Date: Sun, 19 Dec 2021 17:13:42 -0800 From: Mark Millard via freebsd-arm <freebsd-arm@freebsd.org> To: MJ <mafsys1234@gmail.com>, bob prohaska <fbsd@www.zefox.net> Cc: freebsd-arm@freebsd.org Subject: Re: Dealing with slow USB disks, was: Re: Saving environment variables in u-boot Message-ID: <D6C6D413-BE4A-44CA-8E66-A379C9CD79A6@yahoo.com> In-Reply-To: <b0fbd30e-53f9-fe72-568d-1cd368e76285@gmail.com> References: <20211217013613.GA4452@www.zefox.net> <E2F6D50B-694A-4108-BD84-C85BC96AD832@yahoo.com> <20211218005946.GA7670@www.zefox.net> <5C44D0E6-2FF1-4EEB-B21A-83333D6FCF46@yahoo.com> <9D416106-660F-40BB-98D2-1354B53D2FEF@yahoo.com> <EC6CC83C-BC0A-4A12-866A-9FA24083FF7E@yahoo.com> <20211218223543.GA9484@www.zefox.net> <6105a8a6-e760-2183-72fd-92e5a60aa8df@gmail.com> <20211219005134.GA12292@www.zefox.net> <4910504f-3051-9a95-d8e4-95434042196d@gmail.com> <20211219161816.GA14873@www.zefox.net> <6edfdb4a-8ed3-7ef3-c3b2-7d2d7fd3206c@gmail.com> <1387CAB3-CFE7-4C47-8A9C-24D6E56D8C3C@yahoo.com> <b0fbd30e-53f9-fe72-568d-1cd368e76285@gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2021-Dec-19, at 15:48, MJ <mafsys1234@gmail.com> wrote: > On 20/12/2021 9:55 am, Mark Millard wrote: >> On 2021-Dec-19, at 13:39, MJ <mafsys1234@gmail.com> wrote: >>> On 20/12/2021 3:18 am, bob prohaska wrote: >>>> On Sun, Dec 19, 2021 at 10:03:47PM +1100, MJ wrote: >>>>>=20 >>>>> I would think a mechanical USB is going to pull a "lot" of power = when beginning spin-up, but once rotating should be easily powered by a = USB hub. Though this would not explain how it works on RPI4 unless the = powered hub you're using is USB2. >>>>>=20 >>>> That's what I thought too. I certainly didn't expect the disk to = work >>>> without a powered hub. The Pi4 is a different animal; it has USB3 = ports >>>> and more power available. That the mechanical disk works at all on = the >>>> Pi3's USB2 ports without assistance is quite surprising. >>>=20 >>> See here: = https://www.raspberrypi.com/documentation/computers/raspberry-pi.html#typi= cal-power-requirements >>> It mentions 1.2A, which covers your load, but, I would suspect if = you attached, for example, a USB 'thumb' drive or other devices you = would cause problems. >> Unfortunately there is more involved: USB2's standard >> indicates 500mA (0.5A) at maximum on 1 standard USB2 >> port. It takes 2 USB2 ports to get to a total of 1000mA >> (1A) (unless a port is designed to go beyond the >> standard). To my knowledge most RPi*'s are not designed >> to support more than the standard USB2 power on any of >> its USB2 ports. (The 3A+, Zero W/WH, and Zero are >> apparently exceptions, depending on the power supply >> used and such.) >=20 >=20 > You are correct, however, it seems the RPI foundation violates the = "standard" and creates their own for Bs: >=20 > = https://raspberrypi.stackexchange.com/questions/51615/raspberry-pi-power-l= imitations > (Section: How much current can be drawn from the USB ports?) >=20 Quoting that material: QUOTE The USB hub on the B models does not appear to be compliant to the USB specification and does not limit current. Individual ports can supply in excess of 500 mA independent of negotiation, subject to the overall maximum limit and adequate power supply. END QUOTE That such is not documented by the RPi* materials suggests that they consider it unsafe/unreasonable to depend on the behavior on at least some RPi*'s (if it is indeed as described). Still, it may explain why the bus-poered single-USB2 port spin-ups have gone as well as they have. There is also a claim there of a part present but not documented on the published schematics. This might mean that the status relative to that part has varied in time without the model number(s) changing. Depending on the part's behavior without checking for its existence on the specific RPi* might be a risk. =3D=3D=3D Mark Millard marklmi at yahoo.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?D6C6D413-BE4A-44CA-8E66-A379C9CD79A6>