Cutting Electronic Components in Half: A Cross-Section Tour Under the HIROX HRX-02 Microscope
イチケン / ICHIKEN
Ichiken's question in this video is a simple one: what do everyday electronic parts actually look like inside? To answer it, the channel ground through capacitors, inductors, resistors, fuses, crystals, switches, semiconductors, and a printed circuit board, then examined each cross-section under HIROX's new HRX-02 digital microscope. HIROX sponsored the video. Ichiken had used an earlier HIROX microscope in a 2023 video, and here they also report how the new model felt to use. Throughout, the parts were prepared by grinding them down with a rotary grindstone, and the claims below about design intent are mostly Ichiken's own readings of what the cross-sections show.
Multilayer Ceramic Capacitors: Layer Spacing, Odd Stacking, and Tiny Packages
Capacitors come first because, as Ichiken puts it, they appear in almost every electrical circuit. The MLCCs cover a range of dielectrics: high-permittivity materials with X7R characteristics that boost capacitance, and temperature-compensating C0G types that are less affected by temperature and voltage.
The first sample is a 1 µF, 450 V MLCC, slightly larger than the 2220 size. Its cross-section is what Ichiken calls textbook: the internal electrodes overlap in neat alternating layers. The measured spacing between layers was about 16 µm, and Ichiken considers stacking layers that close with that accuracy a remarkable feat of manufacturing.
A 1 nF, 500 V MLCC comes next. It has about the same voltage rating but one-thousandth of the capacitance, and it looks quite different inside. It isn't simply left and right electrodes alternating. A floating electrode sits in the middle and connects to nothing. The stacking density also varies: dense at the top, sparse in the middle, medium at the bottom. The regions without electrodes still show visible traces of the stacking process. Ichiken takes this layout as evidence of considerable manufacturer know-how in how electrodes are placed and spaced.
At the extreme low end, a 0.12 pF, 50 V C0G capacitor has no overlapping electrodes at all. The internal electrodes simply face each other end to end. Ichiken's interpretation is that overlapping them would give far too much capacitance, so the facing arrangement is how such a small value is achieved. This sample also shows the external termination clearly: a three-layer structure of copper, nickel plating, and tin plating, again described as textbook.
Next is a part that is small in physical size rather than capacitance. Murata's 0201M-size MLCC measures 0.25 × 0.125 mm. According to Ichiken, it was the smallest size they could find until 2024, and an even smaller 016008M size (0.16 × 0.08 mm) is now available. Parts this small are easy to mistake for dust once they leave the reel, and hand soldering them is impossible. Under the microscope, the tiny body still has a proper internal capacitor structure. Ichiken sees this as a strong demonstration of Murata's technical capability.
A Three-Terminal Capacitor
A three-terminal capacitor, which has lower ESL than a regular MLCC, shows a different design. In the diagram Ichiken presents, two sets of internal electrodes cross over each other. In the actual cross-section, the left and right external electrodes connect to one set of internal electrodes. The middle electrode, which looks floating in this view, turns out to connect to the side external electrode when seen from above. Ichiken shows it as an example of MLCCs with special internal structures.
Electrolytic, Tantalum, and Film Capacitors
The electrolytic example is a conductive-polymer aluminum solid electrolytic capacitor. Ichiken explains that electrolytic capacitors of all types are built in a broadly similar way: an anode, electrolyte, electrolytic paper, and cathode are rolled up and placed in an aluminum can. A top-down cut shows the roll, and a transverse cut shows the layers overlapping. The black material looks like paper and holds the conductive polymer serving as the electrolyte. The gray material is aluminum foil. Ichiken points out that the two electrodes can be told apart. The cathode is plain aluminum-colored foil, while the anode shows a gray layer sandwiching the aluminum. That gray layer is the oxide film formed by surface-treating the foil, and it is what gives the capacitor its withstand voltage.
Tantalum capacitors are also a kind of electrolytic capacitor, but their electrolyte is solid, where the familiar aluminum electrolytics use a liquid one. Inside, a tantalum wire runs through the middle. The square block around it carries tantalum pentoxide as the dielectric, and together these form the anode. Manganese dioxide serves as the solid electrolyte, and the cathode is built from graphite and silver paste. The square block is porous, which Ichiken notes is hard to see on video, and this porosity is why tantalum capacitors offer large capacitance for their size. Ichiken adds that despite good electrical characteristics, tantalum capacitors fail short-circuit. For that reason they are being replaced by MLCCs and polymer electrolytics, and they are seen and used less and less.
The last capacitor is Rubycon's PMLCAP, a chip-type film capacitor. Conventional film capacitors are made by rolling. PMLCAP is stacked like an MLCC instead. Ichiken describes it as a capacitor with very good characteristics, sometimes used in high-end audio, and correspondingly expensive. The cross-section shows extremely thin layers of roughly 1 µm film. Ichiken compares this with the film in ordinary film capacitors, which they put at a few to several tens of micrometers.
What Changed in the New HIROX Microscope
Midway through, Ichiken covers what is new in the HRX-02 compared with the previous product. The high-resolution camera and a revolver lens unit are now integrated. A 10–80× macro lens and a 100–500× objective lens come standard, and up to four lenses can be mounted so the user can switch to whichever suits the subject. The maximum lens magnification is 10,000×. Ichiken calls that the limit of optical microscopy and says that, in their experience, the effective magnification is about 5,000×.
Lighting has also changed. The previous model needed a separate light source. The new one combines light source and lens in a single unit, which reduces the work of adjusting and connecting the light. A left-hand controller handles zoom, focus, stage movement, light source changes, and other shortcuts. Ichiken particularly liked it. Without it, the stage and focus are controlled by flicking on the screen, and they found the amount of mouse operation inconvenient. With the controller, operation felt easy and intuitive.
The stand now has a vibration-absorbing mechanism in the foot, which reduces shake at high magnification, and it can also be used with conventional lenses. The body can connect two cameras at once, and previously owned cameras and lenses can still be used with it.
Through-Hole Inductors and Ferrite Beads
The drum-type inductor is the first inductor example. Magnet wire is wound around a core and soldered to the leads, and the cross-section reveals an I-shaped core. Ichiken explains that this shape makes the coil easy to wind, but magnetic flux leaks out from the core's ends. Depending on the current, the inductor itself can become a noise source, so its placement and core shape have to be chosen carefully. Other inductors are built to keep the flux inside, as later samples show.
The second through-hole part looks like a resistor but is an inductor. Inside, there is a magnetic core with a coil wound around it. It also uses an I-shaped core, similar to the drum type, but the core and windings are tightly sealed in resin. Ichiken reads this as a sign that the part tolerates harsh environments, and warns viewers not to confuse it with a resistor.
A third inductor has a lead wire passing straight through its core, making it effectively a single-turn inductor. This is a ferrite bead. According to Ichiken, ferrite beads are often placed where high-frequency noise is generated, such as in switching power supplies, to absorb that noise.
Surface-Mount Inductors: Measuring Wire Cross-Sections
Two surface-mount inductors that differ in rated current and inductance make the next comparison. Their windings are visibly different in thickness. The thicker winding belongs to the higher-current part. Ichiken uses the microscope's split-screen function to view both at once. They also show a mapping feature: after capturing a low-magnification map, clicking on it at high magnification moves the stage to that spot.
The microscope can measure area. Tracing the copper wire of the higher-current inductor gives 187,869 µm², about 0.18 mm². The lower-current inductor's wire measures in the mid-0.03 mm² range. Ichiken notes that the wire looks thicker on the outside of the winding than on the inside. Their guess is that the grinding applied too much pressure and stretched the wire.
From this, Ichiken explains that the wire's cross-sectional area, together with the current characteristics of the core's permeability, determines rated current. Winding count affects inductance: thin wire can be wound many times, while thick wire allows only a few turns.
Ferrite-Resin Shielding, a Manufacturing Void, and Depth Measurement
Another surface-mount inductor looks structurally similar but has its windings encased in resin mixed with ferrite powder. Ichiken explains that this gives magnetic flux an easier path than air, so less magnetic noise leaks out and nearby components are less affected.
This sample also has a hollow in it, which shouldn't be there. Ichiken says they were told it is a manufacturing variation, and air bubbles are also mixed into the resin. To measure the hollow, they use the microscope's depth composition function, which uses image processing to measure depth. The cavity turns out to be roughly 500–600 µm deep, and a color map shows the surface relief at a glance. Ichiken attributes this capability to a combination of optical accuracy, mechanical precision in the stage, and image processing.
A metal-alloy chip inductor has its windings solidified in magnetic material. Because that material acts as a magnetic shield, less flux leaks out. Ichiken mentions that multilayer chip inductors, built by stacking layers, also exist, but they didn't obtain one for this video.
The Variable Inductor
Ichiken notes that variable capacitors are fairly common, but variable inductors rarely come up, perhaps mainly in wireless circuits. Inside the example, ferrite cores sit both inside and outside the case, with a winding between them. The inner core is threaded, so turning it moves the core and changes the coupling between the inner and outer cores, which adjusts the inductance. Ichiken calls it a well-thought-out design.
Resistors: Carbon Film, Solid, Multi-Turn, Cement, and Chip
The first resistor is a 1 kΩ through-hole carbon film resistor, a staple of hobby electronics. Ichiken explains that these are coated with a resistive film and then trimmed with a laser or grindstone to reach the target value. The cross-section shows a white ceramic core with a carbon film on top, metal caps at each end, and a beige resin coating over everything. The film and core show evenly spaced grooves from trimming. Resistor specialists Ichiken consulted said this one was trimmed with a whetstone.
A second resistor looks like an ordinary carbon film part but turns out to be a solid resistor. Its black, crumbly center is the resistive element itself, a mixture of carbon and resin. Ichiken lists its characteristics as high pulse tolerance and, because it isn't cut into a coil shape, low inductance. They add that solid resistors are becoming hard to find even when taking apart appliances.
The multi-turn variable resistor contradicted Ichiken's expectation. They had assumed multi-turn trimmers used a longer resistive element. Instead, the part combines a screw and a meshing gear. Under the gear is the resistive element, and a metal leaf spring on the gear's back contacts it. The screw-and-gear reduction shrinks the wiper's travel per turn, and that is how fine adjustment is achieved.
Cement resistors are used in high-power applications, often as discharge resistors, because of their heat resistance. The outer shell is hard ceramic, but the cement inside is soft enough to crumble when poked with tweezers. The resistive element is high-resistance metal wire wound on a ceramic core. Ichiken says it won't burn out unless made to dissipate a very large amount of power.
The chip resistor consists of a resistive film on a ceramic substrate, protected in this case by green resin. With the coating removed, the black resistive film shows an L-shaped trim cut, where a laser burned away material to set the value. Ichiken explains that the vertical cut adjusts the resistance roughly and the horizontal cut fine-tunes it.
Fuses
The glass-tube fuse contains a wire that breaks under overcurrent. A thicker wire carries more current. A second fuse looks like a resistor but contains a similar wire. It is soldered to the board, so replacing it after it blows means desoldering and resoldering. Ichiken finds that a hassle and considers holder-based fuses like the glass tube more maintainable.
A high-power fuse turns out to be filled with sand, which Ichiken explains is there to extinguish arcs and is common in high-voltage, high-current fuses. Its current path is a metal strip with holes punched in the middle. Ichiken says they don't know exactly why. Their guess is that the holes raise the current density so the strip reliably breaks at that spot.
Crystal Oscillators: Same Spec, Different Shapes
Quartz crystals produce vibration with a fixed period and are used in quartz watches and to clock microcontrollers. The first is a 32.768 kHz watch crystal. Its internals are extremely delicate, and Ichiken broke several while removing them from the can. The one shown is a survivor. Inside is a tuning-fork-shaped crystal with a thin metal film on its surface. Applying voltage makes it vibrate at an invisible scale, and that vibration serves as a clock source. The frequency depends on the shape.
That sample was from Citizen. Ichiken compares it with a Seiko crystal of the same frequency and load capacitance, and the crystal shapes are very different. Ichiken is careful to say they don't know much about quartz crystals and aren't claiming either is better. The point is that parts with similar behavior can differ in shape between manufacturers. A 17 MHz crystal, by contrast, is not a tuning fork at all but a simple quartz plate with thin metal films on both sides. Ichiken assumes the cut of the crystal and the film shape involve considerable know-how.
Switches
The DIP switch was cut open mostly out of curiosity. Its cross-section matched Ichiken's expectations: moving the white plastic slider with a click pushes a metal piece to close the contact. Where the plastic rubs the metal, the metal has a visibly different color and a smooth surface, which Ichiken suspects is a surface treatment to reduce wear on the plastic and contacts.
In the toggle switch, moving the central plastic part shifts the metal contact pieces and changes which contacts conduct. The plastic sliding surfaces are coated with what looks like silicone grease to reduce wear. The metal contacts have transparent grease, presumably to improve contact durability. With the grease removed, the contact surfaces were surprisingly rough, not smooth as Ichiken had expected.
Metal-Can Semiconductors: Diodes
For semiconductors, Ichiken chose metal-can packages because their contents are easy to see. Most packages today use epoxy resin, but can packages are still made. Ichiken lists their advantages as excellent hermetic sealing, good heat dissipation, and the can itself acting as a shield. Inside, a silicon chip sits at the center, and the external leads extend into the can and connect to the chip with bonding wires.
The first diode has an extra cathode pad on the chip that isn't connected to anything. The same product in a different package brings out two cathode leads, so Ichiken guesses the pad serves that version. Some pads show damage. Ichiken explains these as marks from the probes of a semiconductor tester used after fabrication. A green line runs around the chip's perimeter. Ichiken wonders if it is a guard ring and asks viewers who know to comment.
A dual-circuit RF diode, probably from the same manufacturer, looks similar. It is wedge-bonded on both the chip and lead sides, with whitish wire that Ichiken takes to be aluminum. The maximum current on these lines is around 20 mA, and they speculate that gold wire would be unnecessarily expensive for that. The two diodes have slightly different wiring lengths, with one having a longer anode electrode. Ichiken wonders whether RF characteristics are involved but considers the difference tiny. In their view, lead length and PCB trace length would dominate.
MOSFET, Power BJT, and an Op Amp
The N-channel MOSFET has gate, drain, and source pins plus a fourth lead connected to the case. Its chip is more complex than the diodes and looks green, purple, blue, and beige. Ichiken explains that the chip isn't actually those colors. The colors come from light interference in the deposited films, the same effect that makes soap bubbles look rainbow-colored.
The NPN bipolar transistor is much larger, rated 60 V and 15 A, and is the kind often seen bolted to heat sinks on the back of old audio amplifiers. The chip, case, and bonding wires are all large and thick. On the chip, strip electrodes interleave: the base electrode surrounds the outside, and the stripes inside are the emitter. Current flows vertically through the chip, so the case itself is the collector, which Ichiken describes as typical of power semiconductors. The chip is soldered to the case so heat escapes efficiently.
The last can-packaged semiconductor is Texas Instruments' LM201 op amp. Ichiken notes that the can version is very expensive compared with other packages, about 3,000 yen each. The chip visibly resembles an op-amp circuit. Ichiken tentatively identifies the differential input stage, some resistors, and what might be output-stage transistors. They suggest viewers pause the video and compare the layout with the datasheet schematic.
A MEMS Microphone
As an unusual semiconductor, Ichiken shows a relatively large MEMS microphone and notes that smartphones use several even smaller ones. A hole on the back of the package lets air vibrations in to move an internal membrane. Opened up, the round diaphragm is made of silicon. Sound makes it vibrate, which changes a capacitance, and reading that change converts sound into a signal. A dedicated ASIC handles sensing and analog-to-digital conversion. It sits under a black epoxy-like potting. Ichiken finds the idea of making a microphone from a silicon membrane impressive.
Ichiken's Own Op-Amp Chip at the Optical Limit
The final semiconductor is a chip Ichiken designed. With support from ISHI-kai, an open-source semiconductor community, they made an op-amp circuit on a shared chip, with separate areas of one die allocated to different participants. The process rule is 0.6 µm and the chip is 1.8 mm square. Ichiken says the price was lower than expected: 20 bare chips for a little over 200,000 yen. The details are in a separate video.
At high magnification, individual transistors are visible, but the image gets harder to read as magnification rises. Ichiken attributes this to the limits of optics and puts the usable limit at about 5,000×. Even at that level, image analysis can show surface relief, such as slight raised areas where metal layers are. Ichiken also points to the vias. They were drawn as squares in the design but came out somewhat rounded. Having been through the fabrication of dissolving layers with chemicals and gas, Ichiken says it makes sense that the corners get eroded.
A Six-Layer PCB and Pad-on-Via
The last subject is a printed circuit board, which people usually see only from above. The six-layer board's cross-section shows the solder resist on top, then the surface copper, then inner copper layers, all on an FR4 substrate of glass fiber and epoxy. The vertical copper sections are vias, and this board uses pad-on-via, meaning the copper pad sits directly over the via. Ichiken thinks this helps greatly in high-density layouts.
Pad-on-via requires filling the via. The one shown is resin-filled, which is why the surface copper has no dimple or tear. Ichiken explains that the fill material affects the board's properties. This board uses resin mixed with copper powder, which conducts heat better than plain resin and also carries current more easily, but costs more. Ichiken recommends copper fill when a via needs to carry current or move heat to the other side of the board.
Closing
Ichiken ends by noting that a microscope can reveal what is inside disassembled parts. They add that it can also show, on the spot, how finely product quality is maintained in manufacturing. They name quality assurance staff, developers, and researchers as people with many opportunities to use one. Several questions raised along the way stay open: what the green ring on the diode chip is, why the high-power fuse strip has holes, and whether the small wiring-length differences in the RF diode matter.
This video is brought to you courtesy of HIROX. Hello. Today we will use a microscope to observe the contents of electronic components. We use this HIROX microscope. In fact, we also used a HIROX microscope in a video in 2023. This time, a new model has been released. We would like to look at electronic components together and see how it feels to use the microscope. We will observe electronic components such as the resistors and inductors shown here, multilayer ceramic capacitors, and semiconductors. Let's get started.
First of all, these parts. We will look at capacitors, which are used in almost 100% of electrical circuits. There are many different types of capacitors. Today we will look at MLCCs, or multilayer ceramic capacitors, as well as electrolytic capacitors, film capacitors, and tantalum capacitors.
First is MLCC. There are various types of MLCCs. Some have X7R characteristics and use high-dielectric materials to increase capacitance. Others have C0G characteristics and use temperature-compensating materials that are less affected by temperature and voltage. We have prepared a variety this time, so let's take a look.
First of all, here is a 1 μF MLCC with a withstand voltage of 450 V. Therefore, the size is a little larger than 2220. In this case, the sample is prepared using a rotary grinding stone like this. If you actually use a microscope to look at it, the internal electrodes of the MLCC overlap in layers as shown here. The structure is textbook. What is the distance between layers? It was about 16 μm. Stacking them at such short distances with this accuracy is a hell of a technique.
We will look at other MLCCs. This is an MLCC with a withstand voltage of 500 V and a capacitance of 1 nF. It's about the same withstand voltage as before, but the capacitance is 1/1,000. What is the structure inside? It has a slightly strange lamination. It's not simply a matter of overlapping the left and right electrodes. In the middle, we have an electrode that doesn't lead anywhere. The density of the laminate also varies from place to place: dense at the top, less dense in the middle, and medium density in the lower area. Also, it's easy to see with this MLCC that traces of stacking are clearly visible even in areas without electrodes. There must be considerable know-how behind the placement, density, and structure of the electrodes in these areas.
The next one has an even smaller capacitance. This is a 0.12 pF MLCC with a withstand voltage of 50 V and C0G temperature characteristics. The inside is structured like this. There is no overlap between the MLCC electrodes. The structure is such that the internal electrodes face each other. Maybe overlapping them would make the capacitance too large. I think they create a small capacitance by making the electrodes face each other like this.
On this capacitor, the colors of the electrodes are easy to recognize, so I'll go into a little more detail. It has a three-layer construction with copper, nickel plating, and tin plating. The structure of this area was textbook.
We will look at this next. This is a small MLCC. It's not the capacitance that's small. We are looking at something smaller in the physical size of the MLCC itself. It is a Murata 0201M-size component, measuring 0.25 mm × 0.125 mm. Until 2024, this was the smallest size we could find, but now an even smaller 016008M size, measuring 0.16 mm × 0.08 mm, is available. These are really small, and if you take them out of the reel, you can't even tell there's an MLCC there. They can easily be mistaken for dust. At this size, hand soldering is impossible. The structure looks like this through a microscope. It's really small, but the inside has a proper capacitor structure. Being able to make such a small part so well is a great achievement. I think it speaks volumes about Murata's technical prowess.
Next we will look at a slightly different MLCC. This is a three-terminal capacitor. Its ESL is smaller than that of a regular MLCC. Structurally, as shown in this illustration, overlapping internal electrodes cross over each other. In the actual cross section, the left and right external electrodes are connected to the internal electrodes as shown here. The internal electrode that appears to be floating in the middle is actually connected to the external electrode as shown here when viewed from above. Some MLCCs have special structures like this one.
Next, we will look at these electrolytic capacitors. This is a conductive polymer aluminum solid electrolytic capacitor. Basically, all types of electrolytic capacitors have a similar structure. The anode, electrolyte, electrolytic paper, and cathode are rolled up inside an aluminum can. In the actual cross section from above, you can see that it is rolled like this. Looking at the transverse cross section, you can clearly see the overlapping layers. The black part looks like paper. A conductive polymer is used there as an electrolyte.
The gray area is aluminum foil. If you look closely, you can distinguish the cathode from the anode. The ordinary aluminum-colored foil is the cathode. The anode has aluminum in the center, sandwiched between gray layers. The gray portion of the anode is an oxide film created by surface treatment of the aluminum foil. This oxide film ensures the withstand voltage of the aluminum electrolytic capacitor. They're pretty clearly distinguishable.
Next, tantalum capacitors. This is actually a kind of electrolytic capacitor. The electrolytic capacitors you may often use have a liquid electrolyte, while tantalum capacitors have a solid electrolyte. The internal structure looks like this. Running through the middle is a tantalum wire. The square part is tantalum pentoxide, the dielectric. So far, that's the anode. Manganese dioxide is the solid electrolyte, and the cathode consists of graphite and silver paste. It's hard to tell from this video, but the square part has a porous structure. Therefore, tantalum capacitors have a large capacitance for their size.
Tantalum capacitors have good characteristics as capacitors, but their failure mode is short-circuit failure. Therefore, they are being replaced by other capacitors, such as MLCCs and electrolytic capacitors using polymers. We see and use them less and less these days.
Last of the capacitors is Rubycon's PMLCAP. It is a chip-form film capacitor. Common film capacitors are made by rolling them up, but PMLCAP is made by stacking layers like an MLCC. PMLCAP itself is a capacitor with very good characteristics. They are sometimes used in high-end audio and other applications, but that's why the price is so high. The actual cross section looks like this. The layers are very thin. The film thickness is approximately 1 μm. Since the film thickness of a common film capacitor is a few micrometers to several tens of micrometers, PMLCAP is very thin.
Let's look at a few things about HIROX's new microscope. There are a few things that have changed compared to the previous product, and we will introduce them to you. First, here is the high-resolution camera and revolving lens unit. The camera and lens unit are integrated into one piece. A 10–80× macro lens and a 100–500× objective lens are standard. A total of four lenses can now be mounted. Appropriate lenses can be used according to the object being observed. The maximum lens magnification is up to 10,000×, but 10,000× is the limit of an optical microscope. After using this product, I thought the effective magnification would be about 5,000×.
Lighting has also evolved from previous models. The previous model required a light source. The new product integrates the light source and lens into a single unit. There is less hassle with adjusting the light source, making connections, and so on.
This new product comes with a left-hand controller for operation. This is pretty good. With your left hand, you can adjust zoom and focus, move the stage, change the light source, and perform other shortcut operations. Without a controller, you move the stage and adjust the focus by flicking on the screen. There was quite a lot of mouse operation, which was inconvenient. With a dedicated controller, these operations are quite easy and intuitive.
The stand has also changed. A mechanism to absorb vibration has been added to the feet, allowing observation with reduced vibration when using a high-magnification lens. This stand can also be used with conventional lenses. As for the main unit, two cameras can now be connected simultaneously. Of course, you can use them in combination with the camera and lenses you have been using so far.
Next we will look at inductors. Inductors come in various types, ranging from through-hole components to chip components. Let's look at the through-hole type first. This is a drum-shaped inductor. Magnet wire is wrapped around the leads and soldered, and the winding is wrapped around the core. Naturally, we will also look at cross sections. The cross section looks like this. There is an I-shaped core with a coil of wire wrapped around it. This shape makes it easy to wind an inductor coil. The structure allows magnetic flux to leak out from the edges of the core. Therefore, depending on the current flowing through it, this inductor itself can be a noise source. Where to use it and the shape of the core must be chosen carefully. I'll show you later, but some inductors have a structure that prevents magnetic flux from leaking outside.
Next up is this inductor. At first glance, it appears to be a resistor. Actually, it's an inductor. A cross-sectional view reveals a magnetic core inside, with a coil wound around it. The core has an I shape, similar to the drum-type inductor mentioned earlier. The core and windings are tightly sealed with a resin-based material. It seems tolerant of harsh environments. The appearance is quite similar to a resistor, so we want to be careful not to mistake it for one.
Next, this is also an inductor. It has a different shape from the previous one, with a lead wire passing through the interior of the core. The actual cross section shows that it is an inductor with one turn. Specifically, this type is called a ferrite bead inductor. Generally, these inductors are inserted where high-frequency noise is generated, such as in switching power supplies. They are often used to absorb noise.
Next up, these are both surface-mount inductors. Can you tell the difference? The difference between these two inductors is their rated current and inductance values. At a glance, the thickness of the windings is quite different. The one with thicker windings has a higher rated current, while the one with thinner windings has a lower rated current.
We will use the split-screen function here to look at each inductor. I'll map it out once at low magnification. If you click on this map at high magnification, you can move to where you want to go, which is convenient. First, look at the cross-sectional area of the winding with the higher rated current. This microscope can also measure area. If you outline the copper wire, you can see the area like this: 187,869 μm², so about 0.18 mm². Let's also look at the one with the lower rated current. Similarly, the copper wire's cross-sectional area measures in the mid-0.03 mm² range.
If you look closely, you can see a difference in thickness between the outside and the inside of the winding. Too much pressure may have been applied when grinding the inductors with a rotary grinding stone, stretching the wire itself. The cross-sectional area of the copper wire and, to be precise, the current characteristics of the core's permeability determine the rated current of the inductor. Furthermore, thin copper wire can be wound many times, whereas copper wire with a larger cross-sectional area can only be wound a small number of times. These differences also affect the inductance value.
Next up is another surface-mount inductor. Its structure is quite similar to the surface-mount inductors we saw earlier, but slightly different. The windings are encapsulated in resin. This is not just any resin; it is mixed with ferrite powder. Covering the windings with resin mixed with ferrite powder makes it easier for magnetic flux to pass through the magnetic path than through air. This means less magnetic noise leaks outside, reducing the effect of noise on surrounding electronic components.
If you look closely at this inductor, you can see a cavity. This is something we don't want there. I'm told it's a manufacturing variation, with air bubbles getting mixed in. We can actually measure how big this cavity is. Using the microscope's depth-compositing function and image processing, you can measure the depth of the indentation. How deep is the hollow area? It is approximately 500 to 600 μm. You can also display a color map to see the surface irregularities at a glance. This is made possible by combining the accuracy of the lenses and other optical systems, the accuracy of mechanical systems such as the stage, and image processing.
Next, we have this inductor. This is a metal-alloy-type chip inductor. A look inside reveals that the windings are encapsulated in magnetic material. Since the magnetic material itself acts as a magnetic shield, less magnetic flux leaks outward. By the way, I didn't get one this time, but chip inductors are also available in multilayer types. Those are inductors made by layering.
Here is the last of the inductors. I wonder if variable capacitors are often used. I think you don't get a chance to use variable inductors very often. If you do use one, it might be in a wireless system. That's about it. Capacitors are more likely to be used. The internal structure looks like this. There are ferrite cores inside and outside the case, with a winding between them. The inner ferrite core is threaded. Turning it moves the core and changes the coupling between the outer and inner cores to adjust the inductance. That's pretty well thought out.
Next we will look at resistors. First of all, here is a carbon film resistor, which is indispensable when doing electronics work. It's a so-called carbon resistor. What you're looking at now is a 1 kΩ resistor. These through-hole carbon film resistors are initially coated with a resistive element, which is then trimmed using lasers or grinding stones to create the resistance value. The white part in the middle is the ceramic core. On top of that is a carbon film, with metal caps on either side. The entire surface is coated with a beige-ish resin. The ceramic core and carbon film have evenly spaced indentations. These are traces of trimming. I asked resistor professionals, and this one was trimmed with a grinding stone.
Next, here is a resistor that looks like a normal carbon film resistor. A cross-sectional view reveals that it is completely different. This is a solid resistor. The black squishy part in the middle is the resistive element. Unlike carbon film resistors, solid resistors are made from a mixture of carbon and resin. Their characteristics include high pulse resistance and low inductance because they do not have a coil shape. Even when disassembling appliances, I think it's getting harder and harder to see solid resistors like these.
Next up are variable resistors. This one is a multi-turn variable resistor. The actual cross section shows that it is made up of a combination of gears that adjust the resistance and screw parts.
Turning this screw also turns the gears that are meshed together. They turn together. If you remove the cogwheel, you will find a resistive element beneath it. If you look at the back of the gear, it has a kind of metal leaf spring. It is in contact with a resistive element.
I imagined that a multi-turn variable resistor would have a longer resistive element that they adjusted, but it wasn't what I imagined. By creating a reduction ratio with screws and gears, the amount of movement is reduced. That means the resistance can be finely tuned.
Next up is a cement resistor. Because of its heat resistance, it can be used at high power. It is often used as a resistor for electrical discharge. The outside is made of ceramic, and the interior is cemented. The ceramic outside is very hard. The cement part falls to pieces when poked with tweezers. It is quite soft.
What about the internal resistive element? High-resistance metal wire is wound around a ceramic core. It is made of metal wire like this. Unless you make it consume a very large amount of power, it never burns out.
The last resistor is this chip resistor. Chip resistors basically have a resistive film on a ceramic substrate. In this resistor, the resistive film is further coated with resin to protect it. In this case, it was coated with green resin. Here it is with the coating removed. The black part is the resistive element.
The resistance adjustment is in this section. There is an L-shaped cut, and the adjustment is made here. Originally, there was also a resistive element in this area. The resistive element is burned off with a laser to adjust the resistance value. The vertical trimming adjusts the resistance roughly, and then trimming to the side fine-tunes the resistance value.
Next we will look at fuses. A typical glass tube fuse looks like this. There is a wire inside. When an overcurrent flows, the wire breaks, interrupting the current. The thickness of the internal wire changes with the current at which it interrupts. The thicker the wire, the more current it can carry.
There are various shapes of fuses. First, this one looks like a resistor, but it's actually a fuse. If you cut it open and look inside, you'll see a wire like this. Again, the fuse principle is at work here. If an overcurrent is applied, the wire inside will break. However, this type of fuse is soldered to the board. When it breaks, you remove the solder and solder in another one. It can be quite daunting. It's better to use something with a holder, like a glass tube fuse. It is easier to maintain.
Next, let's look at these high-power fuses. Disassembled, it has sand inside. This is sand to extinguish arcs. Fuses for high voltage and high current, especially, often have sand inside.
The area where current flows also has a curious shape. It is basically a strip fuse. There is a bumpy hole in the middle. I don't know exactly why it is like this. The current density is higher in the area with the hole. I'm thinking it has this shape because they want to make sure it breaks at this part.
Next, let's take a look inside a crystal. This is called a quartz crystal. It is a component used to create vibrations with a fixed period. It is used in quartz watches and to make clock signals for microcontrollers.
First, here is a 32.768 kHz quartz crystal for watches. The interior is extremely delicate. I've broken the internals several times when taking them out of the can. This is one of the survivors. The inside looks like this. There is a crystal shaped like a tuning fork. There appears to be a thin film of metal on its surface.
When the crystal is subjected to a voltage, it vibrates at an invisible level. We can utilize it as a clock source. The frequency at which it vibrates depends on its shape. Crystals are made for various frequencies.
The one I just looked at is from Citizen. I'll take a look at the Seiko crystal as well. The frequency and load capacitance are the same, but what do you think? The one on the left is Citizen's, and the one on the right is Seiko's. The shape of the crystal is very different. Even for crystals with similar behavior, the shape varies from manufacturer to manufacturer. I don't know much about quartz crystals. It's not as if one is better than the other.
Next is a 17 MHz crystal with an even higher frequency. It is no longer shaped like a tuning fork. The structure has a thin metal film on the front and back of a simple plate. This plate is made of quartz. I'm sure there is a lot of know-how involved in the way the crystal is cut and the shape of the thin film.
Next up are the DIP switches. I was simply interested in what's inside. We will look at the cross section. The structure was pretty much as I had imagined. When the white plastic part is moved with a click, it presses against the metal and closes the contact.
In the area where the moving plastic touches the metal, it looks like there is some kind of surface treatment on the metal. The colors are very different. This is the part that rubs against the surface. A smooth surface may reduce wear on the plastic and contacts.
Next is this toggle switch. When you open the package, you see a structure like this. When the plastic in the middle moves, the metal pieces in contact with it move, changing which contacts conduct. The plastic part was coated with some kind of silicone grease to make it slide easily and reduce wear. Transparent grease is also applied to the metal contact points, presumably to increase contact durability.
This is what the contacts look like after the grease is removed. I wondered if the area that serves as the contact point would be smooth. Surprisingly, it is quite bumpy.
Next, we will look at the semiconductors you have all been waiting for. This time, the emphasis was on the ease of viewing the contents, so we will observe semiconductors in metal cans. Recently, semiconductor packages have often been made with epoxy resin. Semiconductors in cans like this are still being manufactured. These cans are very hermetically sealed and release heat easily. In addition, the can itself has the advantage of being a shield.
Inside the can, there is a silicon chip in the center like this. There are leads extending outside the can. These leads also connect to the inside of the can. The chip is connected to the leads by wire bonding.
This one is a diode. This is the anode, and this is the cathode. On the chip, it is possible to bring out one more terminal for the cathode. In fact, it wasn't connected anywhere. The same product is available in a different package, where two cathode terminals are brought out. Maybe that's what the pad is for.
If you look closely at this pad, there is a mark that looks like damage. When the semiconductor is completed, it is tested with a semiconductor tester. The tester probe is applied to the pad like this. It's a mark from that process.
The green line on the silicon chip goes all the way around. What is this? Is it called a guard ring? If anyone knows, please comment!
The next one is also a diode. This is an RF diode with two circuits. Maybe it's because it's the same manufacturer as before, but it doesn't look much different. The wire bonding method is wedge bonding on the chip side. The lead side was also wedge bonded. The wire is whitish, so it may be aluminum wire. The maximum current through this wire is about 20 mA, so perhaps using gold wire would be expensive or something like that.
The two RF diodes inside are the same, but a closer look reveals that their wiring lengths are a little different. For the diode on this side, the electrode on the anode side is slightly longer. I wonder if RF characteristics come into play in these areas as well. It's a small difference. People who care might possibly care. But I think the length of the leads and the length of the traces on the printed circuit board connected to the circuit would have a more dominant effect.
Next up is the N-channel MOSFET. These pins are the gate, drain, and source, respectively. There is a fourth lead from the package. This is connected to the case. Compared to the previous diode, it has a slightly more complicated shape.
The chip is quite colorful: green, purple, blue, beige, and so on. But the chip doesn't actually have these colors. They are caused by the interference of light from the film deposited on the chip. Thin objects, such as soap bubbles, sometimes appear rainbow-colored. It's the same thing.
Next up is an NPN bipolar junction transistor. It's very big. Compared to what we have seen so far, this one has a large power rating: 60 V, 15 A. Therefore, the silicon chip and case are also larger. If you look at the back of an old audio amplifier, you sometimes see a transistor of this shape fixed to a heat sink.
Inside, it looks like this. The chip is large because of the large power rating. The case is also larger, and the bonding wire is quite thick. It's like a power semiconductor. The chip layout is a little more interesting. Strips of electrodes alternately overlap each other. In this case, the base electrode surrounds the outside, and the striped electrode inside it is the emitter.
For this silicon chip, the current flows through the chip, so the case itself is the collector. It's a vertical chip, like a power semiconductor. The structure allows current to pass through the chip. The silicon chip and the case are soldered together. It is also designed to allow heat generated by the silicon chip to escape efficiently into the case.
The last semiconductor in a can is an operational amplifier. This is an LM201 op amp from Texas Instruments. Compared to other packages, the can version is extremely expensive, at about 3,000 yen apiece.
Looking at the silicon chip, you can see an op-amp-like circuit. This part is the input of the differential amplifier circuit. There are also what look like resistors. I wonder if this is the transistor in the output stage. By looking at the actual datasheet and this video, I think you can understand how the layout corresponds to the schematic. If you're interested, pause and check things out.
Next, we'll look at MEMS microphones as an unusual semiconductor. This is a larger MEMS microphone. Phones use several MEMS microphones in packages even smaller than this. You may wonder if something shaped like this is a microphone. If you look at the back of the package, you'll see a hole. Air vibrations are transmitted through this area and vibrate the membrane inside.
When you open it up, you see a structure like this. This round part is the diaphragm of the microphone. This part is made of semiconductor silicon. The silicon film looks like this. It vibrates with sound, which changes the capacitance. The difference in capacitance is read and converted into a signal.
A dedicated ASIC is used for the sensing and analog-to-digital conversion circuits. The part potted in this black epoxy-resin-like material is the ASIC. For this MEMS microphone, the idea is to make a membrane with silicon and use it as a microphone. I think it's pretty amazing technology.
Here's the last of the semiconductors: a homemade semiconductor. You can actually make silicon chips with your own layouts. This is an op-amp circuit I previously made with support from the open-source semiconductor community ISHI Kai. Separate areas within a single chip allow you to share one chip with others. Just this part is the op-amp circuit that Ichiken made.
The process rule for the transistors is 0.6 µm. The chip is 1.8 mm square. The price was lower than expected: 20 bare chips for a little over 200,000 yen. For more information on this chip, we have summarized it in a video linked in the description. If you're interested, please check it out.
When magnified with a high-magnification lens, it looks like this at the transistor level. As you zoom in, it gets harder and harder to see. I guess this is the limitation of optics. The usable limit is about 5,000×. Image analysis itself can still be performed at this level of magnification. You can see the unevenness of the surface. Where there are metal layers and so on, you can see a slight rise.
I'd also like you to look at the via section between the semiconductors. At the time of design, the via opening was square. The shape is a little rounder and less angular, isn't it? It goes through processes involving chemicals and gases that dissolve material, and various other things. Inevitably, the edges are removed and rounded off like this. When you actually make it, you can see how this happens.
And here is the last part: a printed circuit board. We will observe the cross section. Printed circuit boards are often viewed from above. There are actually not many opportunities to see a cross-sectional view like this. When you look at it, it looks like this. This time we are using a six-layer board. As you can see, there are six layers.
There is solder resist on the outermost surface. Below that is the surface layer of copper foil, and further down are the inner layers of copper foil. The substrate is FR-4, with glass fiber and epoxy resin. There is copper running vertically in this section. This is a via.
This time I used pad-on-via. This allows copper foil to be placed over the via. Being able to use pad-on-via for high-density mounting can be very helpful in many situations. Because pad-on-via places copper foil over the via, the via needs to be filled with something. What you see on the screen now is a via filled with resin. This keeps the copper foil on the surface from being pitted or torn.
The PCB characteristics vary depending on the material used to fill this area. This one is filled with resin mixed with copper powder. Heat is transferred more easily than with resin alone. In addition, current flows more easily. But the price is higher. Copper filler is good for carrying current or transferring heat all the way to the other side of the board.
Today, we used a microscope to observe various things, including electronic components. It can be used to check what's inside after disassembly. In manufacturing, it also allows you to immediately observe fine details and check whether product quality is maintained. I think there are many opportunities for microscopes to be useful to quality assurance personnel, developers, and researchers.
If anyone is interested in a HIROX microscope, you can find the link in the description. Please check it out! If you found this video helpful, please give it a like and subscribe to our channel. If you have any questions or concerns, please leave a comment as well. Thank you for watching the video until the end.
Article published · Updated