How ADAMON 65W Rivals GaN Chargers Without GaN: A Teardown of Its Input-Circuit Trick
イチケン / ICHIKEN
Most compact USB chargers today rely on gallium nitride (GaN) power semiconductors. The ADAMON 65W from E-Power Solutions, a charger developed by a Japanese company and billed as one of the world's smallest, uses conventional silicon (Si) power semiconductors instead. Yet it is roughly the same size as GaN-based competitors. In this teardown review, Ichiken measures its performance, opens it up, reverse-engineers the circuit, and asks how a silicon design reaches GaN-class size and efficiency. Ichiken's answer centers less on exotic components and more on circuit design, particularly an unusual way of controlling the input capacitors.
Form Factor and Weight
The ADAMON 65W is a single-port charger with a folding plug. Its shape is a tall rectangle rather than a cube. Ichiken notes that a vertical rectangular shape is better for heat dissipation. The usual concern with tall chargers is that their weight pulls them out of the outlet, but that is not an issue here because the charger is very light.
It weighs 64.7 g. For comparison, Ichiken measured a CIO 65W single-port charger at 70.6 g and an Anker model at 119 g. That places ADAMON among the lightest chargers in its class. The reason for the low weight becomes clear during disassembly.
Coil Whine, Especially at 200 V
The one drawback Ichiken flags is coil whine. It is most noticeable at low loads, up to about 1 A, and it is audible even with no load. At 100 V input, Ichiken considered the noise acceptable, though users in quiet environments might notice it. At 200 V input the whine becomes much louder, which Ichiken calls a genuine concern. Since Ichiken only uses 100 V in Japan, this is not a problem for them, but they advise caution for overseas use. They also stress that the noise does not mean the charger is broken.
Specs: Size, Power Density, and Heat Headroom
Ichiken compares the ADAMON against the CIO NovaPort SOLO II, which is also rated at 65 W but uses GaN. The ADAMON's volume is 46.6 cm³, which Ichiken describes as among the smallest in the industry. According to the crowdfunding page, the product was released around June 2023, which means it has been in the smallest class for nearly two years. Its surface area is also small, which matters for cooling.
For power per volume, Ichiken says about 1 W/cm³ is typical for 65W-class USB chargers. The ADAMON reaches 1.39 W/cm³. Ichiken found that figure surprising for a silicon design and suspected something unusual inside.
The second metric is power loss per surface area, based on measured power loss. A lower value means more headroom for heat dissipation, since chargers mostly cool by natural convection and larger surfaces shed heat more easily. The ADAMON has more headroom than the CIO. Ichiken attributes this mainly to lower power loss rather than to surface area. The ADAMON's maximum power loss is about 1 W lower than the CIO's, and Ichiken says 1 W makes a significant difference in heat dissipation at this size. Ichiken treats low loss as the area where the designer's skill shows, weighed against what the product can cost to build.
Conversion Efficiency
Efficiency describes how well the charger converts AC wall power into DC for USB charging. Higher efficiency means less loss in the unit and less heat. Ichiken found the ADAMON efficient overall, with a peak of 93.2% at 20 V output. According to Ichiken, 65W GaN chargers typically reach about 91–92%, so a silicon design reaching this figure is notable.
Ichiken also measured efficiency at 200 V input. Most USB chargers are more efficient at 200 V, but the ADAMON was less efficient. Ichiken speculates that the internal design may be optimized for 100 V.
Temperature and Thermal Throttling
A thermal camera showed a maximum surface temperature of 71 °C, with one side face the hottest. The whole body warmed evenly, which Ichiken reads as a sign that internal heat is spreading well to the surface.
Thermal throttling is a protective function that limits output power to prevent heat damage. To check for it, Ichiken attached a thermocouple to the hottest spot and ran the charger at 65 W from the start. The temperature rose, then leveled off after about 30 minutes, peaking at 68 °C. Output never dropped. Ichiken ran the test for about three hours, saw no sign of throttling or shutdown, and ended the experiment. They attribute the absence of throttling to the low power loss. Having tested many GaN-based chargers from other manufacturers, Ichiken calls this result from silicon alone "pretty amazing."
Inside: Three Boards, Graphite, and No Potting
The charger contains three stacked circuit boards, densely populated with surface-mount components. Thermal conductive sheets are applied to both sides. A plastic insulating sheet wraps around the entire assembly like a shield, and a graphite sheet wraps around the outside of that.
Ichiken points out that chargers in this class are often filled with thermally conductive potting compound. The ADAMON uses none. Instead, heat travels from the source through the thermal sheet, the plastic sheet, and the graphite sheet to the outer case. Ichiken compares this design philosophy to Apple's USB chargers. They connect the teardown back to the earlier measurements:
- The even temperature across the body comes from the graphite sheet wrapping all the way around.
- The hottest face in the thermal image matches the side with the thermal conductive sheets over the heat-generating components.
- The low weight comes from the absence of potting, which Ichiken says typically accounts for about 20–30% of a charger's mass.
- The coil whine may also stem from the lack of potting, though Ichiken presents this only as a possibility.
The board is stamped "UNIFIVE." Ichiken found that E-Power Solutions is a fabless company, and UNIFIVE appears to handle board and thermal design optimization as well as mass production.
The Circuit: Undersized Capacitors and Blocking MOSFETs
Ichiken reverse-engineered the full circuit. Its main path is a full-wave rectifier that converts AC to DC, smoothing electrolytic capacitors, an active-clamp flyback converter with a transformer, and a secondary circuit that produces the USB output voltage, such as 20 V.
The unusual part is around the smoothing capacitors. One MOSFET is placed so it can block current flowing through the whole circuit. Another MOSFET is in series with one of the electrolytic capacitors. The electrolytic capacitors are rated for only 200 V or 250 V. The charger accepts 100–240 V AC, and at 240 V a conventional design would charge the smoothing capacitors to about 340 V, which is well beyond those ratings.
Ichiken's hypothesis was that the MOSFET switches on and off according to the instantaneous AC voltage. That would stop current from entering the circuit when the voltage is high, so the capacitors never charge past their rated voltage.
Confirming It with Waveforms
To test the hypothesis, Ichiken measured the AC input voltage, the input current, the voltage across the capacitor with the series MOSFET (Vc1), and the voltage across the electrolytic capacitor just before the DC/DC converter.
At 100 V input, the capacitor voltage looked conventional. It charged to the AC peak, slowly discharged, and recharged at the next peak.
At 240 V input, the behavior changed. Vc1 stayed nearly constant at just over 100 V, because once that capacitor reaches a certain level, its MOSFET turns off and blocks further current. The capacitor before the DC/DC converter charged only during a limited window of each cycle and discharged the rest of the time. Its peak voltage was 233 V, which is within its 250 V rating. The main MOSFET is on during the charging window and off otherwise, and the DC/DC converter runs from the stored charge in between.
Ichiken also observed that the capacitors with the series MOSFET are used at 100 V input but not at 240 V. At the higher input voltage, two electrolytic capacitors alone can supply enough energy to the DC/DC converter.
Why This Makes the Charger Smaller and More Efficient
Ichiken lays out several effects of this design.
Smaller capacitors. Electrolytic capacitors grow larger as their voltage rating increases. In a typical charger, the smoothing capacitors see about 140 V DC at 100 V input and about 340 V DC at 240 V input, so 450 V parts are common. Because the ADAMON controls the capacitor voltage, it can use 250 V parts, which greatly reduces capacitor volume and directly contributes to the small size.
A narrower input range for the DC/DC converter. A typical charger's converter must handle roughly 100–340 V. The ADAMON's converter sees roughly 100–240 V. Ichiken thinks this may make the converter somewhat easier to optimize compared with other chargers.
Lower-voltage switching devices. Ichiken suggests that typical chargers likely use 650 V super-junction MOSFETs or GaN devices here, while the ADAMON's lower internal voltage could allow MOSFETs rated around 300–400 V. Higher-voltage MOSFETs inevitably have higher on-resistance. Ichiken acknowledges that switching losses and other factors complicate the picture, but says at least on-resistance could be lower.
In Ichiken's reading, reducing the voltage the components must withstand makes the components smaller, reduces heat, and together these make the charger smaller.
Conclusion: Design Over Components
Ichiken concludes that the ADAMON 65W achieves its size through circuit-level innovation that shrinks the components themselves, which in turn shrinks the charger. Many recent chargers gain efficiency by adopting cutting-edge components like GaN. Ichiken sees the ADAMON as taking a different approach, improving efficiency through the design of the circuit itself. The remaining caveats from the review are the coil whine, especially at 200 V, and the lower efficiency at 200 V input compared with 100 V.
This video is brought to you courtesy of Yokogawa Test & Measurement.
Hello. Today, I will disassemble and review the world's smallest USB charger, developed by a Japanese company. That's the USB charger here: ADAMON 65W from E-Power Solutions.
This USB charger does not use gallium nitride power semiconductors, which are almost standard in modern chargers. Surprisingly, it achieves this small size using conventional Si power semiconductors. In fact, compared to other chargers using gallium nitride power semiconductors, it is almost the same size or smaller.
Today we will check what kind of circuit and structural design features this charger uses while disassembling it. The link to purchase the USB charger I disassembled today is in the description. If you are interested, please check it out!
But first, here is today's sponsor. Go ahead.
I assume everyone has a measuring device. Are they well calibrated and adjusted? Today I would like to introduce Yokogawa Test & Measurement's standard generators. Yokogawa is known for oscilloscopes and power meters, so you may associate it with measuring instruments. They also make standard generators and calibrators for calibrating measuring instruments. Yokogawa Test & Measurement is capable of both measurement and generation.
We have two power meters here. If we connect the standard generator, this power meter's display is almost identical to that of the generator. The other shows a different value. Actually, the power meter below is significantly outside the product specifications. This doesn't work for my job.
By calibrating the measuring instruments you use regularly with a standard generator, you need to make sure they display reliable values. What do you do with instruments that are out of specification? After this, I'll send it out for adjustment so that the displayed value falls within the specification range.
In addition to high-precision measuring instruments, Yokogawa Test & Measurement offers standard voltage and current generators and AC power calibrators at reasonable prices. If you need in-house calibration, please consider Yokogawa Test & Measurement!
Let's get started. As for the appearance, it is a typical single-port USB charger with a folding plug. The shape is vertical. A vertical rectangular shape like this is advantageous in terms of heat dissipation compared to a cubic shape like the one on the right.
A vertical charger can easily fall out under its own weight, but that is not the case here. The charger itself is quite light and will not come off. I noticed when I touched it that this charger is quite light: 64.7 g. Compared to other 65W single-port chargers, the CIO weighed 70.6 g and the Anker 119 g. This is one of the lightest USB chargers in its class.
One point of concern was the audible coil whine. It seems to make noise at low loads, especially with load currents up to about 1 A. Here's what it actually sounds like.
This is the sound at no load.
I wouldn't mind it that much. It may bother some people, especially when used in a quiet environment. The input voltage for what we were just listening to is 100 V. At 200 V, the coil whine gets even worse.
This is indeed a level of concern! In Japan, I use only 100 V, so it is not a problem. Be careful when using it overseas! Note that even if this sound comes from the charger, it is not broken.
From here, we will look at the performance we confirmed through spec sheets and experiments. First, here is a specs comparison chart. We will compare it with the CIO NovaPort SOLO II, one of the smallest in the 65W class. Although the rated power is the same 65W, ADAMON 65W uses Si power semiconductors, while CIO uses gallium nitride power semiconductors.
ADAMON has a volume of 46.6 cm³, which is fairly small. It is indeed one of the smallest in the industry. If you look at the crowdfunding site, it looks like it was released around June 2023. It appears to have been among the smallest in the industry for nearly two years. Its surface area is also small. This will affect heat dissipation.
Next we will look at performance indicators. First, power per unit volume. The larger this number, the higher the performance as a power supply. In 65W-class USB chargers, 1 W/cm³ is a typical value. ADAMON is 1.39 W/cm³, making it a high-power USB charger for its size. Since it is made so small using Si power semiconductors, I wonder if there's something special going on.
Next is the power loss per unit surface area. The power loss was measured beforehand. The smaller this number, the more room there is for heat dissipation. Most of the charger's heat dissipation occurs through natural convection. The larger the surface area, the easier it is to cool.
Compared to the CIO in the same class, ADAMON seems to have more room for heat dissipation. This is more due to low power loss than surface area. I'll come back to that later, but ADAMON's maximum power loss is about 1 W less. At this size, a 1 W difference in power loss has quite an effect on heat dissipation.
There is also protection against overheating through thermal throttling. I'll check that later. ADAMON did not thermally throttle and could deliver a stable 65W output. I think thermal throttling did not occur because the power loss generated by this USB charger was low. Reducing power loss is where the budget available for the charger and the designer's technical skills come into play.
Next, we measured the power conversion efficiency. This shows how efficiently the USB charger can convert AC power from an electrical outlet into DC power used for USB charging. The more efficient the USB charger, the lower the power loss generated in the unit, resulting in less heat.
Here are the efficiency measurement results. The horizontal axis is output power, and the vertical axis is power conversion efficiency. Overall efficiency is much higher. Maximum conversion efficiency is 93.2% at 20 V output. For a USB charger in this class, it is in a fairly high-efficiency category.
What's more, this result is achieved with Si power semiconductors, without using gallium nitride power semiconductors. Given that the power conversion efficiency of 65W chargers using gallium nitride power semiconductors is around 91–92%, this is a pretty good figure.
We also measured conversion efficiency with an outlet voltage of 200 V. Although a typical USB charger has higher efficiency with a 200 V input, ADAMON's efficiency seems to be lower at 200 V. It may be designed internally for higher efficiency at 100 V.
The next step is to measure the temperature of the unit. Viewed through thermography, the maximum temperature was 71°C. This side seems to have the highest temperature. The entire charger is evenly warmed. Heat generated inside is well diffused to the surface.
Next, in relation to temperature, I'll see if thermal throttling occurs. Thermal throttling is a function that intentionally limits output power to prevent the charger from breaking down due to overheating. The temperature was measured by attaching a thermocouple.
Here are the results of the experiment. The horizontal axis is the time elapsed since the start of the experiment. The left vertical axis is the surface temperature of the charger, measured at the hottest part. The vertical axis on the right is the charger's output power.
Power output is 65W right from the start of the experiment, and temperatures rose accordingly. About 30 minutes into the experiment, the temperature leveled off, reaching a maximum of 68°C. Operation was stable, without any thermal throttling or other problems. I ran it continuously for about three hours. There was no particular sign of it stopping, so I ended the experiment.
I've experimented with a lot of things so far. While other manufacturers use gallium nitride power semiconductors, this one has achieved so much with Si power semiconductors alone, without gallium nitride. I think it's pretty amazing.
We will disassemble it and verify what's actually inside. I quickly disassembled it. It is structured like this. If you look at it, there are three boards: one, two, three. Thermally conductive sheets are also applied on this side, that side, and the surface where many surface-mount components are installed.
A closer look at the exterior reveals a plastic shield-like piece. This component covered the whole area like this. A sheet of plastic wraps all the way around. If you take a closer look, a graphite sheet is also wrapped around the outside. It's like this.
USB chargers in this class are often filled with thermally conductive potting material. ADAMON is not structured that way. Heat is transferred to the outside using graphite sheets like this. The design philosophy is similar to Apple's USB charger.
This is how heat is transferred. There is a heat source, a thermally conductive sheet, a sheet of plastic, and then a graphite sheet. Finally, heat escapes toward the exterior. The unit's temperature was uniform throughout thanks to this graphite sheet being wrapped all the way around. There were thermally conductive sheets on the surfaces that generate a lot of heat. This surface was the hottest when viewed through thermography.
The unit is also lightweight because it uses no thermally conductive potting material. When potting is used, the potting material accounts for about 20–30% of the mass. Not using it helps reduce weight. The coil whine may also be due to the absence of potting material.
A closer look at the exterior of the disassembled board shows that it was stamped UNIFIVE. Upon further investigation, I found that ADAMON's developer, E-Power Solutions, is a fabless company. UNIFIVE appears to handle optimization of the board and heat dissipation designs, as well as mass production of the product.
I disassembled everything and tried to reverse-engineer the circuitry. To summarize the main parts of the circuit, it looks like this. A full-wave rectifier circuit converts the AC voltage coming in from an outlet into DC voltage. Then there are electrolytic capacitors for smoothing. Here is an active-clamp flyback converter and a transformer, followed by the secondary circuit. The voltage is adjusted to the voltage used for USB charging, such as 20 V.
A closer look at this circuit reveals an interesting circuit around these smoothing capacitors. A MOSFET is included here so that the current flowing through the entire circuit can be blocked. In addition, this electrolytic capacitor has a MOSFET in series.
Looking at the electrolytic capacitors used in the disassembled circuit, their voltage ratings are actually only 200 V or 250 V. The AC input voltage of this charger is 100 V to 240 V. With ordinary operation, the voltage across the electrolytic capacitor would rise to 340 V. Obviously, the voltage rating of this electrolytic capacitor is not enough.
Perhaps the MOSFET here turns on and off according to the instantaneous value of the AC voltage to prevent current from entering the circuit. In other words, it may operate so that the electrolytic capacitors are not charged above their voltage ratings. I'll do an actual experiment and check the waveforms.
I experimented with a setup like this. First, let's capture the operating waveform with an input voltage of 100 V. Apply a load.
The yellow line here is the charger's AC input voltage. The red line is the charger's input current. The blue line is the voltage across the electrolytic capacitor with a switch in series. The green line is a little hard to see because it overlaps the blue line. This is the voltage across the electrolytic capacitor just before the DC/DC converter.
If we look at the voltage across the electrolytic capacitor, it has a familiar waveform. It is charged to the peak AC input voltage and gradually discharges. When the next input voltage peak arrives, the electrolytic capacitor is charged to that voltage, and the cycle repeats.
Next, let's set the input voltage to 240 V. This blue line, Vᴄ₁, is the voltage across the electrolytic capacitor with a MOSFET in series. It stays constant at approximately just over 100 V. This is because, once the capacitor is charged to a certain level, the MOSFET turns off, preventing current from flowing.
This green waveform is the voltage across the electrolytic capacitor just before the DC/DC converter. It is charged only for a certain period of time and then discharges. It is charged only during this period. The peak voltage across the electrolytic capacitor is 233 V, so it can operate below its 250 V voltage rating.
This is because the MOSFET controlling the current through the entire circuit turns on and off to control the charging of the electrolytic capacitor. Specifically, the MOSFET turns on during this section to charge the electrolytic capacitor. During the other sections, the MOSFET turns off and does not charge the capacitor. The stored charge is released to operate the DC/DC converter.
These electrolytic capacitors were used when the input voltage was 100 V. They were not used at 240 V because, when the voltage rises sufficiently, just two electrolytic capacitors can supply enough energy to the DC/DC converter.
As we have just seen, the input voltage to the circuit is higher than the voltage rating of the electrolytic capacitors. Through this control scheme, the voltage across the electrolytic capacitors is kept below their voltage ratings.
From here, we will consider why this ADAMON can be made smaller and why it has low losses. Basically, the higher an electrolytic capacitor's voltage rating, the larger its volume. This is the first point contributing to miniaturization.
With a normal USB charger, when the outlet voltage is 100 V, 140 V DC is applied to the electrolytic capacitor. When the outlet voltage is 240 V, 340 V DC is applied to the electrolytic capacitor. Therefore, the voltage rating has to accommodate the higher voltage. What voltage rating is actually used? A 450 V rating is common.
On the other hand, what about ADAMON? Since the voltage across the electrolytic capacitor is controlled in this input circuit, its voltage rating can be lowered. Using a 250 V rating significantly reduces the volume of the electrolytic capacitors. Lowering their voltage rating in this way contributes to miniaturizing the charger.
In terms of ease of design, ADAMON also benefits from the smaller voltage range at this DC/DC converter. In a typical USB charger, the voltage applied here ranges roughly from 100 V to 340 V. In ADAMON, the voltage range is about 100 V to 240 V. I think this might make optimizing the DC/DC converter design a little easier than in other chargers.
Also, since the voltage applied to the circuit is low, MOSFETs with lower voltage ratings can be used here. A typical USB charger might use superjunction MOSFETs with a 650 V breakdown voltage or GaN power semiconductors. In ADAMON, the voltage is only about 240 V, so something with a voltage rating between 300 V and 400 V can be used.
If the MOSFET's voltage rating is increased, its on-resistance inevitably becomes higher. Of course, there are many things to consider, such as switching losses. At least for on-resistance.
So there is a possibility that it can be lowered. Therefore, by reducing the voltage applied to the components, ADAMON can make the components smaller and reduce heat generation. These changes also make it possible to miniaturize the USB charger itself.
Today, I have disassembled and verified E-Power Solutions' ADAMON 65W. By actually disassembling it, I found out that they are making the components smaller through circuit innovations, and thereby making the main unit smaller.
Recent USB chargers can improve efficiency with state-of-the-art components. ADAMON takes more of an approach of improving efficiency through the design and the circuit itself. That was my impression of the product.
Links to the product disassembled today and other recommended USB chargers are collected in the description. If you are interested, please check it out!
If you found this video useful, please give us a high rating and subscribe to our channel! If you have any concerns, please leave your comments! Thank you for watching the video until the end.
Article published · Updated