Now I've got ActiveSync problems...

All Universal Serial Bus devices are assigned an identification sequence code to identify themselves on the USB bus, on the device manifest and to the device driver layer all while the usb device steps through power-on, idle/wait, allocate address. The entire sequence needs to match in order for Windows to properly initialise the device. Once done the device changes state to configured; this doesn't properly take place.

When ActiveSync fails, usually it has missed one of the steps involved and cannot properly reinitialise the device. At which point as I frequently advocate you need to change port - new device instance & bus ID equates to a new identification string being generated - or clear the enumeration from the registry - which is not easy.

As the entire process is done on the Windows driver database... the above statement is... well, unavailing.

Ahhh, it begins to make sense. So my desktop USB port is assigned a number to identify it, and I imagine that my 4 port hub, which is plugged into that is assigned another 4 numbers, each unique for each port.

So if I would have unplugged my sync cable and plugged it into a different USB port, then I probably would not have had the installation problem to begin with...is that correct?

Rich

Edited by Rich Hawley 2006-07-08 12:45 PM

Sort of, it's not quite as simple as being assigned a number. There are a lot of checks in the process. For example, ever wondered why the lights go on and then off and then back on on your hub? The reason is because as part of the enumeration process the system raises the hardware data lines, asks the device what speed it is (and a few other things) then tells it to reboot. The system then mounts it on reboot using the correct driver. This happens on all USB devices, but you don't see it.

Using a different USB port will not universally fix such problems, no way! Don't let me give you that impression. The reasons why enumeration fails are very broad. It can be down to voltage, device drivers, bad connections, lack of privilaves, the system is too slow or doing too much at the time the device is connected. Enumeration failures are unusual in most hardware in my own personal experience, WinCE seems to be a special case because the IrCOMM signal gives up and forcibly terminates the connection if it doesn't instantly (near enough) hear what it wants from the PC end.
If you have ever seen the USB connection jump through all 10 retries in a second or so, that's the reason why - temper tantrum because it's being ignored.

The reason why I say use a different one is because it forces the entire process to happen again. Using the same port uses the existing registration, and if that went wrong, you'll never get anywhere except for those darn 678 errors.