DavR Posted January 18, 2017 Posted January 18, 2017 (edited) Hi all, I have discussed some of this with Promethean support, but I'd be interested to hear others experiences. We have a lot of older Promethean ActivBoards, chiefly ActiveBoard 1 & 2 models, which are still plodding along because on the whole, they do work. I'd love to upgrade to touchscreens, but the money just isn't there, so we just keeping replacing projectors. I've started to notice that these are not all that happy working with USB3 devices. Looking in event logs, the boards are occasionally disconnecting, and then coming back a few minutes later. Anecdotally, teachers have reported the Windows device removed chimes, which would fit in with it. I say USB3 as Promethean confirm this is a recognised bug with older boards, with USB3 seemingly making less power available than USB2. This is confirmed by the one board I can see running connect to an old USB2 PC just fine. Things I have tried to prevent these disconnects: Latest ActivDriver Latest board firmware Latest USB3 drivers Reducing cable runs to under 3m, PC to controller Active USB repeater cable Has anyone else experienced this kind of behaviour with older Promethean boards and USB3 connections? It's with Dell OptiPlex 3020 or 3040 workstations if that makes any odds. Thanks! Edited January 18, 2017 by DavR
FN-GM Posted January 18, 2017 Posted January 18, 2017 I have seen it with a few other devices. Its a pain, I use USB 2 ports if I come across it. Is there any USB setting in the BIOS of the machines that could help? 1
DavR Posted January 18, 2017 Author Posted January 18, 2017 You mean using the USB2 ports instead of USB3 on the same machine? Have you had any success with that? I've heard that the USB3 driver can be fairly dominant and control some USB2 behaviour also :/ I think I'm running my test case PC out of a USB2 port on the back, but that is something worth checking for sure tomorrow.
tech-man Posted January 18, 2017 Posted January 18, 2017 Watching this post with great interest!!! Same issues going on here..... 1
FN-GM Posted January 18, 2017 Posted January 18, 2017 You mean using the USB2 ports instead of USB3 on the same machine? Have you had any success with that? I've heard that the USB3 driver can be fairly dominant and control some USB2 behaviour also :/ I think I'm running my test case PC out of a USB2 port on the back, but that is something worth checking for sure tomorrow. We have Optiplex 3240 machines AIO machines. It wasn't an IWB but a CNC machine. Didn't work properly with USB 3.0, soon as it went into a USB 2.0 port in the same machine it kicked into life. Took ages for me to figure it out!
thebatman Posted January 19, 2017 Posted January 19, 2017 We came across this issue in the summer with updating out teacher PCs. We found if the USB cable was under 3m it was ok, anything over 3m we needed to use an active USB cable.
phillyrovers Posted January 19, 2017 Posted January 19, 2017 Hi All, The USB3 "issue" is a really interesting one, and I know a lot of work went on internally at Promethean specifically, to work out the root cause of the issue. Some high level findings from the investigations: Computers that use the Intel 8 Series USB chipset are the most "at risk" of seeing the issue (this is seen in the computer's device manager under Universal Serial Bus controllers - Intel® 8 Series/ C220 Series USB EHCI) Even if a computer has got USB 2 ports, the USB communication is handled by the USB 3 chipset, so the issue could still be seen. The actual "issue" is caused by the Intel 8 Series chipset being too impatient when attempting to talk to some USB devices. When a USB device is connected to a computer the computer sends a signal that says "who are you, and what can you do?" it has been found that the Intel 8 series chipset is sending another "who are you" request BEFORE the device has had time to send it's information back, in effect crippling the communication link. We believe that portions of this were resolved in a driver update which Intel released, and this issue has not been seen on their Intel 10 Series chipset (so far). The workarounds which have been found so far are: - To use a USB 2 hub inline with the connection, this slows the whole communication speed down and allows the computer and USB 2 device to communicate correctly. - Turn off USB 3 in the BIOS (if possible). - Use a USB 2 repeater (some of these can be hit and miss if they work or not. - Carry out all Windows updates. The above is assuming that you haven't got something else going on which could also cause a disconnect such as power management settings etc, which could be a simple fix if it is found after a period of 10 minutes of none use the boards drop, it is likely to the USB selective suspend setting which is causing the problem, rather than a USB 3 computer. Thanks 2
dhicks Posted January 19, 2017 Posted January 19, 2017 Reducing cable runs to under 3m, PC to controller The issue can be with the model of USB extender / baluns for running longer USB connections over Cat5 cable instead of USB (even three metres seems to be a bit too long for some USB 3 ports to cope with) - we've found this one to be reliable: https://www.amazon.co.uk/gp/product/B001TOG6MM 1
DavR Posted January 19, 2017 Author Posted January 19, 2017 The actual "issue" is caused by the Intel 8 Series chipset being too impatient when attempting to talk to some USB devices. When a USB device is connected to a computer the computer sends a signal that says "who are you, and what can you do?" it has been found that the Intel 8 series chipset is sending another "who are you" request BEFORE the device has had time to send it's information back, in effect crippling the communication link. We believe that portions of this were resolved in a driver update which Intel released, and this issue has not been seen on their Intel 10 Series chipset (so far). Interesting to see a little more detail into the issue, thanks! I was always thinking of it more of a power issue, rather than response timings. I can confirm that the OptiPlex 3020s we have exhibiting this are the 8 Series USB chipset, which I knew was a known issue. Updated chipset drivers haven't resolved the issue though, and I'm sorry to say we have newer model OptiPlex 3040 (Intel 100 Series chipsets) giving the same issue. When you talk about putting a USB2 hub inline, would you recommend powered or just a regular one? I've shied away from doing this thus far as I'd assumed it was a power issue, and was dreading trying to find extra power points in the rooms. 1
markiof Posted January 19, 2017 Posted January 19, 2017 Hi We have had success in this area using regular unpowered USB hubs - worth a go! 1
DavR Posted January 19, 2017 Author Posted January 19, 2017 I think I'm running my test case PC out of a USB2 port on the back, but that is something worth checking for sure tomorrow. Well, on further investigation turns out I was running into a USB3 port on my test case PC. I've moved that over to a USB2 port, perhaps the combination of USB2 port and USB2 repeater cable will be enough, I'll have to monitor it over the next few days. Suspect it probably won't though, as I assume I've still got a USB3 controller running my USB2 ports, and that's where the misunderstanding lies? I have BIOS options to disable USB3 ports, but I doubt this is going to downgrade my USB3 controller, just nobble the ports...
DavR Posted January 19, 2017 Author Posted January 19, 2017 (edited) Hi We have had success in this area using regular unpowered USB hubs - worth a go! Oh cool - I'm hopefully placing a parts order in the next week or so, so I'll chuck one in the basket. If all I have to do is stick a £10 USB2 box between board and PC I could live with that! Edited January 19, 2017 by DavR
Chrism03 Posted January 23, 2017 Posted January 23, 2017 I've tried changing cables, lengths, ports, drivers and a different PC without success but a quick swap with an older laptop (no USB 3) and the board worked fine.
DavR Posted January 23, 2017 Author Posted January 23, 2017 Just checked back on my test machine, and even though it is running through a USB2 port on that machine, it is still having the error, I guess as it's still running off a USB3 controller. Running on something older that runs a USB2 controller does work, I have one board running that way. Not really a workable solution for the others though!
markiof Posted January 23, 2017 Posted January 23, 2017 As phillyrovers said ealier "Some high level findings from the investigations: Computers that use the Intel 8 Series USB chipset are the most "at risk" of seeing the issue (this is seen in the computer's device manager under Universal Serial Bus controllers - Intel® 8 Series/ C220 Series USB EHCI) Even if a computer has got USB 2 ports, the USB communication is handled by the USB 3 chipset, so the issue could still be seen. The actual "issue" is caused by the Intel 8 Series chipset being too impatient when attempting to talk to some USB devices. When a USB device is connected to a computer the computer sends a signal that says "who are you, and what can you do?" it has been found that the Intel 8 series chipset is sending another "who are you" request BEFORE the device has had time to send it's information back, in effect crippling the communication link. We believe that portions of this were resolved in a driver update which Intel released, and this issue has not been seen on their Intel 10 Series chipset (so far)." this agrees with what you guys are seeing, have you tried checking for driver updates direct from the manufacturer (not within windows) for the USB controller? Can you alter the speed from within the BIOS?
DavR Posted January 23, 2017 Author Posted January 23, 2017 this agrees with what you guys are seeing, have you tried checking for driver updates direct from the manufacturer (not within windows) for the USB controller? Can you alter the speed from within the BIOS? Yup, updated directly from Intel, no change. This was a few months back, but doubt they've made any breakthroughs since then. BIOS has on or off options for USB ports, nothing more than that. My next hope is a USB2 hub, which is being ordered as we speak hopefully!
DavR Posted February 6, 2017 Author Posted February 6, 2017 Latest update on this one, just been trying routing the connection through a USB2 hub, with the logic that the board talks to the USB2 controller from the hub, rather than the USB3 controller on the PC. It worked for three days straight last week, and I allowed my hopes to get up... nope. Happening loads this morning! So that was a USB2 hub, into a USB2 port on the PC, with a 3m cable direct to the board controller. With latest drivers and latest firmware. Still hasn't resolved the issue. Can anyone offer any further advice on this one? The only other thing I can think of at this point is fitting USB2 PCI-E cards, which would be far from ideal (not to mention a pain to find these days).
peferraro Posted February 6, 2017 Posted February 6, 2017 Hi Dave, Before you do that, you mentioned that the boards are at the latest firmware version. Please supply the version number and also the driver version number.
DavR Posted February 6, 2017 Author Posted February 6, 2017 Hi @peferraro, Currently on driver version 5.14.21 and firmware 3.75.
peferraro Posted February 6, 2017 Posted February 6, 2017 Hi Dave, At this point please call tech support, here is the link to find the number https://support.prometheanworld.com/contact/.
DavR Posted February 6, 2017 Author Posted February 6, 2017 Hi @peferraro - I already have an open support call with them, they don't seem to be able to offer me anything further. Looks like I may just have to lump this one
Recommended Posts