Graham Sutherland / PolynomialMay 14, 2024

interesting thing of note: if you're looking to buy a GaN USB-C charger above 75W, compare the price-per-watt to a few 65W ones.

chargers above 75W are required to have an active PFC stage, typically requiring four more transistors and a fancier controller, which adds cost, so you can often tell if they're penny-pinching on the design if the price-per-watt is close to that of a 65W one. you should expect to see at least a 20% price-per-watt jump if they're doing it right.

Show threadGD2May 14, 2024

@gsuberland What confuses me is that even with the more expensive and more powerful models, they never have a ground pin?

Does GaN not have leakage current or is there just no market demand for it anymore? Sorry if this makes no sense, I have zero knowledge about physics 😬

Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 they're not required on isolated AC-DC converters, and GaN topologies are typically galvanically isolated with opto feedback.
Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 the ground pin is typically used to ground the chassis of whatever you're plugging in, but if there's an isolated AC-DC converter in between and the output voltage is low then the grounding makes no sense - there's nothing in the device you're holding that can be live. same thing with most laptop chargers.
Show threadGD2May 14, 2024

@gsuberland I've read before that DC chargers like these don't require ground to be safe, but I thought the ground pin on (some) laptop chargers was put there to get rid of that "tingly feeling" when your laptop has a metal body?

I can only parrot dubious internet articles here, I've read that this feeling is caused by a tiny unavoidable leakage current in the power supply, which is harmless but can be countered by grounding the chassis?

Show threadGraham Sutherland / PolynomialMay 14, 2024

@gd2 people often claim it's due to the safety caps (used for EMI suppression) but it's almost always just capacitive leakage across the transformer.

grounding the chassis does avoid it, but for various reasons (electrical and regulatory) you can't just bridge DC GND to earth in an isolated AC-DC design and call it a day. typically you need a separate chassis ground line on the DC side with a bleeder resistor to earth, and you need to carefully spec it to meet isolation requirements.

Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 which, for USB-C, obviously doesn't work.
Show threadGD2May 14, 2024
@gsuberland Wait, but my laptop charger (the standard Lenovo 65W brick, Schuko plug) is USB-C and it does get rid of the effect, as does using my USB4 dock or just plugging in my Screen via USB-C-to-DisplayPort. Only charging the laptop via an ungrounded Apple or phone charger causes this effect; and it even goes away then upon connecting the Screen?
Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 oh you *can* do it, but it's way more expensive than not doing it, and if you get the leakage suppression circuitry wrong you'll fail compliance (which is *REALLY* expensive).
Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 sorry I just realised I was unclear above. you can't just tie DC GND to earth *directly*, but you can implement a bleeder circuit in the charger that allows that leakage current across the transformer to be dissipated. usually they'll pick transformers with very low leakage to help make the design easier too.
Show threadGD2May 14, 2024

@gsuberland Oh, OK, I read it as "you don't have a separate chassis ground line with USB-C", this makes more sense now 😅

So the laptop charger and docking station probably have this feature for brand value but 3rd-party vendors so far have no incentive to incur the extra costs?

But I'm surprised that the screen grounds my laptop then, I don't even think its PSU supplies an extra ground pin (barrel jack)... Could my laptop be grounded indirectly via the tower PC also connected to the screen?

Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 yup. and probably exactly that, yeah.
Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 with the display, the display cable's shield connection is usually hooked to chassis ground via a parallel RC circuit (HV cap + bleeder resistor) which would have a similar effect to how you do it inside the AC-DC supply.
Show threadGD2
@gsuberland I've had an ungrounded macbook visibly discharge (sparks) via the same screen once, is that due to the circuit in the PC being designed for different voltages or amperages? Or is that entirely unrelated?
May 14, 2024 at 4:23pmWeb
Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 only tangentially related. probably the safety cap discharge I mentioned earlier.
Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 what's essentially happening is you've got two devices with isolated DC supplies, so the electrical potential of each device can be literally anything. the voltage delta between them can easily be in the kV range. when you touch them together it'll arc like you see with static discharge (ESD). that's one of the reasons all of the capacitive shield coupling and EMI caps have to be rated for much higher voltage than even the mains input.
Show threadGraham Sutherland / PolynomialMay 14, 2024
@gd2 but that voltage delta is only "backed" by a small amount of charge, just like when you become statically charged by a carpet or whatever, so it's not really harmful as long as you account for it in your design. the current that flows during the discharge event is pretty small.
Show threadGD2May 14, 2024

@gsuberland Yeah, I discounted it as "something like static charge" when it first happened, but I am still glad that the grounded docking station prevents it now 😅

I also guess I'll be sticking with the laptop charger for now, the space savings of GaN are nice, but I don't like the feeling of the leakage current. Luckily my laptop also seems to do enough USB-PD to daisy-chain things from there.

Thanks for the patient explainers by the way, I learned a lot today!