OESF Portables Forum
Everything Else => Sharp Zaurus => Model Specific Forums => Distros, Development, and Model Specific Forums => Archived Forums => C1000/3x00 Hardware => Topic started by: Mjolinor on February 16, 2006, 04:33:09 pm
-
The roms I have had on my 3000 all report two USB host ports so I decided to go looking for them. I suspect that these ports come directly from some crappy BGA device specifically designed to make it hard and I can't trace them from the port on the outside to a chip on the inside.
So what I can find is that the USB port goes to the square arrangement shown in one of the attachments. The other attachment is a drawing of the connections to that square that I can work out.
The two at the top left go to the sync plug for USB connection. They are probably the USB client for the synch cable. Is it allowable to ahve a miniB and the synch cable plugged in at the same time?
Any suggestions for tying it down some more looking for this elusive second USB host port?
-
The PXA270 is a 'system on chip' device. What we may be seeing is the host reporting some of the USB controller capabilities of the PXA270 rather than the supporting hardware that actually implements the bus.
Just a thought.
- Andy
-
The pxa has 3 HOST porhs and 1 client, HOST port 2 and the client can be switched between in software (allowing you HOST one second then CLIENT the next the other two are unbalanced and cannot be used as general perpose USB ports without circutry
that said if you find the others (if they were routed to somwhere else on board) then i can wip you up a schematic to allow you to use it
-
Actually this would be something good to answer on this thread.
Does anyone know the upper limit on power supplied to devices from a Zaurus? either safely or just in terms of reliability?
I'm not suggesting running a USB supplied Kettle, Grill and disco lights off one, however, usbdev -l on OpenBSD reports 100 mA for a typical mouse device and a whopping 300 mA for a Sitecom USB NIC.
-Andy
-
USB specification dictates 500ma or it can't be called USB.
-
Cool, I saw something on an OpenBSD thread from one of the Zaurus team saying that one NIC in particular HAD to be hung off a powered hub though because the Zaurus couldn't power it properly.
- Andy
-
The power to OTG port is povided by TOKO's TK3850 5V voltage regulator. Its datasheet specifies peak output current @ 1000mA. There's a way to control the max. output current by using a resistor between PCL terminal and the GROUND. If PCL is wired to the GROUND, then there;s no limit, but it could be dangerous, since in this case there's no over-current protection. I think Sharp decided to play it safe, and they use R = 6.3K, which limit output current to approx. 200mA. By changing it to R = 2K, the limit can be upped to approx. 400-450mA. R = 1K should further up it to approx. 600-650mA, which should be sufficient to drive most wi-fi usb dongles. On the other hand, there might be some technical reason to why to limit it to 200mA?
The TOKO's datasheet specifies dropout as 170mV @ 500mA.
The input voltage comes from VCC 5V power domain which might have its own limitations.
Here's the datasheet: http://www.toko.com.hk/Catalog/semi/linear...ata738xxb-E.pdf (http://www.toko.com.hk/Catalog/semi/linear_regulator/pdf/data738xxb-E.pdf)
-albertr
-
OK then with that information do Sharp actually advertise it as a USB host port in the original documentation?
-
The power to OTG port is povided by TOKO's TK3850 5V voltage regulator. Its datasheet specifies peak output current @ 1000mA. There's a way to control the max. output current by using a resistor between PCL terminal and the GROUND. If PCL is wired to the GROUND, then there;s no limit, but it could be dangerous, since in this case there's no over-current protection. I think Sharp decided to play it safe, and they use R = 6.3K, which limit output current to approx. 200mA. By changing it to R = 2K, the limit can be upped to approx. 400-450mA. R = 1K should further up it to approx. 600-650mA, which should be sufficient to drive most wi-fi usb dongles. On the other hand, there might be some technical reason to why to limit it to 200mA?
The TOKO's datasheet specifies dropout as 170mV @ 500mA.
The input voltage comes from VCC 5V power domain which might have its own limitations.
Here's the datasheet: http://www.toko.com.hk/Catalog/semi/linear...ata738xxb-E.pdf (http://www.toko.com.hk/Catalog/semi/linear_regulator/pdf/data738xxb-E.pdf)
-albertr
[div align=\"right\"][a href=\"index.php?act=findpost&pid=115605\"][{POST_SNAPBACK}][/a][/div]
Ok, so in lay persons terms the output isn't enough to drive my USB Ethernet Adaptor which is showing...
# usbdevs -v
Controller /dev/usb0:
addr 1: full speed, self powered, config 1, OHCI root hub(0x0000), PXA27x(0x0000), rev 1.00
port 1 powered
port 2 addr 2: full speed, power 300 mA, config 1, LN029(0x182d), Sitecom(0x6189), rev 0.01
#
?
-Andy
-
The max.power that manufactures specify for devices might not be very accurate. Also, the device might not be able to hit its specified peak current consumption under "normal" working conditions. So, there's no simple answer to this question, imho.
All we know is that OTG port on pxa270-based Z's would max. out @ approx. 200mA, but that could be changed by simply replacing the controlling resistor,
-albertr
-
The max.power that manufactures specify for devices might not be very accurate. Also, the device might not be able to hit its specified peak current consumption under "normal" working conditions. So, there's no simple answer to this question, imho.
All we know is that OTG port on pxa270-based Z's would max. out @ approx. 200mA, but that could be changed by simply replacing the controlling resistor,
-albertr
[div align=\"right\"][a href=\"index.php?act=findpost&pid=115614\"][{POST_SNAPBACK}][/a][/div]
or by dragging out that old Belkin powered USB hub that I had stored in a box somewhere, thanks AlbertR, this possibly explains some of the stability problems I was having with this device.
- Andy
-
Myself is using a realtek8150- based no-name usb nic dongle, which has specified power consumption of 120mA. It works fine on the Z, but I had to tweak its driver somewhat.
-albertr
-
The max.power that manufactures specify for devices might not be very accurate. Also, the device might not be able to hit its specified peak current consumption under "normal" working conditions. So, there's no simple answer to this question, imho.
All we know is that OTG port on pxa270-based Z's would max. out @ approx. 200mA, but that could be changed by simply replacing the controlling resistor,
-albertr
[div align=\"right\"][{POST_SNAPBACK}][/a][/div] (http://index.php?act=findpost&pid=115614\")
200mA hey?
I was trying to remember what the max was when I was looking at these in a store:
[a href=\"http://www.x-micro.com/minidisgo.htm]http://www.x-micro.com/minidisgo.htm[/url]
They seem like a nice storage option but they say that they need 500mA (at spin up, I'm guessing) With no option for externel power one would have to get creative to make it work with a Zaurus and without a powered hub.
-
I'm not suggesting running a USB supplied Kettle, Grill and disco lights off one, however, usbdev -l on OpenBSD reports 100 mA for a typical mouse device and a whopping 300 mA for a Sitecom USB NIC.
-Andy
[div align=\"right\"][a href=\"index.php?act=findpost&pid=115598\"][{POST_SNAPBACK}][/a][/div]
Cool.
I can't seem to find it, is there a command like usbdev in pdaXrom?
-
So it looks like if you want an external hard drive then it's probably better to go for something like an iPod and run it in disk mode then... that way you are assured that it is self powered.
Hmm, I wonder if an iPod firmware will complain if I put an OpenBSD disklabel on it.
- Andy
-
Ok, here's something interesting (now I have that old hub plugged in)... I decided to plug in my SD card reader and have now...
# usbdevs -v
Controller /dev/usb0:
addr 1: full speed, self powered, config 1, OHCI root hub(0x0000), PXA27x(0x0000), rev 1.00
port 1 powered
port 2 addr 2: full speed, self powered, config 1, TUSB2046 hub(0x2046), Texas Instruments(0x0451), rev 1.25
port 1 addr 5: full speed, power 500 mA, config 1, USBHS SD Reader(0x1103), vendor 0x07d0(0x07d0), rev 1.13
port 2 powered
port 3 addr 4: low speed, power 100 mA, config 1, Wireless Mouse(0x0011), Wireless Mouse(0x05fe), rev 1.01
port 4 addr 3: full speed, power 300 mA, config 1, LN029(0x182d), Sitecom(0x6189), rev 0.01
#
The reader shows 500 mA, and that thing has been absolutely bullet proof on the Z running without hub/power.
Or, I'm guessing this value is just a 'guess' on the part of the hardware as to what it may require.
- Andy
-
I believe that's something manufactures code into usb descriptors, but not all of them use accurate specs, etc. Some use reference designs/firmware, but then change things to either better or worse. Also, reference designs might be spec'ed for max. possible current requirements, which rarely a case.
-albertr
-
Fair enough, I guess that the device may not peak out at anyway near that value and it may be simply laziness on the part of the engineers to determine what the requirements are. It's safe to say that if they program the device to report a requirement of 500ma then they have themselves covered.
Anyway, the stability issue didn't seem to be down to juice.. I managed to recreate the failure attached to the powered hub also so I will do a little more testing. Thanks for all the info/advice anyway AlbertR.
- Andy
-
200mA hey?
I was trying to remember what the max was when I was looking at these in a store:
http://www.x-micro.com/minidisgo.htm (http://www.x-micro.com/minidisgo.htm)
They seem like a nice storage option but they say that they need 500mA (at spin up, I'm guessing) With no option for externel power one would have to get creative to make it work with a Zaurus and without a powered hub.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=115619\"][{POST_SNAPBACK}][/a][/div]
I found someone that let me test my Zaurus with the 5gig model.
It auto mounted it under /mnt/usbstorage/
...and I was able to copy a text file with the word "test" in it onto the hard drive.
I know it was a very small file but I would think that that would be enough to spin up the drive. Am I wrong?
-
sounds right to me as long as you flushed the buffers (ie manual unmounting, or typing sync on the command line)