USB GPIB interface is not usable if connected at boot time

[ScottieB]

Halden,

We have not seen this behavior on any systems here at NI.  Can you detail more about what you're seeing in the lights on the unit?  There are 2 lights, the READY and the ACTIVE lights.  Which ones are on when it is working?  What color are they?  And when it is not working?

What driver do you have loaded?  You should upgrade to NI-488.2 v2.4 if you haven't done so already.

Scott B.

GPIB Software

[Halden]

Hi Scott and Michael

Thanks for your attention to this issue.
NI 488.2 is version 2.4
Motherboard is ECS RS400-A
When the unit is working the orange "ready" light is on.  When it is not working, the green "active"
light is on.  When it is moving data back and forth, the orange light is on and the green light goes
on and off.  When I unplug and replug the cable into the USB jack (I have tried all 4 rear panel USB
jacks with the same results) the Device Driver list updates to include the interface.  I do not get
messages about the system wanting to automatically install software. 

Halden

[ScottieB]

Sorry, I need more information on the lights.  I'll explain what I think you mean.  When it is working you see this LED state:

READY=Orange

Active=Off

(the active flashes when there's activity, of course)

and when it is not working you have this:

READY=Off

Active=Green

(the active doesn't flash, it's just solid)

Is that correct?

Scott B.

GPIB Software

[Halden]

Scott,
Yes, that's correct.
Halden

[Halden]

Hi Scott and Michael

More information on this problem:

If we leave the USB-GPIB plugged into the front USB port on the computer, then
every time it boots up, the "found new hardware" wizard appears and we have
to click through it to get the unit working.  We would rather not have to do this
every time we turn on the measurement system.  We also don't want to use
the front port.  This is different behaviour
from the back ports where the USB-GPIB is ignored if plugged in at turn-on.

We checked for available bios updates to our motherboard.  There are some,
but they do not address any issue relating to USB.

We updated USB drivers from ECS (the motherboard manufacturer) and
the updates had no effect on the USB behaviour of the system.

Halden

[Albert.Geven]

This looks like something I found out years ago (>20) when a power supply came up very slowly and the hardware did not recognize the powerup condition !
Not all in this world is software, but a lot is.

[fivestar]

Hi Halden-

Do you ever get a "malfunctioning device" or "device not recognized" balloon message from the tray area near the clock?  When the device is in the bad state, try opening Device Manager and scan for hardware changes--that could cause the balloon message to appear if there is a problem.

I notice you are also using the motherboard-supplied USB host controller drivers.  Have you tried to use the drivers supplied by Microsoft with XP?  Those are generally better quality.

Also, are you using XP SP2?  A lot of high-speed USB problems were fixed with SP2, although in my experience most of them were related to multiprocessor systems.

Dan

GPIB Software

[Halden]

Hello Dan and Albert

Thanks for your thoughts.  I might try to swap for another motherboard or another power supply.
How does the fact that it works properly when restarting but doesn't work properly when turning
the computer off then on again relate to the power supply slow risetime theory?
OS is Windows XP Pro SP2.
I tried the GPIB-USB-HS on my notebook computer and it works properly.
I do not get any "device not found" or "unknown device" messages or icons.
Cheers

Halden

[Halden]

Hi Dan

I never get a "malfunctioning device" or "device not recognized" balloon message from the tray area near the clock.
I tried "scan for hardware changes" and this did not affect the list of devices nor cause a message to appear.  In
addition, I checked for icons near the clock and none of them relate to this.

Regarding the idea about the power supply raising its voltage too slowly, does the GPIB USB HS send a message
to the motherboard announcing its existence when it starts to receive power?  Or is the motherboard supposed
to ask each of its ports if anybody's there whenever it turns on?

Halden

[Albert.Geven]

Reset does not switch off the powersupply but issues a reset to devices.

Power on should start the usb communication from the device as far as I know.
Can someone from NI  test this out with a slow power up to the usb gpib device, to be sure.