All of that because they couldn't be bothered to put two resistors for the detect pin.
Remember, line must go up at any cost.
this post was submitted on 24 Mar 2025
173 points (97.3% liked)
AssholeDesign
8470 readers
1 users here now
This is a community for designs specifically crafted to make the experience worse for the user. This can be due to greed, apathy, laziness or just downright scumbaggery.
founded 2 years ago
MODERATORS
Bro have you SEEN the price of resistors lately?
…..cuz I haven’t and I assume it’s fractions of a penny per.
I worked on an embedded product, on prototype there were pullup resistors on the RX/TX lines we used to plug a USB FTDI to do some debug. At manufacture they removed them, it created a floating RX that sometimes injected character that broke uboot...
I love stories from hardware / embedded design. Reminds me that black magic is real, and it's a miracle any of this works.
What is completely incredible, is that there is some simple embedded devices, be it a thermostat, dishwasher, etc, that sold thousands/millions of units, and the firmware is done by one guy, a simple developer, and when it works, it is not touched again.
A fraction of a penny per? Have you seen superman? Office space? How do you think millionaires are made?
But wait. Doesn't this make them both dumb? I'd expect a modern USB-C charger to still support basic 5v low current lazy devices too. If there's a USB-A to C cable that works, it must also still be possible to send the basic 5v down a C-C cable.
I also think there's always going to be a balance between how much a device needs to make and/or how much it needs fast charging to make sense to add the charging circuit for PD/PPS. Even $1-2 on top of the cost can ruin margins in the current electronic market.
PD includes a 5v option, so you’d think that chargers would default to a dumb 5v/4.5w with no other input, but apparently not… It’s probably something to do with the overcharge protection, or to not (further) harm a device with a damaged charge controller.
Or it could be that my PD chargers are pre pandemic and PD 3.1 fixed this.
Well, I'd expect that if they allowed 5v through but with a low current limit (I think the default 5v standard states quite a low current allowance). They could catch anything drawing too much and shut the port off until it detects disconnection/other reset.
I mean, if they're thinking about protecting a downstream device, adding this logic would make more sense than just not supplying any power unless a negotiation is made.
In any case, since standard USB ports on a computer will output 5v without anything being negotiated, then it's really no less safe than any other USB port in that regard.
USB c has active negotiation for power and communication via a cc pin
The usb a to c cable has a pull up resistor that mimics this and says “give 5v”
Some usb c chargers have a fallback 5v mode for this scenario but not all do
Surface Pro X is the worst about this. It knows you're trying to charge it via the USB c ports. And tells you to use their stupid charger instead.
Agreed. Learned this the hard way with my portable 4g router. It's the pickiest gadget I own with cord and plug selection and it left me hanging when I really needed it and I don't carry an A to C cord often. I wish I could find a good alternative to the Netgear Nighthawk M1100
This bothers me more than literally anyone, however as an ee I do feel obligated to explain their crime against humanity.
The usb pd charge negotiation chip costs between $.50 and $3, cheaper for low current, like 12-18w, more expensive for 100w, also those tend to have more logic including battery management and even integrated buck converters.
Still unacceptable by any standards, but there are software aspects like debugging how the different charge and power modes work together, etc.
Largely it's inertia, buying a power brick and just putting a 19v is tried And tested, even though 20v is basically the exact same except for the pd negotiation.
The cheap ones assume you use the same cable that came with it, which are generally 2 wire only and have the resistor in the plug.
Way to complex and expensive. This just needs 2x 5.1 kOhm resistors between the CC pins and GND.
Yeah, the cheap cables they provide are often only vcc and gnd with the data pair shorted in the plug.
Trying to use a normal plug confuses the controller.