If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
USB 2.0 hubs?
I bought two 4-port USB ports to extend one I have on the computer, and they
both work fine, but a bit differently. When I connected each one to the computer, it naturally found the new hardware and installed a driver, as expected. (I'm using Windows XP) However, for one of these items, each time you plug a different USB device into it, it needs to initially enumerate the device, but not so for the other. The only difference I can tell on the specs is that one is called a USB port hub splitter, and the other lacks that term "splitter" (and that the first one appears to be a bit more advanced (for example, allows for external power). Here are the two items. Can someone explain why the difference in enumeration, and that one needs to do it, while the other doesn't? https://www.amazon.com/UGREEN-Splitt...ds=usb+2+ports https://www.amazon.com/AmazonBasics-...=usb+2.0+ports |
Ads |
#2
|
|||
|
|||
USB 2.0 hubs?
Bill in Co wrote:
I bought two 4-port USB ports to extend one I have on the computer, and they both work fine, but a bit differently. When I connected each one to the computer, it naturally found the new hardware and installed a driver, as expected. (I'm using Windows XP) However, for one of these items, each time you plug a different USB device into it, it needs to initially enumerate the device, but not so for the other. The only difference I can tell on the specs is that one is called a USB port hub splitter, and the other lacks that term "splitter" (and that the first one appears to be a bit more advanced (for example, allows for external power). Here are the two items. Can someone explain why the difference in enumeration, and that one needs to do it, while the other doesn't? https://www.amazon.com/UGREEN-Splitt...ds=usb+2+ports https://www.amazon.com/AmazonBasics-...=usb+2.0+ports Items at that price point, aren't likely to have screws for easy disassembly. More likely to be glued together. You can try to use USBTreeView to gather VID/PID, or you can use Device Manager Properties and HardwareID to do much the same thing. https://www.uwe-sieber.de/usbtreeview_e.html If the device had some sort of aggressive power saving, maybe it sleeps and only a plug-in event starts the enumeration process and bus resets. You can check the Device Manager properties and see if there is a tick box to stop it from "sleeping". ******* When you see a device with external power... 1) The power input connector may not be rated to carry enough power to run all ports at full power. A barrel connector might handle 3 to 5 amps on a good day (specs are available). Other connectors might be 1 amp or so (before there is a possibility of the pins burning). The barrel connector is the right solution for this job. 2) A great many powered hubs from China, cut corners. a) There is an app note from a famous chip company that makes USB 4 port hub chips, that shows how to do powering properly. When current starts to flow into the external connector, it operates a small relay. The small relay disconnects VBUS from the PC, from the hub and the hub outputs run off adapter power. No "reverse flow", upsetting the ATX supply, is possible with the relay design. You may hear a "click" if using one of those. But at $7 retail, nobody can afford a relay. The relay costs more than all other parts and the PCB, combined. Forget doing it right! A $35 hub can probably afford to use a relay inside. b) A silly design, could use diodes. This violates the voltage drop spec for the bus. But might still work anyway. The best Schottky diodes I could buy for a home project, have a reasonably small drop at low current. But once you pump an amp through them, they're little better than regular diodes. And of course, in a hub design, everyone treats the power output capability, like the thing is "Niagara Falls". Lots of reviewers are upset because it's not charging their iPad. c) Some designs will just short the hub power input to VBUS, allowing the wall adapter to push current into the ATX PSU. It might not catch fire, but it might not be completely stable either. I don't have a symptom list assembled for this possibility. The notion of external power is nice, but the engineers "went on vacation" when designing the feature. If you use semiconductors to do the power switching and protect the PC, then there are all sorts of corner cases to worry about (where the switch device starts to half-conduct). I had at least one external enclosure, with some pretty weird corner case behavior from a power perspective (the LED would light dimly when the device was "off"). At $7, that's only enough money for one USB chip, and once you acquire enough info to get the part number, you can then Google for "quirks" with the part number in question. That's if Device Manager doesn't have any tick boxes for the hub that explain the "snoozing" habits. It's not likely that there are *any* outboard circuits of significance. And even if a diode costs $0.05, they'll still strip it out to hit the $7 price point for the retail price. I mean, they didn't even add ballast to the thing, to make it handle nicely. Some small electronics, use a weight inside the chassis and rubber feet, to improve handling on the desktop. So it doesn't slide off quite as easily. HTH, Paul |
#3
|
|||
|
|||
USB 2.0 hubs?
Paul wrote:
Bill in Co wrote: I bought two 4-port USB ports to extend one I have on the computer, and they both work fine, but a bit differently. When I connected each one to the computer, it naturally found the new hardware and installed a driver, as expected. (I'm using Windows XP) However, for one of these items, each time you plug a different USB device into it, it needs to initially enumerate the device, but not so for the other. The only difference I can tell on the specs is that one is called a USB port hub splitter, and the other lacks that term "splitter" (and that the first one appears to be a bit more advanced (for example, allows for external power). Here are the two items. Can someone explain why the difference in enumeration, and that one needs to do it, while the other doesn't? https://www.amazon.com/UGREEN-Splitt...ds=usb+2+ports https://www.amazon.com/AmazonBasics-...=usb+2.0+ports Items at that price point, aren't likely to have screws for easy disassembly. More likely to be glued together. You can try to use USBTreeView to gather VID/PID, or you can use Device Manager Properties and HardwareID to do much the same thing. https://www.uwe-sieber.de/usbtreeview_e.html If the device had some sort of aggressive power saving, maybe it sleeps and only a plug-in event starts the enumeration process and bus resets. You can check the Device Manager properties and see if there is a tick box to stop it from "sleeping". ******* When you see a device with external power... 1) The power input connector may not be rated to carry enough power to run all ports at full power. A barrel connector might handle 3 to 5 amps on a good day (specs are available). Other connectors might be 1 amp or so (before there is a possibility of the pins burning). The barrel connector is the right solution for this job. 2) A great many powered hubs from China, cut corners. a) There is an app note from a famous chip company that makes USB 4 port hub chips, that shows how to do powering properly. When current starts to flow into the external connector, it operates a small relay. The small relay disconnects VBUS from the PC, from the hub and the hub outputs run off adapter power. No "reverse flow", upsetting the ATX supply, is possible with the relay design. You may hear a "click" if using one of those. But at $7 retail, nobody can afford a relay. The relay costs more than all other parts and the PCB, combined. Forget doing it right! A $35 hub can probably afford to use a relay inside. b) A silly design, could use diodes. This violates the voltage drop spec for the bus. But might still work anyway. The best Schottky diodes I could buy for a home project, have a reasonably small drop at low current. But once you pump an amp through them, they're little better than regular diodes. And of course, in a hub design, everyone treats the power output capability, like the thing is "Niagara Falls". Lots of reviewers are upset because it's not charging their iPad. c) Some designs will just short the hub power input to VBUS, allowing the wall adapter to push current into the ATX PSU. It might not catch fire, but it might not be completely stable either. I don't have a symptom list assembled for this possibility. The notion of external power is nice, but the engineers "went on vacation" when designing the feature. If you use semiconductors to do the power switching and protect the PC, then there are all sorts of corner cases to worry about (where the switch device starts to half-conduct). I had at least one external enclosure, with some pretty weird corner case behavior from a power perspective (the LED would light dimly when the device was "off"). At $7, that's only enough money for one USB chip, and once you acquire enough info to get the part number, you can then Google for "quirks" with the part number in question. That's if Device Manager doesn't have any tick boxes for the hub that explain the "snoozing" habits. It's not likely that there are *any* outboard circuits of significance. And even if a diode costs $0.05, they'll still strip it out to hit the $7 price point for the retail price. I mean, they didn't even add ballast to the thing, to make it handle nicely. Some small electronics, use a weight inside the chassis and rubber feet, to improve handling on the desktop. So it doesn't slide off quite as easily. HTH, Paul Thanks. I tried usbview but didn't glean all that much from it to explain the difference as to why one hub cares about what's plugged in (installs a new driver the first time it's used with a different USB device) and the other doesn't, for some reason which I still haven't figured out. I'm thinking it's because the UGreen one is acting more like an active hub, if there is such a thing, and the Amazon is acting more like a passive one, and for some reason doesn't care what's plugged into it (no new hardware icon comes up). But I don't really understand all of this. |
#4
|
|||
|
|||
USB 2.0 hubs?
Bill in Co wrote:
Paul wrote: Bill in Co wrote: I bought two 4-port USB ports to extend one I have on the computer, and they both work fine, but a bit differently. When I connected each one to the computer, it naturally found the new hardware and installed a driver, as expected. (I'm using Windows XP) However, for one of these items, each time you plug a different USB device into it, it needs to initially enumerate the device, but not so for the other. The only difference I can tell on the specs is that one is called a USB port hub splitter, and the other lacks that term "splitter" (and that the first one appears to be a bit more advanced (for example, allows for external power). Here are the two items. Can someone explain why the difference in enumeration, and that one needs to do it, while the other doesn't? https://www.amazon.com/UGREEN-Splitt...ds=usb+2+ports https://www.amazon.com/AmazonBasics-...=usb+2.0+ports Items at that price point, aren't likely to have screws for easy disassembly. More likely to be glued together. You can try to use USBTreeView to gather VID/PID, or you can use Device Manager Properties and HardwareID to do much the same thing. https://www.uwe-sieber.de/usbtreeview_e.html If the device had some sort of aggressive power saving, maybe it sleeps and only a plug-in event starts the enumeration process and bus resets. You can check the Device Manager properties and see if there is a tick box to stop it from "sleeping". ******* When you see a device with external power... 1) The power input connector may not be rated to carry enough power to run all ports at full power. A barrel connector might handle 3 to 5 amps on a good day (specs are available). Other connectors might be 1 amp or so (before there is a possibility of the pins burning). The barrel connector is the right solution for this job. 2) A great many powered hubs from China, cut corners. a) There is an app note from a famous chip company that makes USB 4 port hub chips, that shows how to do powering properly. When current starts to flow into the external connector, it operates a small relay. The small relay disconnects VBUS from the PC, from the hub and the hub outputs run off adapter power. No "reverse flow", upsetting the ATX supply, is possible with the relay design. You may hear a "click" if using one of those. But at $7 retail, nobody can afford a relay. The relay costs more than all other parts and the PCB, combined. Forget doing it right! A $35 hub can probably afford to use a relay inside. b) A silly design, could use diodes. This violates the voltage drop spec for the bus. But might still work anyway. The best Schottky diodes I could buy for a home project, have a reasonably small drop at low current. But once you pump an amp through them, they're little better than regular diodes. And of course, in a hub design, everyone treats the power output capability, like the thing is "Niagara Falls". Lots of reviewers are upset because it's not charging their iPad. c) Some designs will just short the hub power input to VBUS, allowing the wall adapter to push current into the ATX PSU. It might not catch fire, but it might not be completely stable either. I don't have a symptom list assembled for this possibility. The notion of external power is nice, but the engineers "went on vacation" when designing the feature. If you use semiconductors to do the power switching and protect the PC, then there are all sorts of corner cases to worry about (where the switch device starts to half-conduct). I had at least one external enclosure, with some pretty weird corner case behavior from a power perspective (the LED would light dimly when the device was "off"). At $7, that's only enough money for one USB chip, and once you acquire enough info to get the part number, you can then Google for "quirks" with the part number in question. That's if Device Manager doesn't have any tick boxes for the hub that explain the "snoozing" habits. It's not likely that there are *any* outboard circuits of significance. And even if a diode costs $0.05, they'll still strip it out to hit the $7 price point for the retail price. I mean, they didn't even add ballast to the thing, to make it handle nicely. Some small electronics, use a weight inside the chassis and rubber feet, to improve handling on the desktop. So it doesn't slide off quite as easily. HTH, Paul Thanks. I tried usbview but didn't glean all that much from it to explain the difference as to why one hub cares about what's plugged in (installs a new driver the first time it's used with a different USB device) and the other doesn't, for some reason which I still haven't figured out. I'm thinking it's because the UGreen one is acting more like an active hub, if there is such a thing, and the Amazon is acting more like a passive one, and for some reason doesn't care what's plugged into it (no new hardware icon comes up). But I don't really understand all of this. This is one of the informal lists of USB devices. Your USB software tool (or Device Manager) should have two four digit hex numbers for vendor and device. http://www.linux-usb.org/usb.ids I don't think there is a distinction on hubs. The hub interacts from a protocol perspective. It reclocks the data and the time to reclock the information is part of the overall USB time budget. The time budget limits operation to a tree depth of five hubs or so. On modern motherboards, Intel added a hub to the PCH, limiting the user to four external hubs in a row. You can hit that limit if using one-port ("active") USB extension cables. A powered hub and an unpowered hub, just differ on where VBUS comes from. An unpowered hub has a limitation on what the subtending devices can draw. I was surprised though, at what kind of loads you could run on USB extension cables (which are unpowered), as I was able to run a webcam with focus motor off the end of three cables. I was expecting to only be able to run a USB keyboard off the (reclocked) cable chain. The address limit on USB is 127 devices and I presume the hub itself has an address. I don't know if USB3 attempted to change these limits at all or not. Some aspect of polling may have changed on USB3, and electrically, USB3 (the 5 pin wiring portion) is full duplex. USB3 includes a high speed section (full duplex) and the low speed USB2 portion (half duplex). While the two interfaces could be used to move data at the same time, software limits operation to the negotiated interface. TX+ \ TX- \ GND \__ USB3 rate portion RX+ / RX- / VBUS \ D+ \___ USB2/UWB1.1 legacy portion D- / GND / Paul |
Thread Tools | |
Display Modes | |
|
|