开云体育

On 12/11/22 8:07 PM, Anne Ranch wrote:

My PC is "open architecture " ASUS motherboard in old desktop case. nanoVNA USB port is via powered D_Link " USB hub.
It has not been powered for about 6 months ,,,, I am going to let it fully discharge overnight.
It looks as setting to "continuous sweep" drains the battery pretty fast - perhaps the flashing
"sweep LED" is "too much load "?

The OS still controls (or at least manages) the hub's power as part of the whole USB tree, even if the hub itself receives external power. That is, the hub can decide if it wants to turn power on to a downstream device or not, and the OS (in theory) can find out. The OS can also force a downstream hub to turn off a device.

Considering that (almost) all USB hubs are 4 port devices at the "building block" level, the interlocking state diagrams up and down the chain can be really complex to figure out exactly what is going on.
(Note that this four-ness, is why you see 7 port hubs. It's actually a cascade of two 4 ports, with one hanging off one of the 4 ports of the other).

There's an equally complex negotiation and state machine for data rate through a USB hub.

I was going to quit this thread, but...
I left the nanoVNA running overnight , connected to my PC and I did "shutdown" and turned the PC power source off.
Turned off the D_link power source too.

In the morning I got no LCD and no LED's.
Turned all AC power on (PC , D-Link) and run "lsusb".
Still no LCD and LED's and no nanoVNA USB device.

Unplugged nanoVNA from USB port.
Unplugged "wallwart " from AC !!!
Plugged nanoVNA to dead wallwart and - LCD came on, so did both LEDs - BUT briefly

Plugged wallwart to AC and have normally operating nanoVNA
LCD on, charging LED on , sweep LED flashing,,,

My conclusion / opinion

in my interpretation of USB "standard operation"

...when a USB device is first plugged into its controller
it runs kind of "here I am " sequence and the controller should acknowledge that and respond ...

How does supposedly dead nanoVNA manages to do this when plugged in DEAD power source ?

5v power should *always* be available on a USB port, so slow charging can
happen even on a dead device. What is negotiated are faster charge rates,
so, as you noted, a dead device will need to slow charge to the point that
it can power up before it can take a fast(er) rate of charge.

(FWIW, I have never charged my NanoVNA H4 on anything *but* my laptop USB
port (no external hub - just plugged it in. Seems to work fine.)

- Tim

--
Tim Dawson (tadawson@...) Owner/Engineer
TPC Services Bellnet: (972)-221-7385
Lewisville, Texas 75067
"The world is complex. Sendmail.cf reflects this...."

On 12/12/22 7:13 AM, Anne Ranch wrote:

I was going to quit this thread, but...
I left the nanoVNA running overnight , connected to my PC and I did "shutdown" and turned the PC power source off.
Turned off the D_link power source too.
In the morning I got no LCD and no LED's.
Turned all AC power on (PC , D-Link) and run "lsusb".
Still no LCD and LED's and no nanoVNA USB device.
Unplugged nanoVNA from USB port.
Unplugged "wallwart " from AC !!!
Plugged nanoVNA to dead wallwart and - LCD came on, so did both LEDs - BUT briefly
Plugged wallwart to AC and have normally operating nanoVNA
LCD on, charging LED on , sweep LED flashing,,,
My conclusion / opinion
in my interpretation of USB "standard operation"
...when a USB device is first plugged into its controller
it runs kind of "here I am " sequence and the controller should acknowledge that and respond ...
How does supposedly dead nanoVNA manages to do this when plugged in DEAD power source ?

Dead might not be really dead. For instance, your device might have dropped below minimum battery voltage threshold, so it shuts down. But upon detecting a change in state of an external interface (e.g. the USB), it might wake up, just enough to probe the interface.

Most USB hubs will pass *some* current downstream, even if their wall wart is plugged into the device, but not the wall. That is, they alter their behavior depending on if external power is supplied.

I have a 4 port hub that will support high current downstream either if it's connected to a wallwart, OR, it's connected to a high current upstream USB. BUT, if the upstream device changes its mind, the downstream device doesn't always figure that out.


And the whole chain/tree can be partially or fully renegotiated. So it's not just "first plug in sets the behavior".

The overall behavior is really complex, partly because you can be compliant with the USB spec, and not do full negotiation. That is, there's allowance for implementations that don't do all the functionality, but part of it, but in a safe way.

And, of course, there are "almost compliant" devices out there too.

I spent months debugging a battery powered system that used USB for custom modules we built, driven by an upstream PC and a Dlink 7 port hub.

If you're on Windows:




From another document


In some cases, Windows will “forget” that the port exists (as documented in a Microsoft Knowledge Base article 817900) , probably due to the device disappearing and reappearing as the microcontroller changes modes. This can confuse the power management logic implemented in Windows. The problem is also known as the “USB root hub power management issue”. The fix is to disable the selective suspend behavior for USB devices, which can be done from the Power Management options in the control panel. There is probably a way to do this using registry keys as described in the MS knowledge base article, but the instructions given there specify registry keys and areas that do not exist.

Figure 2 - Windows Power Management Dialog


Figure 3 - Showing USB selective suspend enabled (undesirable)


Figure 4 - Showing USB Selective Suspend disabled (desirable)




(There's a reason Win NT didn't support USB when it first came out, and why USB support in Linux was really sketchy and unpredictable, 10-15 years ago)

On 12/12/22 7:34 AM, Tim Dawson wrote:

5v power should *always* be available on a USB port,

Not necessarily - there are USB hubs that if they get a request for high power, and can't supply it, just shut down the power and disable that port, other than some microamps to tell if there's a unplug/plug event.


so slow charging can

happen even on a dead device. What is negotiated are faster charge rates,
so, as you noted, a dead device will need to slow charge to the point that
it can power up before it can take a fast(er) rate of charge.
(FWIW, I have never charged my NanoVNA H4 on anything *but* my laptop USB
port (no external hub - just plugged it in. Seems to work fine.)
- Tim
On Mon, December 12, 2022 9:13 am, Anne Ranch wrote:

I was going to quit this thread, but...
I left the nanoVNA running overnight , connected to my PC and I did
"shutdown" and turned the PC power source off.
Turned off the D_link power source too.

In the morning I got no LCD and no LED's.
Turned all AC power on (PC , D-Link) and run "lsusb".
Still no LCD and LED's and no nanoVNA USB device.

Unplugged nanoVNA from USB port.
Unplugged "wallwart " from AC !!!
Plugged nanoVNA to dead wallwart and - LCD came on, so did both LEDs -
BUT briefly

Plugged wallwart to AC and have normally operating nanoVNA
LCD on, charging LED on , sweep LED flashing,,,

My conclusion / opinion

in my interpretation of USB "standard operation"

...when a USB device is first plugged into its controller
it runs kind of "here I am " sequence and the controller should
acknowledge that and respond ...

How does supposedly dead nanoVNA manages to do this when plugged in DEAD
power source ?

I think those who have been around longer, find the better sites for documentation. When they are posted, GREAT!

A side issue to this quality/source, can be the QUANtity/source.

I.E., I used to suggest 2 books for any given computer topic. One ought to be through, maybe even a reference. But the other, while far more than a Quick Start Guide, shod be far far simpler, not much more than an overview.

And ideally, they should be from different authors and publishers. This way, you don't get an abridged version or expanded version, in the same words and from the same point of view, but you get an, "IOW," viewpoint and vernacular.

I've no idea if such a thing is remotely possible w VNA's, and then, there are so many versions now.

If so, the instructions and/or a link to them, per version, would make a good set of list-files. I do download instructions. (So many are not well written.) They can be handy and may be searchable... but I really dislike lack of a printed set, and will print them out myself, sometimes. YMMV...

BillSF9c

"(A very small amount of reading on your part could reduce the amount of time that others take to voluntarily help you.)"

..on the other hand, repeatedly mentioning this fact could reduce the volunteered time for assistance could set the time spent helping to zero as a limit.

Patience and gratitude and good manners go a long way on most of of these groups.
--
William, k6whp
--------------------
"Cheer up, things could get worse. So I cheered up and things got worse."