Problems with external 5V supply

Mon Apr 2 13:39:05 UTC 2012

Hi Håkan,

Have you used a high speed device in you setup? What is the result if 
you do that?

/Klinger

On 2012-04-02 12:39, Jacques Bride wrote:
> Thanks for your answer
>
> so it would mean that simply the SnowBall board does not disable 
> correctly the current it has to provide to the usb OTG when connected 
> to powered hub?
>
> Jacques
>
> 2012/4/1 Håkan Berg <hakan.pm.berg at gmail.com 
> <mailto:hakan.pm.berg at gmail.com>>
>
>     I have done som testing on the hub and OTG hardware for a few reasons.
>
>     1) Somebody told me there is a hardware issue with the OTG port.
>     2) I clearly saw differences between hubs.
>     3) I had built an aluminium casing with a 25 euro built-in hub
>     that did not work with the OTG.
>
>     Especially the last reason made me persue this issue, since I have
>     a cheap hub that works perfectly well, but does not fit the casing.
>
>     The following enhancements did not work:
>     1) There was no voltage regulator for the internal 3.3 volts in
>     the hub. There was only two diodes roughly lowering the voltage. I
>     removed them and mounted a 3.3 volt regulator.
>     2) There was a switch circuitry that should switch the internal
>     power usage between draining from the external power supply when
>     available, else from the host USB power. I removed that and
>     connnected to the external power supply only.
>     3) I supplied the host USB power direct from the external power
>     supply.
>
>     After that I was almost out of ideas. One last idea popped up: I
>     simply cut the red wire, i.e. I opened the circuit on the pcb
>     where the USB +5V (red lead) was going to the OTG port. Then it
>     would at least hinder the hub from consuming any power at all from
>     the OTG port.
>
>     And that works fine. 100%! The keyboard, mouse or whatever leagal
>     device I connect to the hub are always detecte, be it they are
>     connected at boot-up or afterwards.
>
>     Now, this makes the HUB less useful for other purpouses, since it
>     only works with a power supply, but for me that is fine. I have my
>     cool mini-computer now running Linux Ubuntu 10.10 just like my
>     large desk-top.
>
>     If you wish to try this, do it at your own risk. Just cut the 5
>     volt lead on the host connection inside the HUB. It's simple and
>     could be reversed using a soldering iron.
>
>     /Håkan
>
>     Den 30 mars 2012 13:40 skrev Richard CURNOW
>     <richard.curnow at stericsson.com
>     <mailto:richard.curnow at stericsson.com>>:
>
>         Hello,
>
>         I'm having problems supplying a SDK board (V11) through the
>         barrel jack.
>         I've tried not only a couple of "wall-wart" type supplies
>         (both rated 5V
>         5A), but also a bench supply in the lab.  With the lab supply,
>         we also
>         tried varying the voltage in roughly 0.5V steps from 3 up to
>         just over
>         5.  The symptoms were the same regardless.
>
>         What I'm seeing, for all these 5V supplies, is:
>
>         * the LED between the barrel jack and the OTG connector is lit
>         * neither the green "User LED" nor the blue LED on the back
>         panel ever
>          become lit
>         * a voltage around 4.3-4.4 is present on the exposed metal of
>         the jumper
>          by the words "EXT PS"
>         * pin 1 of the JTAG header, which should be 1.8V, is reading 0V
>         * the machine at the other end of the console cable _is_
>         detecting the
>          USB device - in other words a /dev/ttyUSB0 is created by udev
>         – so the
>          FT232RL and the other USB connection are powered
>         * however there is no text displayed in minicom – no boot
>         messages, at
>          all
>
>         In contrast, if I power the board through the OTG connector
>
>         * all 3 of the LEDs become lit
>         * 1.8V is visible on pin 1 of the JTAG header
>         * the boot messages are visible on the console connection and
>         a working
>          root shell is reached through minicom
>
>         We're stuck for what to try next.  Has anybody ever seen a similar
>         problem and fixed it?  Or is the board faulty?
>
>         The other point I've noted is that the reset push-button
>         doesn't result
>         in a successful reset.  The root shell becomes frozen, but
>         nothing else
>         happens.  Neither does a "reboot" command typed into the shell do
>         anything.  Is it possible there is a problem around the reset
>         handling –
>         which is avoided with OTG power because of the logic that
>         limits the
>         initial in-rush current?  Maybe the AB8500 is not starting up
>         (hence the
>         lack of 1.8V output to the JTAG header) ?
>
>         Any ideas would be much appreciated.
>
>         Richard
>
>         _______________________________________________
>         users mailing list
>         users at igloocommunity.org <mailto:users at igloocommunity.org>
>         http://igloocommunity.org/cgi-bin/mailman/listinfo/users
>
>
>
>     _______________________________________________
>     users mailing list
>     users at igloocommunity.org <mailto:users at igloocommunity.org>
>     http://igloocommunity.org/cgi-bin/mailman/listinfo/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://igloocommunity.org/pipermail/users/attachments/20120402/72e20dc0/attachment-0001.html>