New homemade electric clavichord - need electro-magnetic pickup expertise

[gotkovsky]

Hello everyone,

I've been working for a while on a new electric clavichord design. Here are a few pictures of the build:






I'm currently working on the pickups and would love a little help from forum experts, as electronics is not exactly my forte.

I want to design a first humbucking bridge pickup which for now, looks like the following:

1 — The 3D printed plastic bobbin.


2 — 10 cylindrical magnets inside each bobbin.


3 - Copper coil.


4 - 4 bobbins in a row, bobbin 1 and 3 contains north facing magnets (in blue), bobbins 2 and 4 contains south facing magnets (in red).


5 - A single and long pole piece of steel joins all the bobbins.


6 - Schematic view of the wiring that I came-up with. "S" means start of the coils, "E" end of the coils.


I've read articles about pick-up design and understood that in order for two coils to be in a humbucker configuration, one the coils had to be reverse-wound and one of the magnets had to have its polarity reversed. I based my design on this, but I don't know if the fact that my pickup design contains only one long pole-piece which contacts all the magnets will make the whole thing not work. I can always make 4 pole pieces but one long piece would be easier.

I'd be glad to discuss this design with folks here. I kind of need to know if this could work before I spent time building this  . As it's probably pretty simple for some of you here, I thought I should ask.

Thanks in advance, and don't hesitate if you wanna know more about this project.

Cheers,

Kevin

[Jenzz]

Hi :-)

Interresting project !

There are two things that came to my mind: I think all magnets should be facing the same magnetic (sounth) pole. Otherwise, the magnetic fields cancel out at the turnover points, resulting in loss of fundamental tone for these strings.

A humbucking effect can only be achieved when both coils 'see' the same amount hum /magnetic field, so the inverting coil configuration can cancel out the hum and leave the signal intact. Due to this, both coils should be as close to one another as possible (hence the configuration on humbucking guitar pickups).

Jenzz

[gotkovsky]

Hi Jenzz, thanks for the reply, you are absolutely right, the two coils have to "see" at least the same strings, not necessarily at exactly the same spot, but it should be somewhat close (like on a classic electric guitar humbucker which have two stacked or side-by-side coils). Not sure how I completely missed this in the first place 

After a few iterations I finally came up with this design:



It's basically a single coil pickup divided in 4 sections for ease of production. Each bobin has its coil, winding is clockwise. All the magnets (40 in total) are facing the same magnetic pole and linked via a single pole piece (4 mm steel rod). The bobbins have been slightly enlarged to accomodate 2000 turns of 0,1 mm copper wire. I'm planning on wiring the 4 coils on a DPDT switch to test various configurations (parallel vs series, etc.).

I'll also make a wooden cover shield, with its inside faces coated with copper tape connected to ground:



And it will probably be fit like this on the harp:



 
Cheers,

Kevin

[Jenzz]

Hi :-)

If you wind the pickups yourself, you might could incorporate a sort of extra winding in each of the bobbins only for the hum canceling.

This came to my mind as i remember the Bartolini Bass pickups for Music Man Stingray basses. These feature a 3-coil concept were the middle coil has no magnets, only to pick up the hum and do the canceling by the way was it is connected to the other two coils.

Jenzz

[gotkovsky]

Hello!

I'm almost done on the electric clavichord, here's a quick picture of it:



And here's the final version of the pick-ups. I finally came up with a single-bobbin per note solution, like a Rhodes:



Each pickup measures about 230 ohms, and the total resistance measured at the output is about 1100 ohms. I'd like to increase the output volume which is still quite low.

Does anybody have any ideas on how to do that? I must say that every mechanical adjustment to increase volume at the source has already been made, so I'm looking for an electronic solution, i.e a different wiring or a preamp suggestion/design. Or a transformer maybe? The pick-ups are already as close as possible to the strings, and the hammers material can't be changed too.

Here's how the 49 pickups are wired right now (all bobbins are wound in the same direction and all magnets are facing north pole up). I copied the Fender Rhodes and my previous homemade electric piano, so I guess this wiring is supposed to be hum-cancelling, but I'm really not sure:



Would I get a better signal to noise ratio and output volume if all the pickups were wired in series, like this?



Asking because rewiring the whole thing would involve quite some work, so if anybody has advices and/or ways of calculating expected results, that would be super great 

[sean]

I think if you wired them all in series, you would lose the treble, and it might sound tubby.  Plus, putting MORE resistance in series with each pickup might decrease the output of each individual pickup.

I would try groups-of-six instead of groups-of-threes.  There is a sweet spot:  if you have too many in a group, the output is reduced because of all the other pickups in the group create an escape route for the current in the pickup making the sound.  Groups-of-six work fine in a Rhodes. 

Do you have room to add a second magnet?  Or can you get stronger magnets?

Have you tried making a pickup with lower DC resistance?  Wind it with thicker wire?
Have you tried more windings (higher resistance, but more turns gives more current)?

I bet the simplest solution is simply a little more gain at the preamp.
----- -----

Oh wait!  Here is another idea from my history of harp splitting:  split the pickup rail in half, and send each half to a simple mixer.  I got more volume from the harp when it is cut in half, so it might work for you.

It would be easy to test, just short out half the harp with an alligator clip lead.  The remaining part of the harp should be slightly louder than before.


In any case, your clavichord looks insanely beautiful.

Sean

[gotkovsky]

Sean, thank you so much for the kind words, once again!

To answer your questions:
—I don't have room to add a second magnet.
—I'm using neodymium magnets which I've heard are pretty strong.
—I didn't try to make pickups with less DC resistance, neither by using less windings or thicker wire. I just used the stuff that I had which worked on my previous e. piano (0,1 mm wire and about 3000 turns).
—I don't have room to make more windings, the bobbins are side by side and already very close to each other, if not touching.

I tried splitting the pickup as you suggested and got so-so result. Plus this is not very practical as I intend to use this clavichord live, plugged into a guitar amp.

Do you think I'll have a significant result if I rewire the pickups in groups of sixes instead of three? I've made a new diagram, do you see any mistakes? Asking because rewiring implies removing the strings and the yarn, which is a lot of work.



As stated before, my goal is to get a better signal-to-noise ratio. I will also try to shield everything that I can, but i know there's a limit to what I can do because the top of the pickups facing the strings can't be covered. I can surely increase the input volume at my guitar amp (Roland JC50) and get a good volume but there's A LOT of hiss and EMI interferences. Would it help to use shielded cable between the pickups, by connecting the shield of the cable to a kind of ground rail alongside the pickup? I used a shielded cable for the long minus connection at the left of the pickup rail, btw.

Thanks again, hope you can help,

Cheers

Kevin