If you do go with 12 pickups you've got a lot of room to place each one in the place that sounds best for the particular strings it's catching.

You will want to know the relative magnetic polarity of all these pickups. Doesn't matter which end of the earth they want to point to, all that matters is how they relate to each other. As ChrisK said, the wiring polarity is usually easy to change in the wiring. Unless, of course, they are all single conductor plus shield type pickups.

A normal humbucker in a guitar has 2 coils of opposite polarity very close to each other. This is best because one would expect each coil to be capturing somewhere very close to the same EM radiation. We know, though, that strats and teles, when wired correctly, reject a good amount of noise in the "in-between" positions, even though the coils are physically much further apart. I think that Newey was talking about having the humbucking come from pickups paired down the line (horizontally as opposed to verically, sort of) and this seems to work just fine on Precision style bass pickups.

As to how to finally wire the thing... 12 (or even 6) pickups in series is going to be a very big impedance, and likely to be a very dark sound. In parallel, the impedance gets very small, and I'm thinking may be a bit too bright, brittle, or thin. I'm wondering if we might be able to come to some scheme where you'd have 2 groups of pickups wired in series, and then the 2 groups in parallel with one another.

I don't know for sure that you'll be satisfied if you're switching off pickups under different strings. Those strings are not likely to come through with any meaningful volume on pickups not directly underneath them, so they'll essentially be muted. Switching between pickups in different positions along the length of the string (as mentioned above) would give a range of different sounds. Of course, then you'll have to address the overall volume differences between the various pickups. If the pickups are identical, we'd expect lower overall output from those mounted closer to the bridge because the strings just can't move as far in that area. This could probably be compensated for to an extent by pickup height (distance from string).

I sure hope you don't get the idea that we're being negative or trying to rain on your parade. This sounds like fun and I think we're all just trying to help you think through all the details so it works out well for you.

[newey]

In terms of planning where to position the pickup(s), you may find this helpful:
Donald Tillman's Pickup Response Applet.

This is from our "Links Page . . .Sort of" which is under the reference articles section of the board. Tillman's article explaining pickup response is also there if you want the theory behind this.

The Applet is of course designed to sim a guitar, but it looks to me like his parameters are broad enough so that they could be set to the areas where your autoharp will be operating.

This could help you to determine optimum pickup placement for one or two arrays of pickups.

Ash- I was thinking of obtaining some hum reduction in adjacent pickups but also between the 2 rows as well. I envisioned the two rows as being like this:

Pickup row 1: NWNP-RWRP-NWNP-RWRP-NWNP-RWRP

Pickup row 2: RWRP-NWNP-RWRP-NWNP-RWRP-NWNP

Of course, that depends on having a sufficient number of RWRP pickups, and we don't yet know exactly what he has.

On switching, Ash is right, you'd be muting out certain sections by switching individual pickups in the array on and off. If you use the 2 rows of pickups, switching between row 1/both/row 2 would make some sense in terms of giving you the tonal options of the 2 separate rows.

Master vol and tone should be fine, as long as all your pickups are identical, as ChrisK noted.

As far as series vs. parallel, and the resulting resistances, Ash makes a good point, it's an issue you need to consider no matter how you string 6 pickups together. I don't know how you would go about avoiding a "too dark" all-series "meta-pickup" vs. a too-bright all-parallel set up. Buffering the output, maybe? Or, perhaps Ash's suggestion of a combination of series and parallel, although that would probably give you differing sounds along the line of pups.

[ChrisK]

The Tillman model is flawed. As one moves the pickup on either side of the 12th fret, the response should be an identical/mirror image. This is not the case in his model.

Note that when the pickup location goes past the bridge to the right, the response IS mirrored about the bridge. They must be mighty powerful magnets.

I mentioned this to him some time ago, he agreed.


Also, while the model intends to show the frequencies detected for a given position and sensing window width (which is not changeable), one cannot easily discern the response as a function of the harmonic multiples of any given note, for more than a few harmonics.

The best way to do this is with a spreadsheet that calculates the sines of the fundamental and all interesting harmonics, and plots them as well as sums them for a given frequency and positional ratio of string length.


In essence, for any given reasonable position inside a string end, you can't get all the harmonics.