Searching \ for '[EE] USB power in parallel' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/power.htm?key=power
Search entire site for: 'USB power in parallel'.

Exact match. Not showing close matches.
PICList Thread
'[EE] USB power in parallel'
2008\07\03@204951 by Dr Skip

picon face

I have a few external drives that have a dual usb end for tapping more power
for the drive by using 2 ports. In most cases, these drives won't work with
just one port. Looking at drive specs on portable, port-powered drives like the
 Western Digital ones, they can take 1000mA to spin up, and many take 700mA.
With port power restricted to 500mA, it makes sense why I can hear it can't
spin up with just one usb connection.

I'm assuming the second connection to the pc is power only. How can they just
parallel ports? Is that part of the spec, or just a hack (like making a simple
and gate by paralleling OC outputs)?

Has anyone built an add-on pack that would just power during spin-up by
paralleling batteries and perhaps a regulator? Not all the time, so the port
can power the rest of the time and save the batteries. Could it be sensed by
voltage drop? Do any simple, passive circuits come to mind?

Thanks,
Skip

2008\07\03@210057 by Xiaofan Chen

face picon face
On Fri, Jul 4, 2008 at 8:48 AM, Dr Skip <spam_OUTdrskipTakeThisOuTspamgmail.com> wrote:
>
> I have a few external drives that have a dual usb end for tapping more power
> for the drive by using 2 ports. In most cases, these drives won't work with
> just one port. Looking at drive specs on portable, port-powered drives like the
>  Western Digital ones, they can take 1000mA to spin up, and many take 700mA.
> With port power restricted to 500mA, it makes sense why I can hear it can't
> spin up with just one usb connection.
>
> I'm assuming the second connection to the pc is power only. How can they just
> parallel ports? Is that part of the spec, or just a hack (like making a simple
> and gate by paralleling OC outputs)?

They are out of USB spec. Did you ever see a USB logo on this disks?
They can not have it.

Xiaofan

2008\07\03@214146 by Herbert Graf

flavicon
face
On Thu, 2008-07-03 at 20:48 -0400, Dr Skip wrote:
> I have a few external drives that have a dual usb end for tapping more power
> for the drive by using 2 ports. In most cases, these drives won't work with
> just one port. Looking at drive specs on portable, port-powered drives like the
>   Western Digital ones, they can take 1000mA to spin up, and many take 700mA.
> With port power restricted to 500mA, it makes sense why I can hear it can't
> spin up with just one usb connection.
>
> I'm assuming the second connection to the pc is power only. How can they just
> parallel ports? Is that part of the spec, or just a hack (like making a simple
> and gate by paralleling OC outputs)?

It is absolutely a hack, and is not in spec, in the cases I've seen.

The way USB works is you must ask permission to take more then a certain
minimal amount of current. The "main" USB connection likely properly
requests a full 500mA. In all the cases I've seen the second cable
doesn't even connect to the USB data pins, but only takes power.

In a perfect world the ports share the duty, and the current is equally
divided between the two. In the real world it's likely that in many
cases one port will source more then then other.

Ignoring that, many USB ports just can't supply the full 500mA. Laptops
are bad for that, but also any unpowered hub can't supply 500mA per port
since it gets it's power from one port.

While this hack mostly works, it can be hit and miss.

TTYL

2008\07\03@222809 by Jinx

face picon face
part 1 165 bytes content-type:text/plain; (decoded 7bit)

> Has anyone built an add-on pack that would just power
> during spin-up by paralleling batteries and perhaps a regulator?


part 2 3357 bytes content-type:image/gif; (decode)


part 3 35 bytes content-type:text/plain; charset="us-ascii"
(decoded 7bit)

2008\07\03@223124 by Jinx

face picon face
Dr Skip, typo in my gif labelling. Vusb should read Vbus. That
doesn't affect the circuit though

2008\07\03@225413 by Jinx

face picon face
> Ignoring that, many USB ports just can't supply the full 500mA.
> Laptops are bad for that, but also any unpowered hub can't supply
> 500mA per port since it gets it's power from one port

I'm not au fait with USB power standards now, but as explained in
the Silicon Chip article of 10/04 -

"Each USB socket of a PC or self-powered USB hub can supply up
to 500mA at 5VDC ...... Many low-cost USB hubs are also designed
to take their own power from the PC via their "upstream" USB cable...
..things start to get a little more complicated if you try to connect a
number of bus-powered peripherals to your PC via such a hub, because
the hub's "downstream" output sockets can each supply only 100mA
max. That's because all of their power must ultimately come from the
PC"

2008\07\03@230540 by Dr Skip

picon face
I checked and surprise! there's no USB logo on the box with dual head cable.
There also isn't one on the WD portable drive, and they're a big company...
Although they don't have a dual head cable, I've noticed many of my laptops
won't kick it on, and there are lots of complaints on their forums about that.

It's got a small single chip sata interface on it, and its spec is 1A spin up,
but the internal drive can have a 'low power spin up' by jumper, which is in.
It still draws too much for many machines.

Any thoughts on a big cap in parallel with the power lines in the cable to
source add'l current on startup? ;) I'd like to just boost spin up as needed,
externally.

I've also seen where the device (a low power one) didn't work with a particular
length cable. One was an Olympus digital audio recorder. I summed it up as an
impedance/frequency problem with terminations at one end or the other. It
wouldn't work or even be seen with a couple laptops (this was only seen with
laptops) but adding a small extension cable made it work. Others were fine.
I've seen this with the WD drive on a laptop it does work on regularly. There
are certain lengths of cable that it doesn't even get seen with.

-Skip

Herbert Graf wrote:
> On Thu, 2008-07-03 at 20:48 -0400, Dr Skip wrote:
>
> It is absolutely a hack, and is not in spec, in the cases I've seen.

2008\07\03@231506 by William \Chops\ Westfield

face picon face

On Jul 3, 2008, at 5:48 PM, Dr Skip wrote:

> I'm assuming the second connection to the pc is power only. How can  
> they just
> parallel ports? Is that part of the spec, or just a hack (like  
> making a simple
> and gate by paralleling OC outputs)?

The "power only" connection doesn't have the data pins connected.  I  
found out the hard way when I plugged in the wrong (non-data) jack of  
one of those double cables, and a drive wasn't seen at all.

The typical USB port has a PTC fuse or similar in it, if it has any  
protection at all.  OFTEN these don't blow very quickly at moderate  
(2x) overcurrent, so you survive with only a single connection.  If  
you use two ports, the PTC fuses probably make "fine" current-sharing  
resistors.  OCCASIONALLY you get a system that actually measures  
current draw on USB ports, and they complain.  (IIRC there was a  
generation of Apple laptops that did this.)

BillW

2008\07\03@232742 by Dr Skip

picon face
part 0 44 bytes
his is a multi-part message in MIME format.
part 1 3493 bytes content-type:multipart/related; (decoded base64)

NƦ[bgs0Gf 2008\07\03@235502 by Bob Ammerman
picon face
A USB device has to negotiate with the host if it needs more than (IIRC)
100ma. Since the second USB connection on these drives is general
power-only, and thus cannot negotiate, any attempt to draw more than 100ma
is outside the USB spec. This of course applies also to lights, fans, wet
fish dryers, cell phone chargers and all the other widgets that draw power
from USB ports without a data connection. In reality, most systems (not
including hubs and a few laptops) will provide 500ma without negotiation on
each port.

I have an external 2.5" drive case which uses this technique and it has
always just worked for me on perhaps a dozen different computers including
several Dell laptops.

Smalll series resistors on  two Vbus lines can help ensure proper sharing of
the load.

-- Bob Ammerman
RAm Systems

2008\07\04@005222 by Xiaofan Chen

face picon face
On Fri, Jul 4, 2008 at 11:50 AM, Bob Ammerman <.....rammermanKILLspamspam@spam@verizon.net> wrote:
>
> I have an external 2.5" drive case which uses this technique and it has
> always just worked for me on perhaps a dozen different computers including
> several Dell laptops.

We have two 2.5" USB enclosures at home using the same dual USB power
which refused to work with several Dell Desktops and Laptops. It sort of
works. But copying large quantity of files will sometime lead to data error
or even connection loss. Now we would not use them any more and we
would not buy any 2.5" external USB harddisks.

On the other hand, we have four 3.5" USB at home using external power
and they all work fine.

Xiaofan

2008\07\04@012916 by Dr Skip

picon face
Yes, that seems to be a big problem, and the root of my query. One
enclosure I have has a rechargable battery in it for power, and another
allows for bus power or a plug in battery, but both are also USB OTG
type units. The plain external drives have a nice promise - large
capacity and portability without wall warts, but it doesn't come true it
seems, and no other option.

Another question... How tolerant is usb +V? ie, 4 NimH cells may be 6v
fresh, but quickly go to 5v, then drop to 4v at end of use for that
charge. Not enough headroom for a simple regulator, but will these
temporary overages bother a device? How does it fail at low voltage?

An idea as I write this - Use the 5v to charge 4 cells and use the cells
for power too. I suspect this is turning into a switcher with asst
charging and other components though... :(

A large cap (supercap maybe) on +V of the internal drive perhaps? ;)



Xiaofan Chen wrote:
{Quote hidden}

2008\07\04@040248 by Alan B. Pearce

face picon face
>I'm assuming the second connection to the pc is power only. How can
>they just parallel ports? Is that part of the spec, or just a hack
>(like making a simple and gate by paralleling OC outputs)?
>
>Has anyone built an add-on pack that would just power during spin-up
>by paralleling batteries and perhaps a regulator?

Dell do this with their current model laptops. The CD/DVD drive bay has a
USB and a power connection so that it can be used with a CD/DVD, second hard
drive or second battery. They call it a D/Bay (I think short for Docking
Bay), and they replicate the D/Bay port on their Port Replicator and Docking
Station. On my Port Replicator it is on the left, and has a USB port stacked
with the power port in a manner that looks like two USB ports, except the
power port has obviously heavier looking contacts and you cannot plug a USB
device into it in a manner that allows the contacts to mate.

2008\07\04@051809 by Jinx

face picon face
> 4 NimH cells may be 6v fresh, but quickly go to 5v, then drop to
> 4v at end of use for that charge. Not enough headroom for a simple
> regulator

You could go to 6 cells couldn't you ? And a buck-boost SMPS to
get 5V from 4V-6V doesn't necessarily have to be too complicated.
Even just a boost followed by an LDO linear perhaps

2008\07\04@065527 by olin piclist

face picon face
Xiaofan Chen wrote:
> They are out of USB spec. Did you ever see a USB logo on this disks?
> They can not have it.

Why?  What spec exactly do they violate?  It seems to me you could do this
without running afoul of the spec.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\04@070304 by olin piclist

face picon face
Herbert Graf wrote:
> It is absolutely a hack, and is not in spec, in the cases I've seen.
>
> The way USB works is you must ask permission to take more then a
> certain
> minimal amount of current. The "main" USB connection likely properly
> requests a full 500mA. In all the cases I've seen the second cable
> doesn't even connect to the USB data pins, but only takes power.

Yes, that would violate the spec.  And compliant hosts don't need to supply
much power until you ask for it and they feel like giving it to you.  You
are not guaranteed the full power.

However, a second cable by itself doesn't guarantee violation of the spec.
If separate communication is established over that cable, the extra power is
requested and granted, then there is nothing wrong.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\04@071346 by olin piclist

face picon face
Xiaofan Chen wrote:
> We have two 2.5" USB enclosures at home using the same dual USB power
> which refused to work with several Dell Desktops and Laptops. It sort
> of works. But copying large quantity of files will sometime lead to
> data error or even connection loss. Now we would not use them any
> more and we
> would not buy any 2.5" external USB harddisks.

With flash drives being so cheap and available, unless you need to transfer
very large amounts of data there is little reason for external USB hard
drives.  I carry my little 8Gb flash drive in my pocket everywhere I go.
It's my backup and I also use it to keep the three machines I'm routinely on
in sync, which then become further backups.

Flash drives are slow and small in comparison to hard drives, but are
certainly small and low power.

By the way, I found out the hard way they don't go thru the laundry well.
At least that gave me a excuse to upgrade from 4Gb to 8Gb.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\04@072050 by Tamas Rudnai

face picon face
> However, a second cable by itself doesn't guarantee violation of the spec.
> If separate communication is established over that cable, the extra power
is
> requested and granted, then there is nothing wrong.

I thought these kind of devices use the secondary USB plug as a power
connector - like a USB lamp or other dumb gadgets without actually
interfacing with the USB Host. Not sure if that is on the specs and what
current can a dumb device withdraw like this?

Tamas



On Fri, Jul 4, 2008 at 12:05 PM, Olin Lathrop <.....olin_piclistKILLspamspam.....embedinc.com>
wrote:

{Quote hidden}

> -

2008\07\04@123615 by Dr Skip

picon face
I thought so too - there's no intelligence built into the cable AFAIK,
and I can't see any components in it, so it must be paralleling the
power lines.

I noticed there is one of these power ports on a used Dell laptop I
have. Any idea where to get a cable or connector (hopefully cheaply).
That Dell is the worst in terms of power out on usb too. I even tried a
drive that wouldn't spin up on a powered hub (external wall wart) and it
went for a while and then went away a few hours later while writing to
it. So much for portable drives?


Tamas Rudnai wrote:
{Quote hidden}

2008\07\04@172245 by William \Chops\ Westfield

face picon face

On Jul 4, 2008, at 4:16 AM, Olin Lathrop wrote:

> With flash drives being so cheap and available, unless you need to  
> transfer
> very large amounts of data there is little reason for external USB  
> hard
> drives.

My main use for USB drives is backup, and unfortunately flash drives  
are MUCH too small. (Photos, video, etc.)  The ideal drive is big  
enough to backup all the personal data (including duplicated data,  
since first-effort backup is to load things on more than one system)  
from the five family systems (300gb ?), more than once, and small  
enough to fit in a safe-deposit box.  I do wish there was a smarter  
backup utility...

>
> Flash drives are slow and small in comparison to hard drives, but are
> certainly small and low power.

They may not be as "low power" as you think.  Several of mine refuse  
to work on the far side of a hub, and some get noticeable WARM when  
in use.  Almost certainly "low power" compared to a hard drive, I guess.


> By the way, I found out the hard way they don't go thru the laundry  
> well.

Ouch.  And as I commented elsewhere, they keep making them in form-
factor of things that I tend to LOSE all the time (notably pens...)

BillW


2008\07\04@175147 by olin piclist

face picon face
William Chops" Westfield" wrote:
> My main use for USB drives is backup, and unfortunately flash drives
> are MUCH too small. (Photos, video, etc.)  The ideal drive is big
> enough to backup all the personal data (including duplicated data,
> since first-effort backup is to load things on more than one system)
> from the five family systems (300gb ?), more than once, and small
> enough to fit in a safe-deposit box.  I do wish there was a smarter
> backup utility...

Do you really need to *back up* 300Gb.  Chances are you are only actively
modifying a tiny portion of that and the rest would be more appropriate for
archiving.  Photos are a good example.  They take up a lot of space, but
once you're done with scanning, color correcting, annotating, or whatever,
they don't generally change.  I've got 10s of Gb of photos too, but I don't
back them up every day.  Instead if have a few copies archived on CD and
DVD.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\04@182655 by William \Chops\ Westfield

face picon face

On Jul 4, 2008, at 2:53 PM, Olin Lathrop wrote:

> Do you really need to *back up* 300Gb.  Chances are you are only  
> actively
> modifying a tiny portion of that and the rest would be more  
> appropriate for
> archiving.

Ah.  If you want to be exacting about your definitions, I mostly  
"archive" rather than "back up."  The "work" systems are backed up  
automagically by the IT department.

"Archive management" seems to be behind even the sorry state of
"backup management."

BillW

2008\07\04@203628 by Xiaofan Chen

face picon face
On Sat, Jul 5, 2008 at 5:53 AM, Olin Lathrop <olin_piclistspamspam_OUTembedinc.com> wrote:

> Do you really need to *back up* 300Gb.  Chances are you are only actively
> modifying a tiny portion of that and the rest would be more appropriate for
> archiving.  Photos are a good example.  They take up a lot of space, but
> once you're done with scanning, color correcting, annotating, or whatever,
> they don't generally change.  I've got 10s of Gb of photos too, but I don't
> back them up every day.  Instead if have a few copies archived on CD and
> DVD.

Be careful here. CD does not last long. Quite some of my pre-year-2002
archive CDs have some problems. Some of the got the content totally
gone. Luckily I got them backed up to a harddisk back in 2004.

Now I have a DVD/RW but I only use it to create no-essential things
like Linux installation disks.


Xiaofan

2008\07\04@233512 by Peter Todd

flavicon
face
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, Jul 04, 2008 at 05:53:59PM -0400, Olin Lathrop wrote:
{Quote hidden}

The advantage of backups, compared to archives, is that the constant
usage will mean you are much more likely to notice if one of your (many
I hope) backups is going bad. No so likely to happen with an archive.

Of course, you still need to protect against insidious data corruption
as well... Supposedly the XFS filesystem is good for that, but I haven't
tried it yet. I haven't even setup a simple script to check sha1sums on
a regular basis against a master list of hashes for my archive
directories, I really should.


Perhaps a utility that automagically warns you when normally
infrequently changed files get changed? Or, especially, when the last
time the file has been changed isn't what the file timestamp says...

- --
http://petertodd.org 'peter'[:-1]@petertodd.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFIbuLJ3bMhDbI9xWQRAvUDAJ4rcIXfdSOVvQ0R60RaX0Ucn0cE5QCfeWU7
fZQnPXzKret5R3Mwmpy1SHM=
=3mQZ
-----END PGP SIGNATURE-----

2008\07\05@021714 by Dr Skip

picon face
I've had backup DVDs that were burned and verified, and random files retrieved
from them to test at t=0, as backups, then 6 months later only a third could be
recovered - and on the same drive that wrote them. Lower percentage or totally
unreadable on other drives tried. I no longer trust them.

For backups and such I use bigger AC powered drives, and am migrating to rsync.
These 200GB +/- 2.5" drives are good for taking my photography on the road with
some core apps, etc. Also for document repositories for travel. I can't fit one
or the other in under 100GB, which rules out solid state, and it's too tough to
try to pre-filter. I _always_ end up needing other files on site and can't
always trust connectivity (and the time to download several GB for what I need
at the moment is too long.

-Skip

Xiaofan Chen wrote:
{Quote hidden}

2008\07\05@025051 by SM Ling

picon face
>
> We have two 2.5" USB enclosures at home using the same dual USB power
> which refused to work with several Dell Desktops and Laptops. It sort of
> works. But copying large quantity of files will sometime lead to data error
> or even connection loss. Now we would not use them any more and we
> would not buy any 2.5" external USB harddisks.
>
Ask the Sim Lim square guy, technically they can't help much but they should
have enough reject to advise you which brand of the hard disk and what size
you should get, or none.  The enclosure and adapter are innocent.

Ling SM

2008\07\05@072037 by Xiaofan Chen

face picon face
On Sat, Jul 5, 2008 at 2:50 PM, SM Ling <KILLspamsm.ling11KILLspamspamgmail.com> wrote:
> Ask the Sim Lim square guy, technically they can't help much but they should
> have enough reject to advise you which brand of the hard disk and what size
> you should get, or none.  The enclosure and adapter are innocent.

Not really. They are not innocent and since now I know that they violate
USB specification I will not buy any of them.

Xiaofan

2008\07\05@115202 by Bob Ammerman

picon face
>> We have two 2.5" USB enclosures at home using the same dual USB power
>> which refused to work with several Dell Desktops and Laptops. It sort of
>> works. But copying large quantity of files will sometime lead to data
>> error
>> or even connection loss. Now we would not use them any more and we
>> would not buy any 2.5" external USB harddisks.
>>
> Ask the Sim Lim square guy, technically they can't help much but they
> should
> have enough reject to advise you which brand of the hard disk and what
> size
> you should get, or none.  The enclosure and adapter are innocent.
>
> Ling SM

My successes have been with Hitachi ATA drives, mostly 7200RPM! (60GB, 80GB
and 100GB), and also a 40GB 5400 RPM Hitachi.

-- Bob Ammerman
RAm Systems

2008\07\08@110222 by Tamas Rudnai

face picon face
> I thought so too - there's no intelligence built into the cable AFAIK,
> and I can't see any components in it, so it must be paralleling the
> power lines.

Just found that USB Lava Lamp I have (got it from colleagues, I would not
buy those :-)) So there is a 5V1W bulb in there, nothing else... That's
200mA, plus when it's cold it could be even bigger. There is no interface
asking more power than 100mA, just using the power lines. Isn't it violating
the specs? I can't find anything about powering dumb devices like this on
the USB specs? Should I look that harder? :-)

Tamas




On Fri, Jul 4, 2008 at 5:36 PM, Dr Skip <RemoveMEdrskipTakeThisOuTspamgmail.com> wrote:

{Quote hidden}

> -

2008\07\08@112718 by olin piclist

face picon face
Tamas Rudnai wrote:
> Just found that USB Lava Lamp I have (got it from colleagues, I would
> not buy those :-))

Of course.  We understand.  It's from a "friend".

> So there is a 5V1W bulb in there, nothing else...
> That's 200mA, plus when it's cold it could be even bigger. There is
> no interface asking more power than 100mA, just using the power
> lines. Isn't it violating the specs?

Yup, and you won't likely find the official USB logo on the plug.  It will
probably work fine (within the limits of whatever "fine" might be for a lava
lamp) on most desktops.  It will fail to work correctly on computers that
tightly manage their USB power, and then you couldn't blame the computer.
Try it on several laptops.  They can be more picky about USB power to
carefully manage battery drain.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\08@112743 by olin piclist

face picon face
Tamas Rudnai wrote:
> Just found that USB Lava Lamp I have (got it from colleagues, I would
> not buy those :-))

Of course.  We understand.  It's from a "friend".

> So there is a 5V1W bulb in there, nothing else...
> That's 200mA, plus when it's cold it could be even bigger. There is
> no interface asking more power than 100mA, just using the power
> lines. Isn't it violating the specs?

Yup, and you won't likely find the official USB logo on the plug.  It will
probably work fine (within the limits of whatever "fine" might be for a lava
lamp) on most desktops.  It will fail to work correctly on computers that
tightly manage their USB power, and then you couldn't blame the computer.
Try it on several laptops.  They can be more picky about USB power to
carefully manage battery drain.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\08@114129 by Tamas Rudnai

face picon face
> Of course.  We understand.  It's from a "friend".

I swear :-)

> Yup, and you won't likely find the official USB logo on the plug.

Well, you can mould anything you want, and they put a USB logo on it -
probably it's not official and the manufacturer may could have some legal
problems.

Tamas



On Tue, Jul 8, 2008 at 4:29 PM, Olin Lathrop <TakeThisOuTolin_piclistEraseMEspamspam_OUTembedinc.com>
wrote:

{Quote hidden}

> -

2008\07\08@115106 by olin piclist

face picon face
Tamas Rudnai wrote:
> Well, you can mould anything you want, and they put a USB logo on it -
> probably it's not official and the manufacturer may could have some
> legal problems.

In theory the USB organization could sue them for trademark infringement.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\08@131503 by Herbert Graf

flavicon
face
On Tue, 2008-07-08 at 11:29 -0400, Olin Lathrop wrote:
> > So there is a 5V1W bulb in there, nothing else...
> > That's 200mA, plus when it's cold it could be even bigger. There is
> > no interface asking more power than 100mA, just using the power
> > lines. Isn't it violating the specs?
>
> Yup, and you won't likely find the official USB logo on the plug.  

Chances are you will. I find that most "non-compliant" USB hardware has
the logo on the cable (and often even on the box).

It's unfortunate, but it appears that enforcement is non-existent.

TTYL

2008\07\08@142837 by Bob Blick

face
flavicon
face

On Tue, 08 Jul 2008 13:14:37 -0400, "Herbert Graf"
<RemoveMEmailinglist4spamTakeThisOuTfarcite.net> said:
> On Tue, 2008-07-08 at 11:29 -0400, Olin Lathrop wrote:
> > > So there is a 5V1W bulb in there, nothing else...
> > > That's 200mA, plus when it's cold it could be even bigger. There is
> > > no interface asking more power than 100mA, just using the power
> > > lines. Isn't it violating the specs?
> >
> > Yup, and you won't likely find the official USB logo on the plug.  
>
> Chances are you will. I find that most "non-compliant" USB hardware has
> the logo on the cable (and often even on the box).
>
> It's unfortunate, but it appears that enforcement is non-existent.

That's fine with me, the last thing I want is the court system tied up
because of rules broken in a "private club" like the USB association. Or
border agents wasting resources seizing USB lights. One thing leads to
another and the next thing you know, someone will be wanting all email
clients to be standards-compliant <ducks>.

Cheerful regards,

Bob

--
http://www.fastmail.fm - IMAP accessible web-mail

2008\07\08@235024 by William \Chops\ Westfield

face picon face

On Jul 8, 2008, at 8:02 AM, Tamas Rudnai wrote:

> Isn't it violating the specs? I can't find anything about powering  
> dumb devices like this on the USB specs? Should I look that  
> harder? :-)

As far as I know, the USB spec does not allow for ANYTHING like a  
"power only dumb device."  *ALL* that stuff violates the spec, as are  
any cables without an A male connector on one end and a B (or miniB/
microB) male connector (or a low speed device) connector on the other  
end...

BillW

2008\07\09@044209 by Alan B. Pearce

face picon face
>As far as I know, the USB spec does not allow for ANYTHING like a
>"power only dumb device."  *ALL* that stuff violates the spec,

I know my laptop complains that it cannot identify a USB device when I plug
my Bluetooth GPS in on  a power only cable to charge the battery. It detects
the current draw and expects to enumerate a device on the port, but the
device will not answer.

2008\07\09@045632 by Tamas Rudnai

face picon face
> I know my laptop complains that it cannot identify a USB device when I
plug
> my Bluetooth GPS in on  a power only cable to charge the battery.

So the D-/D+ is not connected at all in that USB power cable?

And what happens after saying that the device is not identified, does it
continue to provide the power?

Tamas



On Wed, Jul 9, 2008 at 9:41 AM, Alan B. Pearce <A.B.PearceEraseMEspam.....rl.ac.uk> wrote:

> >As far as I know, the USB spec does not allow for ANYTHING like a
> >"power only dumb device."  *ALL* that stuff violates the spec,
>
> I know my laptop complains that it cannot identify a USB device when I plug
> my Bluetooth GPS in on  a power only cable to charge the battery. It
> detects
> the current draw and expects to enumerate a device on the port, but the
> device will not answer.
>
> -

2008\07\09@061227 by Alan B. Pearce

face picon face
>> I know my laptop complains that it cannot identify a USB device when I
>> plug my Bluetooth GPS in on  a power only cable to charge the battery.
>
>So the D-/D+ is not connected at all in that USB power cable?

Well, it uses a standard cable, but the D+/D- pins are not connected in the
GPS.

>And what happens after saying that the device is not identified, does it
>continue to provide the power?

Yes, but every so often a pop up occurs moaning about the 'bad device' that
won't enumerate. I don't think it can turn the power off totally, just limit
it to the 100mA of the standard, unless an enumerated device asks for more.

2008\07\09@065734 by Geo

flavicon
face
On 8 Jul 2008, at 20:50, William "Chops" Westfield wrote:

> As far as I know, the USB spec does not allow for ANYTHING like a  
> "power only dumb device."  *ALL* that stuff violates the spec,

Yes - but how many devices have we used that pinched power from RTS, DTR
etc and no complaint about violating the RS232 spec.
It /is/ just a connector with 5 volts available - similar to keyboard PS2
connector used by various 'wedge' devices. Just because you are supposed
to /request/ some power....


George Smith

2008\07\09@075729 by William \Chops\ Westfield

face picon face

On Jul 9, 2008, at 3:57 AM, Geo wrote:

>> As far as I know, the USB spec does not allow for ANYTHING like a
>> "power only dumb device."  *ALL* that stuff violates the spec,
>
> Yes - but how many devices have we used that pinched power from  
> RTS, DTR
> etc and no complaint about violating the RS232 spec.

I'm not complaining.  I'm all in favor of creative specification  
violation.  But part of the purpose of USB to avoid the massive  
violations of specs on serial and parallel ports, so that you  
wouldn't have the awful problems inexperienced users had with serial  
(M to M, crossover, at least 7 signals, DB25 here, DB9 there????!!!!)

BillW

2008\07\10@125958 by Bob Ammerman

picon face
Okay, maybe we can interject PICs into this discussion.

How small/cheap a PIC based circuit could be cobbled together to properly
enumerate a device that asks for max power. No fair using PICs with built-in
USB support :-)

Basically, the device would have to enumerate and request 500ma power and
maybe set an output to indicate success.

Anybody think it could be shoe-horned into a PIC10?

-- Bob Ammerman
RAm Systems

2008\07\10@132429 by Dr Skip

picon face
Then we'd have to changed the subject to [PIC] ;)

Seriously though, isn't that controlled by the driver, at least for turn on and
off and value? Why not make a 'control panel' to force it in software? An
on/off switch for your lava lamp on the desktop!


Bob Ammerman wrote:
{Quote hidden}

2008\07\10@145729 by Tamas Rudnai

face picon face
> Anybody think it could be shoe-horned into a PIC10?

If you can spin up the internal clock to 12MHz at least I'll write the
firmware :-)

Tamas



On Thu, Jul 10, 2008 at 5:58 PM, Bob Ammerman <EraseMErammermanspamverizon.net> wrote:

{Quote hidden}

> -

2008\07\10@211544 by Xiaofan Chen

face picon face
On Fri, Jul 11, 2008 at 12:58 AM, Bob Ammerman <RemoveMErammermanEraseMEspamEraseMEverizon.net> wrote:
> Okay, maybe we can interject PICs into this discussion.
>
> How small/cheap a PIC based circuit could be cobbled together to properly
> enumerate a device that asks for max power. No fair using PICs with built-in
> USB support :-)
>
> Basically, the device would have to enumerate and request 500ma power and
> maybe set an output to indicate success.
>
> Anybody think it could be shoe-horned into a PIC10?

Not even a PIC16F/18F without USB or similar. There are firmware based
USB implementations on the AVR but none of them are USB compliant.
USB is more difficult than most of the people think.

Xiaofan

2008\07\10@213158 by Xiaofan Chen

face picon face
On Fri, Jul 11, 2008 at 1:23 AM, Dr Skip <RemoveMEdrskipspam_OUTspamKILLspamgmail.com> wrote:
> Then we'd have to changed the subject to [PIC] ;)
>
> Seriously though, isn't that controlled by the driver, at least for turn on and
> off and value? Why not make a 'control panel' to force it in software? An
> on/off switch for your lava lamp on the desktop!
>

This is actually quite complicated. The device needs to have dual
configuration, one with lower power (say 100mA), one with higher
power (say 500mA). And you need a host driver to set the higher
power configuration after the initial enumeration (with 100mA).
The default Windows host driver will not use the 2nd configuration
so you have to write your own driver.
Reference: https://www.usb.org/phpbb/viewtopic.php?t=13934


Xiaofan

2008\07\10@214831 by Xiaofan Chen

face picon face
On Fri, Jul 11, 2008 at 9:31 AM, Xiaofan Chen <RemoveMExiaofancTakeThisOuTspamspamgmail.com> wrote:
> On Fri, Jul 11, 2008 at 1:23 AM, Dr Skip <EraseMEdrskipspamspamspamBeGonegmail.com> wrote:
>> Then we'd have to changed the subject to [PIC] ;)
>>
>> Seriously though, isn't that controlled by the driver, at least for turn on and
>> off and value? Why not make a 'control panel' to force it in software? An
>> on/off switch for your lava lamp on the desktop!
>>
>
> This is actually quite complicated. The device needs to have dual
> configuration, one with lower power (say 100mA), one with higher
> power (say 500mA). And you need a host driver to set the higher
> power configuration after the initial enumeration (with 100mA).
> The default Windows host driver will not use the 2nd configuration
> so you have to write your own driver.
> Reference: https://www.usb.org/phpbb/viewtopic.php?t=13934
>

The best discussion thread I know is here:
https://www.usb.org/phpbb/viewtopic.php?t=13642

Quote:
"Devices with Y-cables can not comply with the spec and can not be
certified unless they work as two devices so each plug exposes a device
and enumerates before starting to draw power. And manufacturers are
known to sometime lie about being certified."

"USB extension cables (ie cables with a type A plug and Type A
recepticle) are not legal."

Xiaofan

2008\07\10@215052 by Xiaofan Chen

face picon face
On Wed, Jul 9, 2008 at 1:14 AM, Herbert Graf <RemoveMEmailinglist4KILLspamspamfarcite.net> wrote:
> On Tue, 2008-07-08 at 11:29 -0400, Olin Lathrop wrote:
>> Yup, and you won't likely find the official USB logo on the plug.
>
> Chances are you will. I find that most "non-compliant" USB hardware has
> the logo on the cable

The cable might be compliant with USB. So this is okay.

> (and often even on the box).
That is the major problem.

> It's unfortunate, but it appears that enforcement is non-existent.
I think if you stick to more known vendors, you will find that they
dare not to put USB logo on non-compliant device, such as the
2.5" USB harddisks without a power port and use Y-cable.

Xiaofan

2008\07\11@030458 by William \Chops\ Westfield

face picon face

On Jul 10, 2008, at 6:15 PM, Xiaofan Chen wrote:

> There are firmware based USB implementations on the AVR but none of  
> them are USB compliant.

In what ways are they not compliant?  (Aside from picking ID numbers  
and such out of thin air, probably.)  Not defending; just curious...

BillW


2008\07\11@043259 by Alan B. Pearce

face picon face
> Okay, maybe we can interject PICs into this discussion.
>
> How small/cheap a PIC based circuit could be cobbled together to properly
> enumerate a device that asks for max power. No fair using PICs with
> built-in
> USB support :-)
>
> Basically, the device would have to enumerate and request 500ma power and
> maybe set an output to indicate success.
>
> Anybody think it could be shoe-horned into a PIC10?

I guess the trick there is to look at the code at
http://www.cesko.host.sk/IgorPlugUSB/IgorPlug-USB%20(AVR)_eng.htm and see if
the relevant bit can be converted and shoe-horned in.

Essentially one needs three pins, two bidirectional for D+/D-, and another
output to turn on the high current load, so the 10F series has that, it just
remains to be seen if it is fast enough with a 4MHz internal clock. I
suspect not, and the clock is probably not stable enough, so I think an 8
pin 12F from the midrange line would probably be needed, with an external
crystal.

2008\07\11@050556 by William \Chops\ Westfield

face picon face

On Jul 11, 2008, at 1:32 AM, Alan B. Pearce wrote:

> I guess the trick there is to look at the code at
> http://www.cesko.host.sk/IgorPlugUSB/IgorPlug-USB%20(AVR)_eng.htm  
> and see if
> the relevant bit can be converted and shoe-horned in.
>
> Essentially one needs three pins, two bidirectional for D+/D-, and  
> another
> output to turn on the high current load, so the 10F series has  
> that, it just
> remains to be seen if it is fast enough with a 4MHz internal clock.

Last time I looked at the AVR code, they required 12MHz or 16Mh, or a  
couple odd in-between frequencies to get acceptable bit timing, and  
that was at one clock per instruction on the AVR.  I think it'd be  
pretty hopeless on a 1MIPS PIC.

BillW

2008\07\11@050637 by Xiaofan Chen

face picon face
On Fri, Jul 11, 2008 at 3:04 PM, William Chops Westfield <westfwSTOPspamspamspam_OUTmac.com> wrote:
>
> On Jul 10, 2008, at 6:15 PM, Xiaofan Chen wrote:
>
>> There are firmware based USB implementations on the AVR but none of
>> them are USB compliant.
>
> In what ways are they not compliant?  (Aside from picking ID numbers
> and such out of thin air, probably.)  Not defending; just curious...
>
http://www.obdev.at/products/avrusb/index.html
"Fully USB 1.1 compliant low-speed device, except handling of
communication errors and electrical specifications."

So will you use it in a commercial project now? I would not.
If cost is a concern, I would go for low cost USB compliant
MCU like this one: PIC18F13K50/14K50.
http://forum.microchip.com/tm.aspx?m=347894

Xiaofan

2008\07\11@051236 by Xiaofan Chen

face picon face
On Fri, Jul 11, 2008 at 4:32 PM, Alan B. Pearce <spamBeGoneA.B.PearceSTOPspamspamEraseMErl.ac.uk> wrote:

> I guess the trick there is to look at the code at
> http://www.cesko.host.sk/IgorPlugUSB/IgorPlug-USB%20(AVR)_eng.htm and see if
> the relevant bit can be converted and shoe-horned in.

This is better.
http://www.obdev.at/products/avrusb/index.html

But I said in the previous post, it is not compliant to USB specification.
"Fully USB 1.1 compliant low-speed device, except handling of communication
errors and electrical specifications."

The use of "Fully" is very deceptive here as both electrical specification
and error handling are important part of the specification.

> Essentially one needs three pins, two bidirectional for D+/D-, and another
> output to turn on the high current load, so the 10F series has that, it just
> remains to be seen if it is fast enough with a 4MHz internal clock. I
> suspect not, and the clock is probably not stable enough, so I think an 8
> pin 12F from the midrange line would probably be needed, with an external
> crystal.
>

This is for 16F84A. Again, it is not compliant to USB specification.
http://www.telefonica.net/web2/hidlcd/

It is not easy at all to use internal RC oscillator for USB application.
Microchip can not do it yet. I know FTDI and Silicon Labs have
USB product with internal RC oscillator.

Xiaofan

2008\07\11@055935 by William \Chops\ Westfield

face picon face

On Jul 11, 2008, at 2:06 AM, Xiaofan Chen wrote:

> http://www.obdev.at/products/avrusb/index.html
> "Fully USB 1.1 compliant low-speed device, except handling of
> communication errors and electrical specifications."
>
> So will you use it in a commercial project now?

It has the usual tradeoffs of open source stuff.  On the one hand, I  
might be able to fix it (try that if your USB-including micro doesn't  
work!  How many PIC/USB projects support the full communication error  
handling?  I imagine that takes some software as well as on-chip HW  
support?)  On the other hand, it'll probably never be fully compliant  
and might break something that ISN'T fixable...

(On a vaguely related issue, I recently had cause to recommend an  
FTDI chip for an app that needed a USB/Serial bridge.  Cause I *knew*  
it worked on linux and macos as well as windows, whereas other  
vendors were only doing windows + promises.  Alas, FTDI is not on the  
"approved vendor" list for single-source parts, and we went with  
someone else.  Hopefully things will work out OK.)

> If cost is a concern, I would go for low cost USB compliant
> MCU like this one: PIC18F13K50/14K50.

Doesn't seem past the vaporware stage.  Eagerly awaited.

BillW

2008\07\11@065135 by Alan B. Pearce

face picon face
>Alas, FTDI is not on the "approved vendor" list for
>single-source parts, and we went with someone else.  
>Hopefully things will work out OK.)

You mean there are multiple source USB-Serial bridge parts ???

Please do elucidate ...

2008\07\11@070151 by Xiaofan Chen

face picon face
On Fri, Jul 11, 2008 at 5:59 PM, William Chops Westfield <KILLspamwestfwspamBeGonespammac.com> wrote:
>
> On Jul 11, 2008, at 2:06 AM, Xiaofan Chen wrote:
>
>> http://www.obdev.at/products/avrusb/index.html
>> "Fully USB 1.1 compliant low-speed device, except handling of
>> communication errors and electrical specifications."
>>
>> So will you use it in a commercial project now?
>
> It has the usual tradeoffs of open source stuff.  On the one hand, I
> might be able to fix it.

This has nothing to do with open source. It is a blant USB
specification violation. It is a nice hack technically speaking.
But that is all.

I think you can not fix it since the electricalal specification is
out of your control. The communication errors handling
is also part of the SIE and may be too heavy for the
firmware implementation.


Xiaofan

2008\07\11@093827 by William Couture

face picon face
On Thu, Jul 10, 2008 at 9:48 PM, Xiaofan Chen <EraseMExiaofancspamEraseMEgmail.com> wrote:

> "USB extension cables (ie cables with a type A plug and Type A
> recepticle) are not legal."

Out of sheer curiosity, can anyone explain why a USB extension cable
is "illegal" under the USB specifications?

Thanks,
  Bill

--
Psst... Hey, you... Buddy... Want a kitten? straycatblues.petfinder.org

2008\07\11@095512 by olin piclist

face picon face
William Couture wrote:
> Out of sheer curiosity, can anyone explain why a USB extension cable
> is "illegal" under the USB specifications?

This is explained in the standard.  The guaranteed voltage levels and edge
transition times are calculated for the max length cable.  Someone could
make a longer overall cable if A-A cables were available.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\11@102356 by Wouter van Ooijen

face picon face
> Out of sheer curiosity, can anyone explain why a USB extension cable
> is "illegal" under the USB specifications?

My guess:

1. cables up to the maximum length are legal, longer are illegal. An
extension cable can create longer cables.

2. a connector is always a discontinuity. the spec has to take two such
discontinuities into account, an extension cable would add a third (and
a fourth, etc).

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\07\11@102705 by Toms hilidhe

picon face

William Couture wrote:
> Out of sheer curiosity, can anyone explain why a USB extension cable
> is "illegal" under the USB specifications?


I did a little Googling a while back and found out that the maximum
length for a USB capable is 5 metres. After that, you need a repeater.

I wanted a 10 metre lead, so I went looking for one that had an active
repeater. I found one on eBay. And I bought it.

I received the lead and it had no repeater. I tried it out and it didn't
work, the signal was too degraded.

I got back on to the seller but he was a d***head about it, he offered
me a "half refund". Ended up having to start a dispute with P**Pal, and
I'm still waiting on them to step in and make him give me a refund.

The only reason I can see for there being a limit on the length, (when
you have repeaters that is), is propagation delay.

2008\07\11@103942 by Vasile Surducan

face picon face
On 7/11/08, William Couture <@spam@bcouture@spam@spamspam_OUTgmail.com> wrote:
> On Thu, Jul 10, 2008 at 9:48 PM, Xiaofan Chen <spamBeGonexiaofancspamKILLspamgmail.com> wrote:
>
> > "USB extension cables (ie cables with a type A plug and Type A
> > recepticle) are not legal."
>
> Out of sheer curiosity, can anyone explain why a USB extension cable
> is "illegal" under the USB specifications?

If the eye pattern is on the specification range than the cable is
legal even it has 6m lenght. 6 cables of 1 meter lenght chained
together have higher chance to do not met the eye pattern level/shape
than one single cable 6 m long,

Vasile

2008\07\11@104358 by Vasile Surducan

face picon face
On 7/3/08, Bob Ammerman <.....rammermanspam_OUTspamverizon.net> wrote:
> A USB device has to negotiate with the host if it needs more than (IIRC)
> 100ma. Since the second USB connection on these drives is general
> power-only, and thus cannot negotiate, any attempt to draw more than 100ma
> is outside the USB spec.

After negociation, on most USB ports (desktop computers not laptopts
or laptopts PCMCIA to USB cards etc) you may drawn safetly 500mA and
if necessary up to 1A (out of specification but feasable).

This of course applies also to lights, fans, wet
{Quote hidden}

> -

2008\07\11@123440 by Dr Skip

picon face
Better yet, can someone explain this:

The scenario: small devices like voice recorders from Olympus, Sony, WD and
laptops from Toshiba, Dell, Compaq. USB cables between 1/2 and 1m long. Certain
combinations of original cable (came with device) with certain machines cause
the device to not be recognized at all, be seen, but as something else that
remains unidentified (the pc goes through an install again but can't tell what
the device is) or it works fine. It is pretty consistent per combination.
Coiling the cable doesn't help, neither does moving the machine somewhere else
(in case of EMI).

However, adding a short, cheap, dollar store extension cable makes it work fine
and doesn't seem to affect the previously working combinations. A different
length cable also works, so it is a length change that matters, not any magic
to the extension. It's just convenient. In the case of the voice recorders,
it's at USB 1.1 speeds, so terminations shouldn't be an issue, or at least it
should be well understood... The devices are very low power and some are
self-powered. Only a foot+ of cable makes the difference.

This happens in about 1 in 10 combinations.


William Couture wrote:
>
> Out of sheer curiosity, can anyone explain why a USB extension cable
> is "illegal" under the USB specifications?
>
> Thanks,
>    Bill
>

2008\07\11@172806 by William \Chops\ Westfield

face picon face

On Jul 11, 2008, at 3:51 AM, Alan B. Pearce wrote:

>> FTDI is not on the "approved vendor" list for
>> single-source parts, and we went with someone else.
>> Hopefully things will work out OK.)
>
> You mean there are multiple source USB-Serial bridge parts ???

No, I mean that we have other vendors of single-sourced
parts that have already gone through some "qualification"
process to "allow" us to buy parts from them, and they're
much easier to specify as the source for additional single
sourced parts.  A sort of "well, if TI goes belly up, we're
so screwed that another chip won't be any more of a problem"
attitude, I guess (My being, as a SW engineer, FAR from the
"manufacturing" end of the hardware business!)

BillW

2008\07\11@184352 by William \Chops\ Westfield

face picon face

On Jul 11, 2008, at 6:38 AM, William Couture wrote:

> Out of sheer curiosity, can anyone explain why a USB extension cable
> is "illegal" under the USB specifications?

Because cable lengths greater than the max defined in the standard  
run the risk of not operating correctly due to timing issues, AND the  
connector probably represents an impedence discontinuity that is also  
not supposed to be there.

Does anybody (Xiaofan?) know whether the immensely popular thumb-
drive like things (including flash drives, wireless transmitters of  
various sorts, security dongles, microcontroller evaluation modules,  
etc, etc) (ie USB devices with no cable at all) are allowed by the  
USB spec?  I don't seem to recall the initial hype including devices  
without ANY cable, thought there is a special case for devices with a  
built-in cable (mice in particular); that's why you won't find a  
(legal) USB A/B cable as thin and flexible as a mouse cable...

The "tightness" of the USB specs are directly aimed at the horrors  
caused by rs232 in previous generations (and parallel ports to a  
lesser extent.)  It's all very amusing to watch the way in which  
companies violate the specs, not to mention the ways in which  
customers demand that the specs be violated.

BillW

2008\07\11@201645 by olin piclist

face picon face
William Chops" Westfield" wrote:
> Does anybody (Xiaofan?) know whether the immensely popular thumb-
> drive like things (including flash drives, wireless transmitters of
> various sorts, security dongles, microcontroller evaluation modules,
> etc, etc) (ie USB devices with no cable at all) are allowed by the
> USB spec?

Of course they are allowed.  What rule do you imagine they migh violate?
The spec gives a maximum cable length, but there is no minimum.  Why would
there be?


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\07\11@202548 by Xiaofan Chen

face picon face
On Sat, Jul 12, 2008 at 6:43 AM, William Chops Westfield <TakeThisOuTwestfw.....spamTakeThisOuTmac.com> wrote:
> Does anybody (Xiaofan?) know whether the immensely popular thumb-
> drive like things (including flash drives, wireless transmitters of
> various sorts, security dongles, microcontroller evaluation modules,
> etc, etc) (ie USB devices with no cable at all) are allowed by the
> USB spec?

It is allowed.

>I don't seem to recall the initial hype including devices
> without ANY cable, thought there is a special case for devices with a
> built-in cable (mice in particular); that's why you won't find a
> (legal) USB A/B cable as thin and flexible as a mouse cable...

It is because that mouse is low speed device. The low speed USB cable
is of different specification so that thin and flexible cable is
possible.

Regards,
Xiaofan

2008\07\11@204203 by Marcel Birthelmer

flavicon
face
On Fri, Jul 11, 2008 at 5:16 PM, Olin Lathrop <TakeThisOuTolin_piclistKILLspamspamspamembedinc.com>
wrote:

> William Chops" Westfield" wrote:
> > Does anybody (Xiaofan?) know whether the immensely popular thumb-
> > drive like things (including flash drives, wireless transmitters of
> > various sorts, security dongles, microcontroller evaluation modules,
> > etc, etc) (ie USB devices with no cable at all) are allowed by the
> > USB spec?
>
> Of course they are allowed.  What rule do you imagine they migh violate?
> The spec gives a maximum cable length, but there is no minimum.  Why would
> there be?
>

The Ethernet physical layer has some conditions for a minimal cable length,
so it's conceivable that USB might have a similar limitation (though it
doesn't).
- Marcel

2008\07\11@204609 by William \Chops\ Westfield

face picon face

On Jul 11, 2008, at 5:16 PM, Olin Lathrop wrote:

> Of course they are allowed.  What rule do you imagine they migh  
> violate?
> The spec gives a maximum cable length, but there is no minimum.  
> Why would
> there be?

Well, for example, there are (maybe?) physical specs for the  
dimensions and appearance  of the external shaped strain-relief part  
of the USB connector, right?  I myself have run into the annoyance of  
being unable to plug a flash drive into a USB slot on a computer,  
because the slot next to it was occupied and didn't leave room for  
the THICKNESS of the drive.

BillW

2008\07\11@205903 by Carl Denk

flavicon
face
I had a similar situation with a Trendnet 811G and bluetooth adapter.
was able to buy an extension piece, male/ female, adds an inch.

William "Chops" Westfield wrote:
{Quote hidden}

2008\07\12@083524 by Lee Jones

flavicon
face
>> Of course they are allowed.  What rule do you imagine they migh  
>> violate?  spec gives a maximum length, but there is no minimum.

> Well, for example, there are (maybe?) physical specs for the  
> dimensions and appearance  of the external shaped strain-relief part  
> of the USB connector, right?  I myself have run into the annoyance of  
> being unable to plug a flash drive into a USB slot on a computer,  
> because the slot next to it was occupied and didn't leave room for  
> the THICKNESS of the drive.

I've run into that same situation.  The USB memory dongle in
question (from a name brand) came with a male-A/female-A cable
to be used in that case.  Raises interesting questions of "is
it an extension if it's the only cable involved" - eh? :-)

                                               Lee Jones

2008\07\14@040344 by Alan B. Pearce

face picon face
>The Ethernet physical layer has some conditions for a minimal cable
>length, so it's conceivable that USB might have a similar limitation
>(though it doesn't).

AIUI Ethernet also has conditions about message collisions which require a
minimum cable length to work properly. The same collision conditions are
involved in the maximum length.

2008\07\15@070915 by Xiaofan Chen

face picon face
On Fri, Jul 11, 2008 at 9:38 PM, William Couture <.....bcouturespamRemoveMEgmail.com> wrote:
> On Thu, Jul 10, 2008 at 9:48 PM, Xiaofan Chen <RemoveMExiaofancspamspamBeGonegmail.com> wrote:
>
>> "USB extension cables (ie cables with a type A plug and Type A
>> recepticle) are not legal."
>
> Out of sheer curiosity, can anyone explain why a USB extension cable
> is "illegal" under the USB specifications?

Because the standard says so. ;-)

*******************************************
Section 6.4.4 of USB 2.0 Specification.

6.4.4 Prohibited Cable Assemblies
USB is optimized for ease of use. The expectation is that if the
device can be plugged in,
it will work. By specification, the only conditions that prevent a USB
device from being
successfully utilized are lack of power, lack of bandwidth, and
excessive topology depth.
These conditions are well understood by the system software.

Prohibited cable assemblies may work in some situations, but they cannot
be guaranteed to work in instances.

• Extension cable assembly
A cable assembly that provides a Series "A" plug with a series "A"
receptacle or a
Series "B" with a Series "B" receptacle. This allows multiple cable
segments to be
connected together, possibly exceeding the maximum permissible cable length.

• Cable assembly that violates USB topology rules
A cable assembly with both ends terminated in either Series "A" plugs or Series
"B" receptacles. This allows two downstream ports to be directly connected.
Note: This prohibition does not prevent using a USB device to provide a bridge
between two buses.

• Standard detachable cables for low-speed devices
Low-speed devices are prohibited from using standard detachable
cables. A standard
detachable cable assembly must be high-/full-speed. Since a standard
detachable cable
assembly is high-/fullspeed rated, using a long high-/full-speed cable
exceeds the
capacitive load of low-speed.

*******************************************

And here is a more detailed explanation from Intel.
http://www.usb.org/developers/whitepapers/cablew~1.pdf

Regards,
Xiaofan

2008\07\15@071237 by Xiaofan Chen

face picon face
On Mon, Jul 14, 2008 at 3:56 PM, Alan B. Pearce <spamBeGoneA.B.Pearce@spam@spamspam_OUTrl.ac.uk> wrote:
>>The Ethernet physical layer has some conditions for a minimal cable
>>length, so it's conceivable that USB might have a similar limitation
>>(though it doesn't).
>
> AIUI Ethernet also has conditions about message collisions which require a
> minimum cable length to work properly. The same collision conditions are
> involved in the maximum length.

I tried Google and it seems that the standard does not say this. But there
are non-IEEE source mentioning the limit.

http://www.duxcw.com/faq/network/cablng.htm
http://www.ctrlink.com/2006/07/no-minimum-length-for-cat-5.html


Xiaofan


'[EE] USB power in parallel'
2008\08\17@104852 by Dario Greggio
face picon face
Xiaofan Chen wrote:

>>>http://www.obdev.at/products/avrusb/index.html
> This has nothing to do with open source. It is a blant USB
> specification violation. It is a nice hack technically speaking.
> But that is all.
>
> I think you can not fix it since the electricalal specification is
> out of your control. The communication errors handling
> is also part of the SIE and may be too heavy for the
> firmware implementation.

I played with it and absolutely agree with Xiaofan: it works, it's nice,
but it's out of spec.
20MHz PIC16F628A

--
Ciao, Dario

More... (looser matching)
- Last day of these posts
- In 2008 , 2009 only
- Today
- New search...