Hacking together a USB-C charger for a cheap Chromebook
blog.filippo.io
blog.filippo.io
Disclosure: I'm the author
There are so many USB charging standards - I had been thinking that it should be possible to construct a universal USB power converter. (Even in the case where the charger can't supply the watts the device wants, you could accumulate some energy in a small battery large cap and charge intermittently). But I suspect the reason it hasn't been done so far is liability - there are bound to be a few devices that don't play ball and blow up.
The reason why he was doing this became clear when one of the cables he tested was so bad that it fried the USB port on his Chromebook Pixel 2015.[0]
[0] https://www.androidpolice.com/2016/02/02/google-engineer-ben...
I think as long as these hacks stay in 'prototype' mode, and is clearly called out as such, then it is fine. Once it goes beyond that (ie for sale without proper testing, etc), then it gets worrisome.
The switch has a couple issues with how it requests power but they're minor and mostly on the dock rather than the actual console.
Nnnnnnot really.
He was checking for cables that were misrepresenting themselves and might cause the wrong amount of current to be drawn.
He found one cable that had the power wires completely backwards. That means it was a very blatant mistake that could just as easily happen to any kind of cable.
The wall to 12V brick I had was massive, and caused lots of noise on the mains (that could be heard in the speakers).
Eventually I found on Amazon a USB-C to cigarette lighter outlet cable, so now instead I'm using a compact GaN USB-C charger to power it, it works great!
It's also more amps than USB-C is rated for, so depending on the model of charger and fridge, this might not work reliably.
There are refrigerated coolers out there, which have essentially the same structure in miniature.
You still have cheapo compressors, insufficient or bad foam (some foams embrittle when exposed to sealed car temperatures, I don't know if that affects the R value but I wouldn't be surprised), but at least you can fix the seal, and have a form factor people are still familiar with.
It's only rated at 2.8A (34W) so it's within USB-C spec. It can probably exceed the spec when the compressor starts, but it hasn't been a problem.
Mine does have a thermostat control, and it can keep things frozen well. Used it with a portable 220Wh Li-ion battery to keep ice frozen during a party, it kept it frozen solid on a hot day in direct sunlight for 5 hours
As an engineering estimate the system wide COP of a peltier device is at most 0.05 vs pumped freon running around ten (to one sig fig). So with a solid state peltier device the 30 watt device sucks 1.5 watts of heat out of the fridge and dumps 31.5 watts into the heatsink (your room, eventually). A pumped freon unit like the one in your kitchen would suck 300 watts of heat out of the fridge and dump 330 watts into your room.
Anyway the definition of a BTU can't be more accurate than 1% for a bunch of boring thermodynamic reasons, but a good engineering estimate is one BTU is 1/3 of a watt-hour. So running your cooler will suck about 5 or so BTU out of the fridge every hour.
A BTU unit is very convenient as it changes the temp of one pound of water by one degree. And to one sig fig a 20 oz soda bottle is about a pound. So one bottle in an otherwise empty cooler should drop about 5 degrees per hour in your cooler. Figure a "few" bottles means they'll drop about one degree per hour. Yeah its gonna take "a couple hours" as you state to cool it down.
A pumped freon fridge could cool your soda close to 200 times faster than your cooler. That seems "about right" as a case of beer takes a couple hours to be drinkable.
I'm not sure if there are any residential "convection fridges" on the market. I'm sure the moving parts fan would increase future sales figures. Another interesting idea would be something like "NEST" thermostat to control the cooler temp while using as little power as possible. Might make some interesting startup ideas.
Anyway if you want to check your math, I did a quick experiment. My fridge cooled a 1.5 liter bottle of room temperature (25.6 C) tap water to 18.4 C in one hour, and 11.9 C in 2.
The new USB Power Delivery 3.0 PPS (Programmable Power Supply) standard even supports requesting arbitrary voltages, but I haven't yet found an adapter that is programmable and supports this.
Also the sound is better with the external power bank, as the DSP limits volume and bass response when running on internal power. The PB100 is good value for the money, but don’t expect hifi quality. Its currently the best off-the-shelf solution for something you can still stuff in a backpack, unless you spend the big bucks and go with a Soundboks.
They really half-assed it tho:
- the LED lights are erratic and don't look that great
- the BT range is crap
- no built-in connector to connect external DC power like in the PB300
- internal battery life is not great, just 2400mAh, also not swappable
- no second set of tweeters like in the PB300
- takes a long time to boot (~20secs)
- no app to adjust the DSP or disable the startup sound
Very, very few USB PD power supplies support PPS 20V at full 5A.
https://www.kickstarter.com/projects/zendure/supertank-pro-t...
This is why a lot of charger makers went for "workarounds" for 5A cables rather than risk being blamed for setting mismarked cables on fire.
Their cable marked "65W" did not actually work on 5A 15V, but on 3A 20V where it is not the issue, but will melt when used by actual 5A 15V.
On windows there are the MUTT tools, but looks cumbersome: https://docs.microsoft.com/en-us/windows-hardware/drivers/us...
https://www.youtube.com/watch?v=hMlySYTHbyk
https://www.banggood.com/POWER-Z-PD3_0-QC4_0-USB-Tester-Volt...
I understand the many, many articles that claim it's a mess. And although that may be true, I'm very happy with USB-C.
On the other hand, this is expected behavior for a USB-A charger, and thus for a USB-A-to-C cable.
I'm thinking of trying to convert it to USB-C, similar to what this guy did: https://youtu.be/V-vFtiDYiIw?t=538
Sort of like the cat and mouse of lighting cables.
I fully expect it to let the smoke out some day, but with a supply voltage less than half what it's delivering, I don't think it can fail in a way that'll hurt the Chromebook, and until it does I'm happy to have it as an option. And I bet those dinky little un-heat-sinked boost converters from the article likewise get pretty warm, too.
(Also, passive USB-C to USB-A female converters are trivially inexpensive, enough so that I just got a big bag of them so I don't have to care if they disappear on account of being tiny. For a new-build project, it makes sense to go straight to the connector, but for back compat purposes I can definitely recommend having a good supply of them on hand - USB-C cables seem in general to be either expensive or unreliable still, so being able to adapt old cables to new sockets saves a lot of headache. Especially when the old cables are permanently mounted to devices!)
[1] https://www.amazon.com/gp/product/B01ID90K4A - no longer available, but plenty of similar ones are. Don't trust any of 'em, would be my advice
Wait, for $190 you can buy a better used i3/i5 laptop? At least in pre-codvid times, not sure about today.
Covid really ramped up prices. The "big" webcams we use in our conference rooms just got 30% more expensive since the start of the whole Covid situation. Prices also seem to stay up high. Also delivery takes 8 weeks or more now instead of one week.
Supply and demand I guess...
I have a little array of SBCs powered off of a USB charging station, except for the switch, and one SBC that demands a 12V barrel connector.
USB-C to barrel connectors as packaged goods would have been useful. As far as I'm aware USB is the only standard charging connector that allows for negotiating voltage and amperage. We should use it more (or perhaps have another standard that can also handle USB charging).
Comparable to ChromeOS's USB-PD Sniffer Twinkie [1]
They are around 5 dollar.
I wonder if some adapters couldn't be designed to replace a USB-A port? It would be nice to solder them in place, and pull two extra cables to the charging circuit.