To all

I had an interesting interaction with my computer system, its USB controllers, and the function of the BIOS that I thought I would share.  There are so many things going on in the background that we are not aware of--these things effect the functionality of our systems--but, we often have no idea what's going on--where or why.....

So, this may not apply to very many out there--it's an older motherboard--the system was built about 2002 or 2003.  It has 2 USB 2.0 controller chips--one is the Intel built into the north/south bridge chip set that creates 6 USB 2.0 ports.  And there's an added chip installed on the motherboard by NEC--that creates an additional 4 USB 2.0 ports.

Two of the Intel ports are *hard wired* on the back of the I/O panel of the motherboard--the rest are accessed by attaching connecting wires from peripheral ports to header connectors on the motherboard--each header will allow for 2 USB ports--so one header goes to the 2 case USB ports on the front panel.  And, the motherboard came with two rear slot adaptors that go in empty add-on slot openings on the back.  So, I have plugged the wires from those slot adaptors to the 3 other available headers to create an additional 6 USB ports on the back.

So, enter the first *issue*--the documentation for the motherboard does not identify which controller chip is connected to which header--so it's impossible to tell which USB port is going to which controller.  Why is this important?  Well, at least on this older system, if you want to play around with booting your system from a USB device--flash drive or HDD--you can only do that from the USB controller that is integrated into the north/south chip set--i.e. the Intel USB controller.  The NEC controller is ignored as a possible boot device source.

And, an additional problem--if you hook up more than one USB device to the north/south bridge chip-set that could possibly be a boot source--again on this older system--the boot device that is hooked up to the first seen port will be considered the boot device--and any other device will be ignored--and/or you will possibly get a system that *freezes* when attempting to boot from a USB device--the system's BIOS doesn't like to manage more than one USB boot device!

So, what does the above paragraph mean?   If you have two flash drives hooked up to the Intel north/south bridge chip set--that may create a problem (I sometimes keep a flash drive hooked up that really isn't meant to be a boot device--just a storage device to transfer files between computers--if I then add an additional flash drive specifically as a boot device--things might not go well if the two devices are in the wrong order as to which port is being used for each).

Also, if I have my USB flashcard reader hooked up--with a memory card in the reader--if it is on a port seen before the flash drive that I am attempting to boot from--there often is a *freeze* of the system.

So the bottom line--which port is seen first for the north/south bridge chip set by Intel makes a difference as to success or failure to boot from a USB device.  And, again, just like there is no documentation as to which header is to which controller--there is no documentation as to what the port sequence is, even if you figure out which header goes to which USB controller.

So, enter the next piece of hardware--I have a USB 2.0 add-on card, a Adaptec DuoConnect card.  It has 3 USB ports on the card's slot side, and two fire wire outlets (that's the *Duo* part of the card).  I had been using that card on another older system that only came with USB 1.1 controllers--so to have USB 2.0, I installed this card on that system.

Adaptec created a nice USB mapping software program called *USBControl* that will show the USB connections with a nice visual display.  The downside, the software will only work if you have a very limited set of Adaptec USB hardware attached to the system.  The above USB controller and/or several different USB hub devices.  I also invested in one of the Adaptec USB hub devices.  I have found that the older versions of the USBControl will detect and work if the hub is connected without the DuoConnect add-on card present.  But the newer versions of the USBControl program seem to work only if the add-on card is installed--and will not work if just the hub is connected.  I have not figured out why that is so--at least not yet.

So, this is the setup *teaser*--I installed that Adaptec USB 2.0 controller add-on card on this older system so I could use the USBControl software to map the USB controllers, map how many ports are present (note, I have discovered that there are more ports present than what can be actually accessed), and to map which USB port each port is assigned--so what the sequence of port assignments are for each port.

So, more to come later--

To all

NightOwl wrote on Feb 8^th, 2013 at 12:02pm:

Nobody?! 

Really, no one has ever tried to trouble shoot their USB setup, and needed to know which USB port went to which USB controller?

I'm amazed at how little is available to trace what the *inner workings* are going on with the USB ports and the controllers of a system--mostly kind of a *black box*!  Either it works or it doesn't, and there's no easy way to figure out the what, where, why......

I've come across a couple things that may be helpful.  I'm testing the software to see if if gives me the information I'm looking for in a fashion I (a layman--not a computer hardware engineer!) can understand.

More later....

To all

NightOwl wrote on Feb 22^nd, 2013 at 11:41am:

Back to the *Details* tab in the *Properties* listing in Device Manager--after the *PCI\VIN_1033....*, for each of the *NEC PCI to USB Enhanced Host Controller (B1)*, there was a *DEV_xxxxxSUBSYS_xxxxxx*.  These numbers where unique to each of the NEC chips as well, but still did not identify which NEC chip was from the motherboard and which was from the Adaptec add-on card.

So, I decided to remove the Adaptec add-on card to see which of the two listings in Device Manager disappeared.  (Thinking back now, I probably could have used the feature in Device Manager under the *General* tab at the bottom where you can use the *Device usage:* drop down box to *enable* or *disable* that given device--and then tested to see which USB ports where still active--but, that didn't occur to me at the time.)

So, looking at the *PCI Slot # (PCI bus #, device #, function #)* in Device Manager I now knew which NEC chip was the motherboard chip.  But, the *DEV_xxxxxSUBSYS_xxxxxx* changed!  It now was the same numbers as the Intel's *DEV_xxxxxSUBSYS_xxxxxx* numbers!  At this point, I don't know what those numbers represent--they are apparently not numbers specific to a given device or its location.  But, they apparently change depending on what's hooked up to the system.  And why would the on board NEC chip now match the Intel's numbers after removing the Adaptec's NEC USB chip--don't know!

So, I was going to test the various BIOS settings for the Intel USB chip--it allows for enabling 6 ports, 4 posts, 2 ports or disabled--again testing to see which ports where controlled by what BIOS setting.  Well, low and behold what did I find--the motherboard NEC USB chip was again listed in the BIOS for enabling/disabling!

Apparently, whatever method is used to enable/disable those NEC USB chips must effect both the on board chip as well as the add-on card chip--so the BIOS is designed to remove that functionality if it sees two *identical* chips.

As I said at the beginning, it's interesting to see what effect different configurations have on the behavior of the the BIOS.  It was also interesting to see that the system identification numbers changed in Device Manager depending on whether there was that extra NEC chip present or not.

I will post some information I have come across for trying to map current system's USB controllers and device interconnections that I have found in my next post.

Christer

Quote:

Not to worry--it's a little different topic, but of a similar nature!

Quote:

I understand completely--I too have a newer Gigabyte system--I think we built our systems about the same time--mine is the following:

GA-P55A-UD4P (rev. 1.0)

What is your model?

The BIOS shows a possible 13 IDE Channel connections--but the BIOS list does not specify which ones are the SATA ports (6-SATA2, 2-SATA3, 2-eSATA) and/or 2 PATA devices--and those add up to 12 and not 13 possible devices!  And the channel numbering skips channels 5 and 8

If you look at the motherboard user guide, the ports on the motherboard are identified and numbered--and there are a possible 12 devices that can be hooked up--but they do not correspond to the channel numbering in the BIOS.  And there is no 13th hookup for a 13th device. 

And during the POST, 2 of the channels listed in the BIOS do not show up in the list of possible devices and channels in the POST screen--so that screen is down to only 11 possible channels.

So....what does that all mean ......    ! 

I plan on mapping which motherboard port goes to which POST channel and BIOS listing--if you want to start another thread and explore the possibilities, that would be fine with me.

A little boasting. My new computer has 6 USB 3.0 ports and 10 USB 2.0 ports. The MB has room for another 2 USB 2.0 ports.

Christer

Christer wrote on Mar 6^th, 2013 at 1:57am:

I have already started a new topic thread here:  SATA HDD Ports, BIOS, and POST Info Mismatch

I'm in no hurry as far as getting the exploration of the issue done--so you shouldn't feel compelled to do so either.  I will have to open the case and trace all the SATA communication cables (I thought I had recorded that--but can not find it--buried in some obscure, forgotten notepad somewhere I'm sure!).

I'm also waiting for an *eSATA* to SATA hookup cable that I've ordered, so I can attach one of my spare SATA HDDs through the eSATA ports on the back I/O panel to see where those show up in the BIOS and POST listing of HDDs.

So, it will be a while before I continue with additional post to that thread.

Brian

Brian wrote on Mar 6^th, 2013 at 1:19pm:

Nice lead-in to my ongoing discussion here!

That's exactly right--most of us have no idea which USB port on our system goes to which controller--and which port is *seen* first by each controller, and what the order is of all the ports going to a given controller!

And, often times, it probably doesn't matter to most of us!  That's *okay*--USB just works!

But, when I was doing some testing of USB flashdrives and HDDs for booting from them into DOS to do various Ghost backups--it became *important* to know this information, especially on my older system where the BIOS only has rudimentary support for booting from a USB storage device.  Success or failure to boot depended on which USB port was used--one had to hook up to a USB port from the Intel north/south chipset--hooking up to a USB port from the additional NEC USB controller that was hardwired on the system would fail--as would hooking up to the add-on Adaptec USB PCI card.  The BIOS on this system will recognize only the north/south USB controller for booting to DOS!

And, in some cases, if another device (in particular--the USB card reader for various flavors of memory cards) was hooked up to a USB port seen before the USB device being tested for booting to DOS--this could often result in a *hung* system that stops, and needed to be power cycled to regain control.

So, the USBControl software from Adaptec allowed me to map the USB devices on my older system just fine, and quickly.  But, back in my reply #2 I said:

NightOwl wrote on Feb 8^th, 2013 at 12:02pm:

And, of course, I can still hear the crickets in the distance--the silence is deafening!  At the end of this post is a screen shot showing how USBControl works on my new system that uses an i7 Intel CPU, and the P55 north/south chipset.  It's not working well!

(I have my 4 port Adaptec USB Hub hooked up that allows the USBControl program to see a qualifying Adaptec product hooked up to the system so the program loads!)

There are two components to the USBControl program--the small rectangular box (it shows the individual components hooked up to the system in a vertical list--and that box is actually showing most of the USB devices that are hooked up to the system--however not exactly correctly!)--I have moved that small box onto the larger rectangular box which shows the whole USB tree--so both are in the same screen shot.

I clicked on the top listed item in the small box--it is the HP Printer that is correctly identified.  It is hooked up to controller *B*, and is on the second port.  But on the USB tree in the larger box--all ports are showing as having *nothing* hooked up--and all ports are *available*.  The USB controllers *B* and *C* are the Intel P55 controllers--they are both USB 2.0 controllers--and they should have a small number *2* like seen next to the mouse in the small box.  And, controller *B* actually only has 8 ports and not 12!  And controller *C* actually only has 6 ports and not 13!

USB controller *A* is the Renesas (formerly NEC) Electronics USB 3.0 Host Controller.  USBControl is reporting the correct number of ports for that controller--but, obviously being as the software is over 10 years old now--it has no programing that allows it to understand that it is a USB 3.0 controller--USB 3.0 was only on the drawing board back then, and not implemented.

So, obviously USBControl is of no value on a newer system.  Apparently, the new P55 Intel chipset for i3, i7, and newer CPUs is a radical change from previous USB controllers--and that older Adaptec program can not properly interpret the USB device listing and implementation by the P55 USB controllers.

So, my previous statement in reply #4:

NightOwl wrote on Mar 4^th, 2013 at 1:36am:

Okay, so it wasn't exactly in my *next post*    --but, here it comes.....in my *next post*  !

To All

Quote:

Well, sort of.....at the end of this post is a screen shot showing how Microsolf's *USBView* from the previous post's link to FTDI's Utilities page displays one's USB layout.....

So, you see two panes--left side and right side.  The left side has the USB *tree view* of how your USB devices are hooked up and laid out.  And the right side has details about a given device as highlighted in the left side.  So, the *tree view* is not quite as *graphical* as the Adaptec USBControl view of the USB interconnections--but, it's functional......

I have collapsed one of the Universal USB controllers and the two USB 2.0 Hubs in the left side tree view in order to shorten the tree and get all the controllers to show up in one screen shot of my newest system:  GA-P55A-UD4P (rev. 1.0)

Let me go over some of the information you can gather:

I have circled the terms *Universal* and *Enhanced* controllers--Universal is for support of legacy USB 1.00 and and USB 1.10 devices which gives *low* and *full* USB speeds respectively.  Enhanced is for USB 2.0 devices--which is *high* speed.

If you count, you will find 7 Universal Controllers listed--and each Universal Controller controls two ports--so a total of 14 ports--which is how many USB ports the system has.

But, you will only find 2 Enhanced Controllers--but one has 8 ports and the other has 6 ports--so again a total of 14 USB ports.

In the left side tree, each of the ports has a *Loc* listed--I assume this is short for *Location*--the first number is the Controller number and the second number is the port number on that controller.  As mentioned above, I collapsed the Universal Controller #3, so you will not see it's two ports listed and the *Loc* is not displayed in this screen shot.

In the right pane--the top circled item--the *bcdUSB*--appears to report what version of USB the device reports itself to be.  The reason I say it in that fashion is because certain devices on my system report themselves to be USB v2.00 devices, but they can only operate at version 1.00 or 1.10 bus speeds--so they are really not actually USB 2.0 devices in terms of how they function!

Further down, the *iProduct:* reports the products name that's embedded in the device that's hooked up.

And below that I have highlighted the *Device Bus Speed:* which shows the actual communication speed that the device is operating at.

Back to the left pane--down at the bottom is the *Renesas Electronics USB 3.0 Host Controller*--this is the *NEC* USB 3.0 chip that is embedded on the motherboard.  Notice it does not get a number assigned to it for the first *controller* number like the *Universal* and *Enhanced* controllers do.  That's why I said at the beginning that USBView *sort of* works!  I have to suspect that this version of USBView may be an older version that does not fully recognize USB 3.0 controllers.  I'm going to have to explore that a little bit more to see if that's true.

So, it's easy enough--get a USB 2.0 device and plug it in to the various physical ports and see which port it's listed on under the two *Enhanced* controllers and you will have the outline of the controller/port interconnections.  Then take a USB 1.00 or 1.10 device and do the same thing and you will then know which *Universal* controller is associated with which *Enhanced* controller port and visa versa.  The two types of controller *Enhanced* or *Universal* are sort of *virtual* controllers in that they both are connected to the same physical port and are waiting to take over once a USB device is hooked up--but which one takes control depends on whether it's a USB 2.00 or USB 1.x device.

More to come later.....