Studies in wandmaking - Fitting electronics inside a straw [UPDATE 2020-07-25]


Active Member
For the last few years I've had the idea of making an illuminating wand kicking about in my head. Looking around the internet I found a few creators who have done just this, and mixing their ideas and techniques set me on the path. Here's to giving back. I've tried to document this properly after that fact, as I didn't take the time while working.

My goals when setting out was to include a few features: to have it be made out of wood with as little visible damage to the grain as possible, replaceable electronics and a tip covered in thin veneer, all inside a body thin enough to fit the profile of the movie replicas.

Also the sun, the moon and the rainbow in a basket, please?

The idea of drilling out a dowel seemed bound to fail. Even if a drillbit would make it through without going out the side, the wood wall would be too thin and fragile to hold up. If I made the wall thicker there would be no room inside for the electronics, or the wand would be too bulky. This was the blockade that had me stuck for a long time in the planning phase, and it was sitting at a convention that I suddenly realized a possible solution.
Taking inspiration from bentwood ring tutorials, where wooden veneer is put on top of metal piping, I could theorically extend the structure to a hollow covered tube.
So I set about sourcing materials for my first prototype.


The battery is the least flexible part of the design, and I looked at many battery types before I settled on using a 8 mm diameter AAAA as the power source. Then I bought 0.5 mm walled brass tubing in RC shops, ranging from 4 to 9 mm outside diameter. Aiming for a 10 mm outer diameter at the thickest, this allowed me a covering of 0.5 mm veneer, which turned out surprisingly workeable.
Looking for premade circuitry, I found plenty of arduino based prototyping boards available. None were even close to small enough to fit inside my required footprint, which on top of being a measly 7 mm wide was also round, so the effective headspace for any components Still, I bought a few boards that had tiny components and simple layouts, because I had a plan...
A 3.3 V regulator, a capacitive touch sensor, a PNP transistor, a variety of bright 3x3 mm SMD LED's and a few variable regulators to test it all rounded out my materials list.

The most frustrating part of the early electronics was soldering the LED's, as the pads were so small and the bulb tended to overheat and burst from my shoddy work. My soldering probably improved a lot from doing this stuff, but I'm glad that I had the sense to buy lots of backups. Finally I got one to work, and it was easily strong enough to shine through the veneer I planned to put on top. The battery life suffered a bit, but I was having fun with it.

Moving onto the chassis, I cut, soldered and polished the brass piping into a stepped tube. Keeping it fairly straight was the hard part, as the small space between the pipe diameter was enough to make them go pretty wonky.
The brass pipes would also allow me to pull the ground through the wall itself instead of using a loose wire on the side of the battery, saving crucial extra space.

2_Brass wand.jpg

Next came fitting the electronics inside of the brass chassis. The PCB's were small, but a far cry from fitting inside my 7 mm inner diameter requirement. So, using a handy little hacksaw, I set to frankensteining my solution.


Marking up the components and creating a simple diagram of connections, I carefully took to sawing through PCB's, cutting them clean in half and trimming off the edges as much as possible with a dremel sanding drum.
After gluing the bisected boards in a straight-ish line, I then resoldered the broken connections back together. This was even more time consuming than the LED's, and the wires I used were far too big. Switching to thin winding wire later on helped solve the issue, but early on I had way too much material in the way.
After hours of work I had the circuit lighting up in its new, slimmed down form. Using winding wire I recreated the broken capacitive sensor coil in a footprint that would fit snugg against the chassis wall.
I now looked for a way to securely mount it inside the wand. I needed something that could act both as the positive battery pole, and also allow some kind of tool to hook onto it, to remove the circuit from inside the wand in a controlled way. I hoped for premade solutions, but nothing fit the bill or the footprint. This kind of had me stumped for a while, but in the end my dad suggested a bikewheel spoke nipple, and this worked out excellent.
After some grinding, gluing and soldering, the design had a working lower connector combined with anchoring point. More superglue on top of it ensured that it was isolated from the grounded wall, and it looked like my prototype was all set for its first test assembly.

4_Gen0 Boards.jpg

Covering it in hot glue seemed like a good idea at the time, to provide insulation and give it a flexible press fit , but this turned out to be a major misstep.
This is also when I ran into my first major unseen issue: I had spent hours carefully making a circuit that fit inside the brass wand, only to find out that capacitive sensors don't work behind conductive materials. At all. The press fit was also far too tight, and the glue joint I had made was not strong enough, so the circuitry broke off inside of the wand. I had to use a fair bit of force to get it out, and then carefully repair it, now covered in thermal glue. Not fun.
After taking a frustrated break, I got back to work. If the problem was conductivity, then simply removing the piping over the sensor was of course the way to go. So my trusty hacksaw made a reappearance, helped along by a diamond wheel for the dremmel.
I also quickly made a battery ground plug out of brass tubing and a brass motherboard standoff screw, plus a top diffuser out of some acrylic rod.


After assembling it all, the prototype lit up for the first time as a self contained unit, free from the breadboard.

6-Gen0 Lit up.png

This had me really excited. I had managed to fit all I needed inside. From here, I was sure that it would be smooth sailing, and that most of the really difficult work was behind me.
Oh boy....
So I started on plugging the sensor window. It couldn't be left open, since the veneer might break without a solid backing, and the coil inside might hook onto the wall and break, or get the circuit stuck again.
A plastic pen came to the rescue, and after some sawing and shaping, it provided a non conductive covering to the sensor.

7- Bic pen window.jpg

Finally, I started work on the wood covering.
I used the same kind of technique that the bentwood ring makers do: boiling/steaming wooden veneer and then bending it around a similar size tube to dry overnight. This worked out well, although it proved difficult to get an easy bend onto the thinner piping without breaking it.

8- Bentwood veneer.jpg

Having that done, sanded down the edges, and then set about glueing it to the brass using CA glue.
This was also time consuming and annoying work. As I had no jigg to hold it in place, I had to use my hands to keep pressure onto it as it dried. The thinnest part of the chassis was tricky yet manageble. It was when it came time to cover the steps between the thicker pipes that problems arose. I had covered the lower parts and sanded down a smooth curve, but the slight compound curves and not-quite-straight pipes caused the veneer to crack in several places as I applied it.
Not exactly a clean, untouched wooden finish, but I did manage to seal it up and puzzle it all toghether fairly well. After sanding, It actually looked like a wooden wand, although it was a bit thicker at the top than I had originally planned.
Still, it worked.

9- Gen0 v1 finished.jpg
10- Gen0 v1.2.jpg

I finished this just in time for a convention. In fact, the last glueing was done while sitting and waiting at my table!
Driving back for 8 hours gave me a lot of time to think about my design, and I set out making a better one almost right away.
Repeating the same process with the second wand, I adjusted the top pipe thickness to be 4 mm OD instead of 5, to give it a more pointed look. The circuit also was far easier, as I had learned to use winding wire to save on space. CA glue insulated it and held it together more firmly than the hot glue from before.

011- Gen0 v2 Finished.jpg
011- Gen0 v2.2.jpg

The second wand worked and looked far better overall, veneer transitioning more smoothly between the different layers and the circuit being more reliable.
Still, it also showed me the limitations of the current design and techniques.

For one, the work on each wand took way too long. One circuit board was almost a full day of soldering and glueing on its own, and although I likely could have gotten more efficient, the cost of the materials and likelihood of me messing up were way too high for my taste.

The second issue was the sensor. While the plastic window did work, it required precise measurements in creating each one, you had to insert it exactly right each time or it wouldn't function at all. It also made each wand circuit custom made for its chassis. On top of that, the veneer worked really well; well enough to make the sensor completely invisible from the outside, which might have been what I wanted, but it also made finding the activation spot very difficult unless you really knew what you were looking for.

Thirdly, the veneer application was really messy, and the result wasn't as nice as I had hoped. I'm sure that it could have been made better, if I practiced a lot and gave each wand enough time, but the concept was flawed from the beginning. It was clear that it would never give the finish I was aiming for, it limited me to wood types available in veneer, and it barred me from staining the woodgrain as the glue seams would leave blotches and show through.

The success did give me a platform to stand on, but it was time to go back and improve on the foundation of the core idea.

More on the way.
Last edited:


Active Member
Thank you.

I'm trying to find the time to photograph the second iteration, but things are a bit hectic at the moment.


Active Member
Having something to work off of was far easier than starting from nothing.
I took the wands apart and tried to analyze each individual piece, to see if I could improve both the end result and make the build process easier and more reliable.

The metal tubes had seemed like a great idea at the start, but more and more they proved to be less than helpfull. The sensor would always be a custom job with those. If I could instead make the middle part out of a nonconductive pipe while keeping the battery wall conductive, I would have the best of both worlds.
8 mm fibreglass tubes looked promising, so I bought some and tried to insert them between the top and bottom pipes. They fit well, but the inside diameter was 6 mm and I just couldn't take more off of the pcb's; I had already shaved off the top corners of the components to fit them, so any room would have to come from the pipe.
Despite fibreglass being a very sturdy material, boring an tube to widen it didn't work out. The ones I could find in the right size were extruded, and the drilling deformed and cracked them along the length, which removed a lot of their sturdiness.
Still, could probably be salvaged, so I considered this a step up.

11- Gen0 v2.2_5.jpg
11- Gen0 v2.2_6.jpg

Moving onto the veneer, I really wanted to improve the look. Early on I had written off drilling solid wood as a method. It just seemed impossible, but if there was the slightest chance that I could do it the end result would look a lot better.

I bought the longest drillbits I could find, and taking information from penturning I tried drilling the wood blanks in my lathe. This taught me that modern twist drills are pretty much useless for making very deep holes straight and true. What I wanted was considerably longer than a fountain pen, so this was not unexpected, but a bit dissapointing.

Researching long bore drilling technique references, I found information on flute making, and a book by recommendation on a woodworking forum.
"The amateur wind instrument maker" had illustrations of jiggs and special drillbits, custom crafted to make the pilot holes for flutes, along with some other helpfull advice about working with woodgrain that will come in handy for a future wand..

11- Gen0 v2.2_9.jpg

I bought myself silver steel blanks and then hand filed a simple nose auger in 4, 5 and 6 mm diameters, and armed with those I went to it again.

The nose auger is great for staying centered. The slanted cutting edge doesn't want to follow the grain and the solid body and matching neck of it gets support from the hole itself as it's being drilled.

Their downside is that they can't clear swarf out as they go, which turns out to be the main culprit for them missaligning. That is why modern twist drills were invented: to clear the cut material away as it goes, so I reused the twist drills to clear out the bore as I went.

11- Gen0 v2.2_3.jpg

It was still difficult, but after several tries I managed to make a few blanks with fairly straight holes.
It could be done.

11- Gen0 v2.2_12.jpg
11- Gen0 v2.2_13.jpg

I'll update with the next steps as soon as I have time photograph them.


  • 11- Gen0 v2.2_1.jpg
    11- Gen0 v2.2_1.jpg
    439.3 KB · Views: 101
  • 11- Gen0 v2.2_2.jpg
    11- Gen0 v2.2_2.jpg
    343.3 KB · Views: 111
  • 11- Gen0 v2.2_4.jpg
    11- Gen0 v2.2_4.jpg
    437.3 KB · Views: 104
  • 11- Gen0 v2.2_7.jpg
    11- Gen0 v2.2_7.jpg
    489.1 KB · Views: 116
  • 11- Gen0 v2.2_11.jpg
    11- Gen0 v2.2_11.jpg
    505.6 KB · Views: 103


Active Member
The plan was to fit the brass and fibreglass pipes inside like in penturning, but they wouldn't go. The brass tubes didn't work great because of the conductivity anyways, but it was my only plan up until then, so I was stuck.
After a lot of thinking I decided to try switching materials, and fibreglass reinforce the inside.
I had never laid fibreglass before, didn't know if it was possible in a space like this, or if it would be strong enough to hold the wand shape, but I had come this far so I might as well try everything.

11- Gen0 v2.2_14.jpg

Using old brass pipe material leftovers, I made a polished last to shape the inside diameter, filled the pipe with two part liquid epoxy and let it sit in a short test piece. After removing the last, I used more piping to make a mandrel to support the hollow tube on the lathe, since it might wobble and break otherwise, and turned it down carefully as far as I dared.
It turned out great and I moved onto a full scale test.

11- Gen0 v2.2_15.jpg
11- Gen0 v2.2_17.jpg

Using more of the old brass piping I had left, I made a full size last and mandrel, and glued up a proper wand blank this time.
I also bought some collet chucks to use for holding the mandrels perfectly centered. Having it centered is crucial, as the wood covering will be turned so thinly that any wobble or misalignment would cut right through to the fibreglass.

11- Gen0 v2.2_20.jpg
11- Gen0 v2.2_23.jpg
11- Gen0 v2.2_24.jpg
11- Gen0 v2.2_28.jpg

Getting the last out of the longer blank was a real struggle, but after a long fight I could turn it on the lathe. I almost cut through the side and left an imperfection on the side of the handle, but it was a test piece and it provided valuable practice for later. While it dried I had been making more lasts and had set more blanks to be reinforced, so I only had to wait a day for my next attempts. They were designed to open before the battery, to avoid ruining them the same way. I want to return to this fully covered design later, but I need more precision in my tools before I can attempt it.

11- Gen0 v2.2_29.jpg
11- Gen0 v2.2_30.jpg

Onto electronics!
I had a shiny new way of making wand flutes which the PCB's I used would never fit, so I moved onto the internals.

Designing my pcb, I based it off of the NCP1402 and TTP223 but in a radically thinner format with different components to save space.
All printed leads sit inside a 4 mm footprint to allow shaving off the sides as much as possible, and to keep current away from the sensor coil.

The most difficult part turned out to be the inductor. To fit inside the old 7 mm diameter they needed trimming down with files as much as possible, and I had set my mind on trying for a 5 mm inside diameter. Because of the small scale and the square-cube law doing its thing, that 2 mm difference pretty much halved the space inside, so the old inductor just wouldn't do.
After tons of sourcing for the right specs inside of one suitable package, solution was simply two smaller inductors put together. Go figure.

11- Gen0 v2.2_32.jpg

On top of that I upgraded the end anchor. Now it's a brass motherboard standoff, covered later in shrinktube and then a ground layer made up of brass piping set around it.

11- Gen0 v2.2_33.jpg

Taking some Valnut and Hazelnut wood, I made two flutes.
I set a brass pipes into matching wood handles and turned those down to size, put some brass bands and ebony as a decorative pommel on the valnut, and carved a pattern into the thicker handle on the hazel.

11- Gen0 v2.2_35.jpg
11- Gen0 v2.2_37.jpg
11- Gen0 v2.2_38.jpg
11- Gen0 v2.2_39.jpg
11- Gen0 v2.2_40.jpg

With this, my second try is finished.
There are things I want to improve and new things to try out, but the basic techniques work and I have the components for several more, so the next attempt will be some character wand replicas.


Active Member
Beautiful. I can imagine a curved wand would be a challenge and would require a completely different approach to the external material.


Active Member
Thank you.

A bent wand is the next thing to attempt after I make an elder wand.
I thought it would need a different technique too, but here is a section in the book about crumhorns and how to create bends in drilled flute blanks, so it might be possible, and I'm happy to try.
Right now the thread is caught up to where I am, so the next posts will be experiments and results.


Sr Member
Another thread I'd missed. Love how you leveraged flute making skills into this build. Look forward to seeing where this goes next.


Active Member
That's really awesome! Any way you can make a sorta shorter how-to for us electronic noobs?
Thank you!
That depends a bit on what information you're after. There were parts of this project I had no idea about, like sourcing components, but maybe you're looking at a specific project?
I'm actually not some great electronics wiz, and learned most of the stuff through using arduino style prototype boards and online tutorials. I can point you in the right directions if you like but I'm mostly a noob myself, which goes to show how far you can go with patience and googling.

Another thread I'd missed. Love how you leveraged flute making skills into this build. Look forward to seeing where this goes next.

Flutemaking being the key component was a surprise for sure.
The prototype elder wand will be glued up tomorrow, and the day after that it'll be turned and shaped if nothing goes awry. As soon as I have things photographed the thread will get a picture heavy update.

Depending on how that goes I'll either make another elder wand or try for a bent one.


Sr Member
These are beautiful! I made myself a one off light up wand for one of the premiers out of that same brass tube, a pen light, and the “wood” being shaped from miliput and that was enough of a headache, this is a whole different level of craftsmanship.


New Member
Thank you!
That depends a bit on what information you're after. There were parts of this project I had no idea about, like sourcing components, but maybe you're looking at a specific project?
I'm actually not some great electronics wiz, and learned most of the stuff through using arduino style prototype boards and online tutorials. I can point you in the right directions if you like but I'm mostly a noob myself, which goes to show how far you can go with patience and googling.
I'm mostly talking about getting & assembling the electronics for a light up wand like this. I have a habit of making paper-based wands and I think that this can kick them up a notch


Active Member
I'm trying to document as I go now. Not as nice of a photoshoot but what can you do.

Buying a replica of the original I measured and calculated, and then started digging up materials from my storage.
The tiniest battery I could find, and it might not light up properly but we'll see.


Birch for a first prototype, with a bit of moose antler for the glyph piece.
The top wood is split because I'm reusing otherwise faulty pieces, and the top part needed to be 4 mm inside instead of 6.
If this proves succesfull I'll attempt to drill a single piece, but first I hope to hide the seamline in the darkened knobbly portion
Custom last made the same way as before. Steel core and brass pipes stacked, polished and covered in Marbocote 227.
One issue is that the fibreglass sleeve fitting. The space inside the thinner portion is so small that the top sacrificial part needs to be saturated in CA glue and filed down to go inside, and even then the sleeve catches and bunches up due to how small it's compressing.
I'll do a test with a scrap piece to see if the epoxy makes it smoother. Otherwise I'll have to devise something using ultrathin kevlar weave instead.

Reporting back tomorrow with updates.


These are beautiful! I made myself a one off light up wand for one of the premiers out of that same brass tube, a pen light, and the “wood” being shaped from miliput and that was enough of a headache, this is a whole different level of craftsmanship.
Thanks! Using a penlight is an excellent idea! Do you still have it?
I'd love to know how you attached the pen light to the brass. Glue or some bridging piece as well?
Did you extend the diode up to the top using wires or some other technique?

I'm mostly talking about getting & assembling the electronics for a light up wand like this. I have a habit of making paper-based wands and I think that this can kick them up a notch
Sourcing first. In summary it went like this:

To get my component limitations I first decided on an interior dimension. Diameter is the limiting factor for these, so that was where I had to start.
Since I knew that I wanted try making an elder wand, I went for the smallest diameter (5mm) I thought possible, and then calculated the component dimensions + pcb thickness to fit in that.

I went with OSHpark since they were very helpfull and their 0.8 mm PCB option was a good balance between optimizing and price.
You need the PCB to be 0.8 or bellow to fit inside a 5 mm diameter, unless you want to go with very small components, or shape the board and components themselves using files (not recommended).

Knowing the maximum width and height, I made list of components from the prototype boards I had used before and searched for those one at a time, filtering for dimensions below my limits, but above sub mm since they would be too small for me to handle currently.
Digikey and Mouser Electronics are great sites that you can use for finding and buying component, but if they don't sell certain ones you can take the product codes to other sites to get everything you need.
I ended up having to buy one chip on ebay because it was delisted from the sites where I could find everything else, and it was only sold in packs of a thousand otherwise.

With components arriving, I used a free PCB design software to create the blueprint to send to OSHpark. There was a bit of back-and-forth to get the thing just right, but it was mostly copying and pasting measurements into windows and being patient.

When the PCB's arrived I tried to solder by hand, but it was a mess. So I turned to reflow soldering.
Essentially it uses a paste that you smear on the copper pads on the PCB with a specialized cutout stencil, then heat and let cool in a controlled fashion using a special reflow soldering oven. I used an old business card for a stencil and a heatgun, and the result worked just fine.
I would recommend soldering all the cards in one go if you have the option though. Reflow paste has a short shelf life and is a bit costly, but it works much better than regular soldering for these small components and applications.
Aside from that, remember to first clean the copper pads using a fibreglass brush and solder flux. Prepwork is the difference between clean soldering being impossible and it being laughably easy, so deoxidize that copper before you start.

With the PCB assembled, you need wires between it and the LED, plus a structural mounting point underneath that allows both negative and positive to connect without short circuiting. The way I've done it is the brass pipe holding the battery grounds it, and there's a smaller pipe inside that fits it snugly. Inside that there's an insulated brass standoff screw that connects to positive.
I still haven't found a way to do it within 8 mm diameter without serious custom crafting, so if you come up with something let me know!

All of this could potentially be made easier if you meant to mount it permanently inside though. It's me wanting things to be repairable and modular that complicates things.

If you have any further questions just ask and I'll try to answer.
Also, if you want to try things out with my design I have some 1.6 mm PCBs in the same design that was wrongly manufactured due to a miscommunication. There are no surplus components so those you'd need to source, but if you want I could send you a few boards to experiment with.


Sr Member
It’s packed away currently but I might be able to dig it out and get a picture. If I remember correctly I think I attached the pen segment to the brass with masking tape but then it was reinforced with miliput (this was a quick build). The pen segment holds a single AAA battery which is then wired to a small momentary switch buried in the sculpt, and from there the wires run up the brass to the single LED. It’s definitely not as sophisticated as your approach lol.


Active Member
The test was a failure. The fibreglass couldn't take the strain and the last broke through.
After cleaning up and salvaging what I could, I've taken some time to think.
Clearly brute force will not solve this, so maybe I'm going about it all wrong and there's another way or something that I'm missing.

Improving my workshop is keeping me busy and gives me time to think it over, and there are a few ways that have come to my mind while cleaning today. After I get more superglue I'll do some experimenting with the process and try again.

There's definite value in working quick, and it seems you made a lot with a little.
Did you put the switch at the hilt on the pen part, or on the brass part. If the later, did you drill into it to get the wires inside?
Maybe it's not as complex but it resulted in a functioning illuminating wand, and that's some serious macgyvering.

Again, I would love to see it if you find it! I wouldn't have been able to make this without relying on the work others put online, so it's not like I came up with it all on my own.


New Member
Also, if you want to try things out with my design I have some 1.6 mm PCBs in the same design that was wrongly manufactured due to a miscommunication. There are no surplus components so those you'd need to source, but if you want I could send you a few boards to experiment with.
That would be super cool, if you can send to my country and give me a list of components


Sr Member
We’ll see if these pictures work, it’s showing up weird on my phone. I screwed off the end cap to get a look at the batteries and I’m not sure if they’re even changeable lol, past me didn’t care about future me apparently. Now that I’m looking at it again I’m remembering it was made from a dollar store pen with a light in it, so I just built the battery compartment into the handle and added the wire to move the LED out to the end of the brass tube. I think the switch was a surface mounted momentary switch from a broken headlamp. All the surface detail is miliput and some wood-esque paint work.


  • 470F5341-3A0E-424F-9FEC-B102DF38827A.jpeg
    1.4 MB · Views: 104
  • 4532E217-8D73-4830-A0BE-5194A0320A6F.jpeg
    1.5 MB · Views: 95
  • 1F06F100-56EB-497B-8EC3-CC3184182F02.jpeg
    1.6 MB · Views: 89


Active Member
Sorry for the wait. Turns out my parts list is on my other, older computer that's not with me right now. I'll try to compile a list out of the parts numbers but it will take a little while longer.

Credit where it's due, from those pictures that sculpture and paint work would 100 percent have fooled me into believing it's carved wood.
Handling something gives a different feel of course, but you did a lot of good work with those materials. I especially love the smooth transition to the lighted tip. It looks really good!
Did the button have a cover on it? I ask because it looks like there's a sunken divot around it to house a larger cover button, like a broken off branch or similar.

You've got me itching to make some of my own! While I wait for the glue to arrive I'll see if I can do some carving on another one.

Your message may be considered spam for the following reasons:

  1. Your new thread title is very short, and likely is unhelpful.
  2. Your reply is very short and likely does not add anything to the thread.
  3. Your reply is very long and likely does not add anything to the thread.
  4. It is very likely that it does not need any further discussion and thus bumping it serves no purpose.
  5. Your message is mostly quotes or spoilers.
  6. Your reply has occurred very quickly after a previous reply and likely does not add anything to the thread.
  7. This thread is locked.