USB disconnect issue

[R. Meijn]

problem description
USB connected devices disconnects unexpectedly at regular times, ranging
from minutes to hours, without user initiative; sometimes printer is
spontaneously deleted by the system

configuration
Operating system: WinXP, SP2+SP3.
USB connected devices: scanner, Bluetooth dongle, printer, WiFi client (all
USB directly from motherboard) and Card Reader (front USB port)
Recently both printer and scanner (USB1.1) were replaced by new devices
(USB2.0): HP K5400 OfficeJet and Canon CanoScan 5600F.

observations
After installation of the printer, including lastest software, the printer
would unexpectedly disconnect and reconnect, Windows generated at these
instances the audible USB 'on'/'off' sounds.
Note the printer functions properly and even two-way communications are
maintained after disconnect/connect.
Selecting another USB port did not cure the problem, neither did replacing
the USB cable by an USB2.0 hi-speed certified one.
Device Manager shows that all hardware is connected and reports that is
'functions properly'.
Disabling USB power management to turn the device off did also make no
difference.
The printer is shared in a network with another computer from which can be
printed without problems.

The printer self diagnostics logged a firmware error: 69666.
For this reason the was replaced by a new one; however, also this printer
exhibited the same firmware error.

Focussing on the printer the software was completely removed (using a HP
scrubber, including registry entries) and reinstalled succesfully; this did
not help.
Following instructions from HP Support Desk the printer drivers were removed
and manually installed; this also did no cure the problem.
During installation virusscanner, firewall and anti-bot software were
disabled.

Installing a freeware application USB Deview (www.nirsoft.net) allows to
monitor USB port connected devices (acts as USB watchdog).
This revealed that not only the printer, but also the Card Reader
disconnected, sometimes simulataneously.

The printer can switched on and off normally.
Upon computer reboot, the printer sometimes finds itself as new hardware and
installs succesfully.

The problem aggravated when the printer spontaneously was deleted without
any initiative; Event Viewer logged Event ID 4:
Event Type: Warning
Event Source: Print
Event Category: None
Event ID: 4
Date: 27-12-2008
Time: 23:31:52
User: NT AUTHORITY\SYSTEM
Computer: MASTER
Description: Printer HP Officejet Pro K5400 Series is pending deletion.
And next Event ID 3:
Event Type: Warning
Event Source: Print
Event Category: None
Event ID: 3
Date: 27-12-2008
Time: 23:31:57
User: NT AUTHORITY\SYSTEM
Computer: MASTER
Description: Printer HP Officejet Pro K5400 Series was deleted.
Next the printer could easily be reinstalled using the add printer wizard,
upon which Event Viewer logs Event ID 20:
Event Type: Warning
Event Source: Print
Event Category: None
Event ID: 20
Date: 28-12-2008
Time: 00:07:11
User: NT AUTHORITY\SYSTEM
Computer: MASTER
Description: Printer Driver HP Officejet Pro K5400 Series for Windows NT x86
Version-3 was added or updated. Files:- UNIDRV.DLL, UNIDRVUI.DLL,
hpwk5403.GPD, UNIDRV.HLP, hpwhk540.cfg, hpwk540a.ini, hpzst589.dll,
hpz3c589.dll, hpzur589.dll, hpwk5403.xml, hpzsc589.dtd, hpzui589.dll,
hpz3r589.dll, hpzpr589.dll, hpcdmc32.dll, hpbcfgre.dll, hpzsm589.gpd,
hpz3m589.gpd, hpzev589.dll, hpzhl589.cab, STDNAMES.GPD, hpfie589.dll,
hpfig589.dll, hpfrs589.dll, UNIRES.DLL.
After installation Event Viewer logs Event ID 1015 several times:
Event Type: Warning
Event Source: MsiInstaller
Event Category: None
Event ID: 1015
Date: 28-12-2008
Time: 00:21:18
User: MASTER\René
Computer: MASTER
Description: Failed to connect to server. Error: 0x800401F0
The printer works fine except for status information; HP software finds it
disconnected whilst with the same software two-way communications can be
established.
After the Card Reader was unplugged the printer keeps (dis-)connecting, but
is no longer spontaneously deleted.

It is noted that when the scanner is on (usually off) the USB connection is
also regularly lost.
The Bluetooth dongle and WiFi Client, being on adjacent port with scanner
and printer never disconnect.

analysis
As not only the printer, but also Card Reader, suffers from USB connection
problems, the USB ports of the computer may be suspect; however, peripherals
on adjacent ports are not disconnected and Device Manager finds all hardware
functioning properly.
The printer may be suspect as it still reflects a firmware error; it is not
clear whether this would impair other hardware.

question
By elimination it would make sense to exclude potential problem causes one
by one.
First of all, I would like to verify USB port functionality.

Any comment, suggestion or advice for a diagnostics tool is more than welcome!

--
Kind regards,

René Meijn
Hoeven
The Netherlands

 

[Woody]

You are probably having driver conflicts. That happens when you get too many
USB devices on at the same time. You need to isolate which devices conflict
with each other and then only power or plug one of the devices at a time.
Check for system dumps at the time of the failures and analyze them as they
can show which drivers are failing. You may have to clean up old drivers
there left behind when older devices are removed or you may have to rebuild
your system from scratch to eliminate them.

 

[R. Meijn]

Woody,

first of all thanks for reading my lengthy story.
And thanks for your suggestion; I did not think of it, but the application
USBDeview I mentioned, indeed offers the possibility to deinstall items which
were formerly connected to the PC by USB and left entries behind.
I will uninstall all old entries and herewith deleting old drivers;
following your rationale this will exclude any conflict of drivers.
Let you know of any (positive) result.

--
Kind regards,

René Meijn
Hoeven
The Netherlands

 

[R. Meijn]

LS,

in addition to my earlier mails and besides the driver issue, I am wondering
whether I am overdrawing the current limitations on USB.
At the same time I feel the equipment which is connect via USB onto my PC is
certainly not excessive:
USB Dongle (Sitecom); no interrupts, no problems
WiFi Client (SpeedTouch); no interrupts, no problems
CardReader (Sitecom); spontaneous interrupts, disconnected
Canon CanoScan 5600F scanner; spurious interrupts, normally off
Hewlett Packard K5600n printer; regular interrups, spontaneous
uninstalles/installs, communication problems et cetera
No USB hubs are used.
So am I overstressing the performance of my ASUS P4PE motherboard concerning
USB capabilities?
Please comment.

--
Kind regards,

René Meijn
Hoeven
The Netherlands

 

[Paul]

LS,

in addition to my earlier mails and besides the driver issue, I am wondering
whether I am overdrawing the current limitations on USB.
At the same time I feel the equipment which is connect via USB onto my PC is
certainly not excessive:
USB Dongle (Sitecom); no interrupts, no problems
WiFi Client (SpeedTouch); no interrupts, no problems
CardReader (Sitecom); spontaneous interrupts, disconnected
Canon CanoScan 5600F scanner; spurious interrupts, normally off
Hewlett Packard K5600n printer; regular interrups, spontaneous
uninstalles/installs, communication problems et cetera
No USB hubs are used.
So am I overstressing the performance of my ASUS P4PE motherboard concerning
USB capabilities?
Please comment.

There are two hardware limits on your board.

If the USB was powered from +5VSB, and the +5VSB was overloaded
by the computer, the power supply would switch off. That isn't
happening, so that is not the issue.

Asus uses Polyfuse technology to protect the USB ports. The
Polyfuse is a polycrystalline device, which melts internally
when overcurrent occurs. The state change makes the fuse an
open circuit, and then the USB device loses power. It might
take 30 seconds to a minute for the Polyfuse to cool enough,
such that it started to conduct again. If your USB interruptions
occur faster than that, it isn't likely to be a Polyfuse problem.

Polyfuses are self-repairing and need no maintenance.

A test you can try, is to evaluate USB bus behavior in an
alternate OS such as Linux. Knoppix, from knopper.net, is
an example of a LiveCD, which doesn't need to install anything
on the hard drive in order to be used. You download the
700MB CD ISO9660 image and burn a CD with it using Nero.
You boot the CD. When the desktop appears, open a terminal
and type "dmesg". That displays detected hardware. When I
used a webcam, its detection was added to the end of the
"dmesg" output. If your hardware continued to have
spurious interruptions and re-detection, you should
see it adding entries to the end of dmesg.

If nothing funny happens in Linux, but does in Windows,
then it is some kind of software problem.

Cleaning out the USB stack in Windows, and letting
Windows rediscover all the hardware, is a thing you can
try. One method of doing that (which I haven't tested),
is renewusb from this site. Note the "use entirely at your
own risk" warning. Prepare a restore point and a backup,
just in case.

http://www.robvanderwoude.com/devcon.html

An alternative, is the recipe here, which uses Device Manager in
Safe Mode.

http://www.usbman.com/Guides/Cleanup Device Manager Safe Mode.htm

Paul

 

[w_tom]

Asus uses Polyfuse technology to protect the USB ports. ...
It might take 30 seconds to a minute for the Polyfuse to cool
enough, such that it started to conduct again. If your USB
interruptions occur faster than that, it isn't likely to be a
Polyfuse problem.

I was using Polyfuse in my electronics designs even 20 years ago.
Polyfuse will cool sufficiently in less than 5 seconds. Problem with
an improperly selected Polyfuse is one that is too small; that
increases voltage drop. If a Polyfuse current limit is too close to
0.5 amps, then some systems may have insufficient voltage on a USB
device that is not so tolerant.

Meanwhile, that is not his problem since his USB devices are not
(should not be according to his posts elsewhere) drawing power from
the USB port. But then procedures suggested in another thread also
started by R Meijn might identified that or other anomalies.

As Woody noted here and as also noted in that other thread,
sometimes a USB device does not play nicely which is why plug fests
were so often sponsored by the USB consortium. Is he or has he used
USB devices that did not participate?

Also that other thread, AC power leakage is also a suspect which is
why AC power sources must be considered.

Also assumed - the OP is only using USB approved cables on High
Speed USB devices which mean no USB extension cords.