[arm-allstar] One of two USB URIs dropping off
Randy Neals
randy at neals.ca
Sat Jan 12 00:40:34 EST 2019
I agree. The suppliers on Amazon for those potted dc dc converters change
like the weather.
The 3B+ uses an entirely new DC regulator module.
A custom design for the Raspberry Pi.
I’m thinking of using POE Hats and powering my Pi’s from an Ethernet
switch. If the switch is “Managed”, it also provides a way of turning off
the port and rebooting the Pi.
The second gen Pi POE Hat would seem like an elegant solution.
I may also revert to a Pi per repeater.
I dislike having multiple Pi’s to admin and multiple firewall port
mappings, but I like reliable nodes better :)
https://www.raspberrypi.org/blog/poe-hat-revision/
On Fri, Jan 11, 2019 at 10:11 PM David McGough <kb4fxc at inttek.net> wrote:
>
>
> Hi Randy,
>
> Okay, yes, looking at the dmesg output, this issue looks power related,
> one way or another. Note below that just after the "under voltage"
> warning, and before the "voltage normalized," the URI(s) aren't happy.
>
> The micro-USB power connection on the RPi3b+ may be the problem. Another
> factor that we've noticed recently is that some brands of the epoxy potted
> 5V3A DC-DC converters don't perform as well as others. Frustrating.
>
> A subtle ground loop, even with the hardware in same chassis, might be the
> culprit, too.
>
> 73, David KB4FXC
>
>
>
> On Fri, 11 Jan 2019, Randy Neals wrote:
>
> > Hi David,
> > Thanks for the email...
> >
> > Usually just one URI stops. It seems to be the first port, my VHF
> repeater.
> > Earlier it seemed like it was alternating with no pattern between
> repeaters.
> > (Thus I wasn't suspecting a bad URI)
> >
> > No reboot of the system, no hang. Other repeater usually continues to
> > operate.
> >
> > The repeaters are ICOM FR-5000/FR-6000
> > Both repeater RF modules are in the same metal Icom rack chassis, ie:
> same
> > ground for both repeaters.
> > The Audio and COR/PTT come out on DB25 connector on back of repeater.
> > 1' cable with signal/audio to URI
> >
> > The repeaters are powered from 12VDC power distribution.
> > 12V power is from a large communications power supply (Switch mode N+1
> > redundant) with battery backup.
> >
> > The 5V Pi power is a DC to DC converter rated at 3 Amps. 12V from the
> same
> > distribution as repeater. 5V on micro USB to Pi.
> > At first I had a different PS, and swapped early thinking that was the
> > problem.
> >
> > The Pi is in an aluminum chassis with a tiny fan.
> > It is on a rack shelf above the repeater. The Pi chassis is not grounded,
> > but it has 2 USB cable, and the power supply cable to the DC DC converter
> > which may have ground.
> >
> > Here's dmesg
> > I see several "under voltage" messages. That doesn't seem good.
> >
> > I rebooted just a few minutes ago. When the URI stops again, I'll run
> dmesg
> > again.
> >
> >
> > [ 7.215868] sd 0:0:0:0: [sda] Assuming drive cache: write through
> >
> > [ 7.222428] sda: sda1
> >
> > [ 7.227933] sd 0:0:0:0: [sda] Attached SCSI removable disk
> >
> > [ 8.300598] smsc95xx 1-1.1:1.0 eth0: link up, 100Mbps, full-duplex,
> lpa
> > 0xD1E1
> >
> > [ 13.608033] *Under-voltage detected! (0x00050005)*
> >
> > [ 17.229524] retire_capture_urb*: 119 callbacks suppressed*
> >
> > [ 21.938034] Voltage normalised (0x00000000)
> >
> > [ 26.348512] retire_capture_urb*: 18 callbacks suppressed*
> >
> <snip>
>
> --
Sent from mobile.
More information about the ARM-allstar
mailing list