Re: Boot from USB on RPi4 8GB? [SOLVED]

On 2021-Jun-12, at 18:18, William Carson <freebsd at dsllsn.net> wrote:

> On Jun 10, 2021, at 9:57 PM, Mark Millard via freebsd-arm <freebsd-arm@freebsd.org> wrote:
>> 
>> On 2021-Jun-10, at 17:22, William Carson via freebsd-arm <freebsd-arm at freebsd.org> wrote:
>> 
>>> On May 30, 2021, at 6:54 PM, William Carson via freebsd-arm <freebsd-arm@freebsd.org> wrote:
>>>> 
>>>> Thank you for hopefully pointing me in the right direction.
>>> 
>>> Alright, so, I ended up buying a WDS500G2B0C, which seems to only use a maximum of 75 mW (this seems ... low, to say the least, but the column is unlabeled in the spec sheet I found here https://documents.westerndigital.com/content/dam/doc-library/en_us/assets/public/western-digital/product/internal-drives/wd-blue-nvme-ssd/product-brief-wd-blue-sn550-nvme-ssd.pdf).
>> 
>> That document says for WDS500G2B0C:
>> 
>> QUOTE
>> Maximum Operating power    3.5W
>> END QUOTE
>> 
>> So at about 5V: about 0.7A maximum. Of itself, that is under the 1.2A
>> figure. The X872 must take some power as well. But I do not find any
>> W or A specifications for it.
>> 
>> I do not know what other power draws in the overall system might be.
>> Any combination that might use more than about 0.5A? (So: more than
>> about 2.5W total)?
>> 
>> Unless you specify otherwise, I'm going to presume: no. So I'm not
>> expecting power to be a problem.
> 
> Here is my setup, trying to be precise as possible:
> 
> Static components:
>    - Apple 30W USB-C Power Adapter
>    - Apple USB-C Charge Cable (2m)

The original RPi4B's (v1.1) were missing a resistor and
caused various power supplies to misidentify the RPI4B
context. But I doubt this is your problem: normally that
blocks the supply from working at all if I understand
right.

But there is the problem that the RPi4B only supports part
of the Power Delevery specifications and does not negotiate
current or power at all: just uses a register to tell an "e"
PD power supply connection that it needs about 5V.

As one RPI engineer/forum monitor put it in response to
reported experimental results (and a misinterpetation of
them) for a RPi4B:

QUOTE (2021-May-31 material):
The PI does NO negotiation with power supply under any circumstances.

Your results are down to different power supply ability, not any sort
of negotiation. i.e. some supplies are better than others.
END QUOTE

Apple USB-C power adapters (only listing 5.?V
combinations supported):

Older:
29W USB-C: 5.2V/2.4A
61W USB-C: 5.0V/2.4A (an older model)
87W USB-C: 5.2V 2.4A

Newer (USB-C Power Delivery 2.0+):
18W USB-C: 5V/3A
30W USB-C: 5V/3A
61W USB-C: 5V/3A (a newer model)

Note that none of those with 3A are 5.1V, which is what
the RPi4B is indicated to require. Using a true 5V has
more risk of voltage droop problems as I understand.

https://www.raspberrypi.org/documentation/hardware/raspberrypi/power/README.md

reports:

QUOTE
The power supply requirements differ by Raspberry Pi model.
All models require a 5.1V supply, but the current supplied
generally increases according to model.
END QUOTE

(So far as I know, the 5.1V requirement only fits with
USB-C at all via the margins allowed around the nominal
USB-C figure.)

The CanaKit 5.1V 3.5A ones that I've been using have worked
well. The Power Supply from the RPi folks is 5.1V/3.0A.
Look for 5.1V and 3.0A to no more than, say, 4.0A or less.

It might be worth testing with such a power supply instead
of the Apple one. (Unless you can measure and see that you
are getting the 5.1V.)

>    . . .

For the moment I'm going to stop with this one point
before looking at your other material.


===
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)