My laptop charges with USB C, so when the standard charger broke, I just used the USB out port from an EcoFlow battery. The display on the battery said the laptop pulled 25-30 watts while charging. So, why can’t I use just any USB brick that can output more than 30 watts?

Is there something that is bound to go wrong that I don’t know about?

Are laptop chargers really that special?

(Edited for clarity)

  • @j4k3
    link
    English
    32 months ago

    Are laptop chargers special?

    Yes and no. If you are going to use it regularly, you’re better off with the OEM brick, especially if the OEM came with a charger with over 100 watts.

    There are several factors involved here. It really depends on the battery circuit topology and how/if it bypasses the battery to avoid fatigue, and how it deals with peak current requirements.

    I would not use USB-C regularly for a few reasons.

    First, the strain relief for the connector, the cord just behind the connector, is not designed for regular handling. You know how people that use their phone while it is plugged in always seem to damage the cable near the end of the connector? Yeah, that is the problem. That will eventually short out stuff between the power and other lines and there is a lot that can be damaged.

    Second, you may like the smaller form factor of some USB-C chargers, but Maxwell’s equations are not impacted by aesthetics or convenience. The magnetics inside the OEM power brick are larger. Inside a typical modern power brick there are larger metal strips against the inside of the case to radiate heat more continuously. There is also theoretically more space to separate the heat from the output smoothing electrolytic capacitors. These capacitors are the primary failure mode in most power supplies.

    Third, the way higher power devices are made into smaller form factors is with higher switching frequencies. Basically it means the magnetics can be smaller and a different material. These kinds of supplies are relatively new to the dumpster fire that is consumer electronics where penny pincher accounts reign supreme over their diminutive electrical engineer slaves. Most OEM power bricks are based on the TL494 chip. You can take apart a power supply from 30 years ago and it will have this chip too. It is a workhorse and super reliable. Newer stuff like higher frequency designs are constantly changing in a super volatile market. These chips are discontinued constantly, so a power supply based on one is likely something that was quickly thrown together by a subcontractor that barely knew the device and checked the boxes to get paid. The TL494 is the Chad of chips in this niche. While the switching frequency should not matter in an ideal world, in practice, it does. When aesthetics or convenience trump engineering, things get stupid fast. The tiny chargers are unlikely to have sufficient heat dissipation for continuous charging. You probably won’t find a decent company that gives you a duty cycle like specification indicating what the peak power versus continuous power should be, but these are not the same number. One of the biggest problems will be smaller output capacitors that get hotter. The higher frequency on these will cause them to fail sooner in most instances, even when the lower equivalent series resistance versions are used. The heating will make this happen even sooner.

    Four, USB-C is actually terrible from the perspective of the back side of the connector. There are too many connections that are too close together in order to double all connections. Everything in engineering is a compromise and being ignorant if these compromises is foolish, so hear me out. The spacing between the pins on USB-C is so small that it requires advanced nodes from PCB houses. You can’t order the cheap node from PCB Way and use a USB-C connector because that pin pitch is so tiny it is beyond their resolution. If I etch my own circuit boards, I have to use photolithography with a photoresist and UV/transparency to etch the required resolution. This takes me twice as long as just using toner transfer. I can technically do it with toner transfer, but my failure rate is higher than 50% at this resolution, and I need to etch to know if it fails.

    When a lot of power is placed across a tiny little wire pin like this, that point gets quite hot. This connector is near the exterior of the laptop enclosure. Debris and moisture from the plug and outside world will inevitably build up around this area, and the heat will tend to attract junk. Over time, this tiny pin pitch spacing will develop resistance from the build up of junk and start to short itself out. This may take years longer than the life of the device or it might not. The connector lacks robustness, especially when it is compared to the connector designs typical of most laptops. Every connector has a rated life span under specified conditions such as plugging cycles and handling while connected. USB-C is much lower in these specifications compared to a typical laptop charger connector.

    Five, the circuitry for USB-C power delivery is digital logic and a point of additional failure. This becomes more likely with heat, and failing output capacitors. As resistance builds on the laptop connector side, it will also cause issues with this PD circuit.

    I could go on further, but those are my top 5 in no particular order.

    • dave@hal9000
      link
      32 months ago

      Thanks for that thorough explanation! As someone that knows enough to be suspicious of the usual problematic factors, but not really aware of the details, this was great info

    • @UsernameblankfaceOP
      link
      32 months ago

      The OEM brick is rated for 65 watts, does that change things?

      • @j4k3
        link
        English
        22 months ago

        Yeah, you should be better off with more random stuff. I have taken apart several dozen power supplies like these for various projects in the past. The ones with power factor correction start as low as 80 watts, but don’t become common until around 120 watts. In my experience, the supplies with power factor correction tend to also have a smoother output with far less noise overall. The lower power stuff can be like the wild west. Connect up some Arduino project and power supply noise can make you chase your tail.

        The same basic thing applies here. The laptop was likely designed for a specific type of filtering on the supply side. There are always margins, and battery managing chips have gotten a good bit more sophisticated, but in the end doing outlier things to a circuit that was designed to a price is always an iffy proposition. The PD circuit is a known input randomness factor for the laptop engineers, but ultimately the engineered reliability factor of the OEM supply is going to have slightly more potential quality than some contract developed and contract manufactured venture capital driven USB-C PD power supply where a fraction of a penny is an absolute warzone.