Free-Formed Solar Chirping Bird Pendant Using 0603 SMD Components (BEAM Electronics)

I had one weekend between finishing quite a stressful university term and starting to study for my exams so I thought, "You know what would be really relaxing, finally trying to make a free-frormed circuit using 0603 components.". No way that could be stressful right? 😂😂

The final circuit chirps on and off for about 20-40 seconds (watch the video to hear it chirping) and then stays off for another 1-2 minutes before chirping again.

I originally just wanted to make this as small as possibly but in doing so I realised it could actually be made as a pendant and I am very glad I turned it into a pendant in the end as I think the frame adds so much to the circuits aesthetic and protects the circuit from getting damaged. I don't intent to wear this often (in fact I hardly expect to let it see the sun at all because, as I know from previous chirping circuits, the noise gets old quick) but I reckon this would be the absolute perfect thing to wear to a maker faire or similar event!

In the final photos in the introduction you can see my first attempt to miniaturize this circuit by designing a PCB back in 2019 (this was one of my first PCB projects and it doesn't even work, the solar panel is too small!). As well there is a comparison in size between the breadboarded version, PCB, and free-formed SMD version

Edit: wobbler has kindly amplified the chirping audio as it was a little quiet in the original video. If you are struggling to hear the chirping in the video try the audio file attached below. It is still very true to how the circuit sounds in real life just louder and clearer =).

EDIT: I don't know what happened here but my supplies section seems to have disappeared at some point so I will summarise the important components again.

Resistors: All 0603.

Capacitors: 0603 for the 1nFs and the rest had to be 0805 as they are much easier to source. I also used a couple of axial 10nF caps.

Diodes: 1N4148WS.

Photodiode: I used a VEM6010X01CT-ND though any choice will probably do and you may be able to find smaller with a more suitable pinout.

Solar panel: KXOB25-02X8F-TB which is a very small and efficient 5.5v panel. Quickly becoming a favourite for me!

Super capacitor: 100mF, 5.5v square super capacitor from KEMET.

Piezo element: 9x9mm externally driven piezo speaker.

Inverter IC: 74HC14 SOIC (hex Schmitt inverters)

Brass: 0.02" (0.51mm) for most of the circuit. 3/64" (1.19mm) for the frame 1/32" (0.81mm) for the loops.

Tools: Most important for this project is a clean soldering iron tip, flux, and good solder. I used multicore 60/40 solder at 0.7mm which was fine but at times I did want thinner solder. Also important for soldering are fine tweezers, Blu Tack, and perhaps some form of magnification (I was lucky I did not need magnification for this project as I am not used to working through a microscope or anything so it gives me headaches). You will also need various jewellers pliers (I bought a set from my local electronics hobby store that work great) and sids cutter for forming the brass.

This is yet another circuit that was originally designed by Wilf way back in 199 (found on the solarbotics.net site). I have modified it a little, mostly to change the frequency of chirps, and a few other modifications to make it sound less robotic.

The circuit is essentially just 4 oscillators. From left to right, the first and second oscillators turn on and off slowly to turn the other oscillators on and off. In Wilf's original circuit this creates some kind of randomness in the number of chirps but the circuit is constantly chirping. With the components I chose, the circuit turns off for a very long time (1-2 minutes) and is on for a shorter time (20-40 seconds) based on the second oscillator values (10uF, 10M, 5M). The first oscillator values (1uF, 10M, 3.3M) controls that little pause you hear in the video between sets of chirps.

The third oscillator controls how long each chirp is and the distance between each chirp. By placing a light dependant element in that path we can make the chirps faster and slower. Wilf's original circuit used an LDR but I tried to replace it with a photodiode as I don't think SMD LDRs are too common. I had some trouble with this and the final circuit doesnt get affected as much as with the LDR so if I were to attempt this again I would try find a SMD LDR.

The final oscillator creates the chirp frequency which I left as is.

I have also added a diode between the third oscillator and the frequency generating oscillator which quickly shuts down the chirping. I found that without it, the chirps would "linger" which sometimes sounded ok and sometimes sounded awful and robotic.

The only other modifications I made were changing the 2k resistor to 1k (idk why it was 2k to begin with but I have more 1k SMD resistors at my disposal) and adding a capacitor across the piezo element to take away some of the harshness from the chirps. It does lower the volume of the chirps a little and using too large a capacitance can drastically alter the sound so this capacitor would probably be better placed elsewhere but I wanted a little less volume anyway. In the end I used 10nF but I think you will have to play around here as my larger piezo speaker had none of that harshness and I only needed this cap for the smaller speaker.

The panel and main storage capacitor are just in parallel with Vcc and GND of the circuit.

The final image shows how everything is laid out in the free-forming.

First of all, let me start by saying this was very hard to solder so I wouldn't recommend this project unless you feel comfortable you can give it a fair shot. Even if you are comfortable, buy many extra parts, the photo above shows all the components I was able to find laying around on my desk after finishing this project. Counting the parts I dropped, I reckon I would have wasted more than twice the number of components I actually used in the end.

One of the difficulties of free-forming at this scale is obviously the size. I found that with some practice soldering 0603 components and smaller on regular PCBs by hand, actually soldering these components wasn't too bad. That is, provided you have a fine soldering tip, fine tweezers, and think about the order you solder components.

Holding everything is also difficult. I actually never use helping hands because the jaws are way too big so I use blu tack, my fingers, and tweezers to hold everything while soldering.

The real difficulty is heat. As everything is so small and so close together that soldering one side of a component will quickly heat the other side and any other joints nearby. This means that everything has to be done in specific orders so that the brass wire can be supported as you solder components to it such that if any of the nearby joints do melt, the whole circuit doesn't fall apart.

I did a couple tests to see what the best way to solder such small free-formed parts was and during that testing I discovered that making the brass wire much longer than it needed to be and then supporting the brass wire with solder joints far away from where you intend to be making the actual connections makes the brass wire rigid enough to solder to.

There were a couple points where I made mistakes and had to jerry-rig some kind of support before I could remake any connections.

Hence, the general approach to soldering everything together is to support all the brass somehow, solder one side of the component to the brass using tweezers to hold the component in place. Then once one side is temporarily soldered (I say temporary because the joint is likely to be cold), you can solder the other side in place. Then go back and clean up the first solder joint. Once a component has two connections or a wire has two connections that are far enough apart, it is pretty solid but you still have to be careful with how much heat you put into the joints.

The other thing that helps is keeping everything super clean. This means using flux and a clean soldering tip, and then cleaning any flux residue after each joint.

The final technique I used sometimes is to soak a q-tip in isopropyl alcohol (water is probably safer but alcohol evaporates faster after the fact), and then place the cotton on one side of a component or on a piece of brass wire. This helps keep the side of the component or wire cool as I heat the other side of the component or I heat the wire nearby. Again, you still have to be very careful with the heat but I feel like it did help.

The final thing I wanted to mention before starting is that normally I like to build these circuits in sections so I can test them. Most logically for this build you would want to build the final tone generating oscillator, test it with the speaker, build one of the other oscillators, test it, connect them, test it.... etc. But due to the nature of everything being so compact I mostly just had to put down the next most logical component instead of building in blocks so it was longer between testing sections. As such, I don't really have great names for each step in this Instructable so hopefully the progress pictures I have included of the layout help you understand what is happening. It is still quite confusing because the schematic is a top-down view but we build the circuit underneath, or "upside-down". I also made a couple of mistakes, so let me know if anything is not clear.

In the picture above, on the right, you can see my first attempt to see if soldering these tiny components free-form was even possible. On the left you can see that I did test the final oscillator before starting but seeing as it was the "easier" side to free-form, on the final circuit I had to leave that section until last. All just things to think about if attempting a similar project.

Everything is free-formed around the 74HC14 chip so we will start there. First I carefully bend all the pins up as this is the strongest in the end based on my inital testing but you have to be careful not to weaken the leads too much or they may snap. Then I placed the IC in blu tack and soldered the brass wire on. Again. the joints don't need to be good at this stage, once everything is held in place with a small piece of wire soldered along the top of all of the brass, then you can go back and make sure the joints are good.

Components are soldered in a similar way to how I would do hand soldered SMD components on a PCB. First I pre-tin the component and/or brass. Then I place the component where I want with a pair of tweezers in one hand and reflow the solder with the soldering iron in the other hand. Once that is secure I carefully solder the other side properly and then go back to the first joint and make sure it is not cold. Lots of flux helps here.

The pre-tinning only needs to be done on one side, and the component will lay flatter if you only tin one side, but with how close everything is I often found it difficult to apply solder to the other side once the first side was soldered so I tinned everything.

Note: I found that the frequency of chirping with the 510K resistor was too infrequency so I ended up stacking a second on top in the end for ~250k

This first diode is soldered in much the same way and then I cut a chunk out of the brass for the next diode. Before cutting the wire you have to unsolder it from the support at the top because otherwise the side cutters will push everything apart and probably crack components (ask me how I know). That red sharpie is just a little extra assurance that I wont accidentally solder the brass where I don't want to. It is very easy to accidentally put solder in places you don't want which spoils the aesthetic a little.

I only bought 10M 0603 resistors so to achieve this 5M (and later 3.3M) resistor I stacked resistors in parallel.

Incredibly, this tiny SMD photodiode was the largest component and I almost couldn't fit it in. It also doesn't alter the speed of the chirps very much so if I were to attempt this again I would probably leave it out or find an SMD light dependant resistor.

This diode is added in the same way as the previous.

Next I added the 3x parallel 10M resistors to act as a 3.3M. I then carefully removed the small bit of brass connecting the 3.3M to the diode and replaced it with a bent section of brass.

This was my first mistake as the connection that actually needed to be bend across the chip was the lower section of brass.

Next I added the 0805 capacitors and connected them to ground. I left the ground wire long for now.

Now we can finally do a test to see if anything is broken so far. To do so I connected a wire to the positive side of the 74HC14 and connected it to a power supply. Then I can probe the outputs with an oscilloscope to see that they are changing. these oscillators are slow enough that you could test the outputs with a multimeter.

The second side of brass was added the same as the first however I did not solder the length of brass that was to go to the piezo speaker and left it long as I did not know how I was going to make that connection yet. This ended up being a poor idea as it was very easy to bump the long wire and bend the lead on the IC.

I added the 1K resistor and first 1nF cap the same way I had done most of the other 0603 components and then made a bent section of wire to connect to the other side of the circuit. In attempting to make bends like this in situ I often broke the components and had to resolder everything anyway so it may be better to make any bent connections before putting them on the circuit even though it will not be as supported when you solder to it.

This bend is in the wrong place so this was now my second mistake.

This is where I noticed my first mistake from "Step 10: Free-Forming 7". Even though I only had to replace two sections of wire, because everything was so close together, it all had to come apart and everything needed to be supported again to remake the connections.

Because the top connection was not just going to be between the diode and 3.3M resistor, I did not think that was going to be strong enough and I could easily bend the resistor and diode connection. So I ended up separating the 3 parallel resistors out which provides a little more strength to that point of the circuit.

I then soldered a diode across the bent sections of wire I had made across the IC.

This is where I noticed my second mistake from "Step 14: Free-Forming 10". Again it was quite a big mistake and warranted removing all the other connections I had made before I was able to fix it.

After fixing this mistake I put back most of the brass wire and the 1nF cap but left out the 1K resistor and piezo connection as it didn't need to be there yet and I kept bending it accidentally.

Next, the two 100K resistors were added. I first soldered them together by placing them on some double sided tape (the tape lost stickiness quick with the heat but it helped a little).

Then the 1uF capacitor to connect the slow oscillators to the final tone generating oscillator was added. This was done now because it will be hard to reach after the next steps

Next, some bends were made and the final 1nF capacitor was connected.

And one final diode completes the bulk of the circuit.

Now I can connect the positive wire back again as well as the piezo element and check that everything works!

I thought the finished circuit looks pretty cool as is so here are some pictures.

I didn't yet know how everything was going to go together so I did a little prototyping with blutack and double sided tape. I originally wanted a section of brass behind the speaker as I think, being designed to be mounted onto a PCB, the sound comes out the back of the piezo speaker a little and it sounds better with something behind (not too sure how that is actually working, this is just what I found while testing).

I used my limited knowledge of jewellery making from watching youtube videos to build the pendant body. First I bent a square with one side overlapping and cut through both overlapping sections with a fine jewellers saw to arrive at my final square shape. I then soldered the joint shut (I made two of these). Next I added four upright supports. These are not soldered super well yet, just like the components, I can go back and fix up the solder joints once the other side is secure.

Next I made two loops and soldered them to the other square and then soldered everything into a rectangular prism.

I thought the pendant actually looked really nice without the circuit even so here is a quick video. I wouldn't trust it not to break though, it is very weak structurally =P.

Now we can finally start connecting everything together.

I first bend the negative connection and then started on the positive. I made the positive connection connect to the negative with a 100nF 0603 capacitor just for mechanical stability because otherwise the positive is only connected to the circuit at one spot. Then I can bend the positive as desired.

Next I soldered the circuit to the panel and my god, it is finally starting to look like something. I used a small dab of super glue to hold the IC in place, I don't think this is necessary but does make it a little more solid.

The storage capacitor was then connected to the positive of the panel with a short length of brass and the negative was bent up to make contact with the pendant body which we will connect to GND in the end. I also super glued the capacitor to the panel.

This was one of the more difficult sections to bend as there is not much supporting this wire at this stage and there is not much room to hold things securely as I bent them. Honestly this is another connection where I probably would have been better off taking it completely off, forming it, and then soldering it back on.

The second speaker connection was then added. This one I did bend most of the way first and the later bends I can support with pliers. I connected another 1nF capacitor to both of the speaker connections for support.

As I mentioned, we will connect the body of the pendant to ground as it is helpful to route the last leg of the capacitor through the body, it provides more stability, and we may as well connect it to something. This small section of wire accommodated for this connection.

The pendant body is then connected in three places.

1. The negative connection from the previous step.
2. The negative of the capacitor.
3. A 10nF axial capacitor connects the positive of the panel to the frame. Like the 100nF capacitor in step 28, this does not affect the circuit and is just there for added mechanical stability.

I did add a 10M resistor from the second speaker connection to ground as it was a little flimsy. This shouldn't affect the circuit at all and I even made sure to put it on the speaker output that is normally grounded. The speaker has a 10nF capacitor connected across it and the only reason I am using an axial capacitor here instead of a SMD one is because I didn't think to buy any 10nF SMD capacitors!

Once again, I have undoubtably forgotten things, so feel free to ask questions.

I do eventually want to coat the circuit to keep it shiny but it is currently cold and humid in Sydney, very poor conditions for spraying a coating, so that will have to wait.

I also wanted some kind of on off mechanism but even 2 pins and a jumper was too bulky so I will have to think on that

Well, the weather report got it wrong again and we had a very hot day today. Was still very humid but I was eager to coat the circuit so the brass wouldn't go dull. I removed the piezo speaker because I couldn't figure out how to adequately protect it from overspray and then covered the contacts in electrical tape so I could solder to the after spraying. I initially let it cure in a box to protect it from dust and then left it in the sun all day. The datasheet says to bake the coating for maximum protection but I don't think that is feasible for me.

I used clear silicone conformal coating from Electrolube. This was also a test to see if this would be a suitable coating for my BEAM head in a jar and I have a couple of notes:

1. Cleaning all of the flux residue was very hard, took a long time, and required the use of IPA and an aerosolised PCB cleaner. I should really invest in an ultrasonic cleaner already.

2. Interestingly the datasheet says make absolutely sure you do not shake the can as it introduces bubbles. This is really counterintuitive for me and is something I almost missed.

3. The spray went on well but there were two tiny bubbles on the panel I couldn't quite fill (and I didn't want to spray it on too thick so I just left them).

4. The actual circuit looks perfect, as if there isn't even any coating on them. It looks so good I couldn't even justify posting photos of the coated circuit because it looks exactly the same. I can tell they are coated well as the coating has a UV sensitive additive so you can use a blacklight to check for full coverage.

5. This UV additive seems to be visible in sunlight unfortunately, it makes the brass a little "whiter" and the solar panel turns blue.

6. While the spray is invisible on the circuit, it is quite wavy on the flat sections such as the panel and super capacitor faces. I expect this from a can and all in all I am very happy with the new coating assuming it is durable.

I also don't think I have ever posted this circuit anywhere so figured I should share it here now. This was my first free-formed chirping circuit that I think I built in late 2018. The amorphous cell means this starts up very early in the day. I added a potentiometer across the piezo element to control the volume and 99% of the time it stays off because the chirping gets annoying =P.