Re: N900 will not allow USB OTG!
 

looks like he has set the N800 up as host/bridge to transfer the files from the N900. Pretty cool setup might have a go laters. :)

Re: N900 will not allow USB OTG!
 

Although fun, this is not really the solution most are looking for! This is pretty much the same as plugging an N900 and USB flash/hard drive into a PC and dragging files from one drive to another, using the PC as USB host. I think most people here are looking to be able to connect USB devices directly to an N900, carrying a minimum amount of extra equipment.

Re: N900 will not allow USB OTG!
 

I still argue that an N8x0 tablet is a minimum of extra equipment. You're just not going to find a PC this small anywhere else.

I'm pretty confident that the hardware limitations of the N900 mean that you're going to need the battery pack, power injection cable and the hub even if you manage to hack the N900 to be a USB host.

The N8x0 tablets are very compact little devices, and are much more capable and easier to hack than anything else proposed so far. The problem is finding one if you don't have one already.

Oh, and if you don't want quite so many cables, you can also push files from the tablet (and the connected USB device) to the N900 via bluetooth or wifi. I saw my N900 show up on the N800's file manager under bluetooth, even before I connected it via USB.

Re: N900 will not allow USB OTG!
 

I'd have to add that an externally powered USB hub's with the correct wiring, resistors, etc are quite smaller than an N8xx - I've build several working prototypes in the past for various devices and they can be quite compact, let alone a fraction of the cost of an N8xx.

We may do R&D in getting USB host mode working on the N900; if and when that happens, we may very well decide to produce all in one hubs specifically for the N900 in the ~$20-25 range since we have the resources to produce them.

Re: N900 will not allow USB OTG!
 

neopwn:

Oh man I would love one of those USB hubs for my N800, even if you never got a USB hack for the N900 working. Of course, the hack would be vastly superior.

It would be especially cool if the hub and a 4AA battery pack were combined together, making it completely portable, and useful as a charger for the N900 when on the road.

Re: N900 will not allow USB OTG!
 

I have doubts the n900 usb is very usable as host. Presumably for their production kernel nokia fixed almost all problems they could find related to peripheral but did not prioritize host. I don't know how many changes there were between 2.6.28 and 2.6.29 for usb, but there are 30 patches for usb against linux-omap-2.6.29 in the openembedded kernel.

Re: N900 will not allow USB OTG!
 

For the mean time (and when I get time) I'll do a write up and include schematics for you.. pretty much all of the components are off the shelf to DIY.

Cheers

Re: N900 will not allow USB OTG!
 

Neopwn, I have electronics experience and would be very interested in building a proto and helping you test.

I also have a laser cutter and would be willing to build test enclosures for you. I can cut acrylic, plywood, etc. I'm very skilled at programming the laser.

Furthermore, if you get this working, I'll treat each member of your team to a free N900 stand or set of stands from my store -- just as a gift. This is my number one disappointment with the N900, anyone working on this is working to make the N900 what is should have been out of the box, and deserves to be rewarded.

Re: N900 will not allow USB OTG!
 

I have used a similar arrangement for transferring photos from Canon A530 to Nokia N800, using Nokia 770 as host and running gphoto2 to transfer the files. (You may remember the abysmal performance of N8x0 in trying to run gphoto2.) For your purpose as well, there would be advantages in using the 770, notably:

1. You can buy non-WSOD Nokia 770 for about $55 (three recent auctions on eBay).
2. 770 does not automount, and it will therefore not attempt to mount devices and partitions that you would prefer it to leave alone. (Undisciplined automounting is one cause of the problems with gphoto2 on N8x0.)
3. Lack of internal power connection to the USB transceiver is actually an advantage if you are going to use power injection in any case. This isolation protects the host from fluctuations in the USB voltage line and ESPECIALLY from static buildup in the "ground" circuits. (Static buildup is possibly the biggest problem for camera connection to N800 host, i.e. not using 770. This varies in importance from one camera to another. With one of my cameras, the N800 is practically useless unless it has a genuine ground connection, whereas the other camera gives less trouble. It may be significant that the more sophisticated camera is the one that needs proper grounding. No doubt other gadgets also vary widely in the extent to which they create static buildup that could affect a USB host.)

The disadvantages of the 770 as host are:

1. The low-speed bus. This should not be a problem unless you want to play video via real-time USB transfer.
2. The need for a USB gender-changer and the poor quality of the few that are available. I have found that most problems with host mode on the 770 can be traced to sloppy receptacles on the gender-changer.

There is a prevalent myth that host mode on the 770 is difficult and mysterious, requiring use of the flasher utility. I hope that an earlier post in this thread has helped to dispel that myth. (Use of a USB keyboard could lead to problems if you activated host mode without the flasher. Those problems are avoidable.)

ADDENDUM 1, Feb 24: I don't want to pretend that grounding the USB cable is a workaround for the bug reported in https://bugs.maemo.org/show_bug.cgi?id=3281 . The ground connection seemed to improve things somewhat for a while, that is all. The unfortunate fact is that you can't rely on N8x0 as a USB host for many devices beyond the basic HID and Mass Storage. Nokia 770 is a better choice.

ADDENDUM 2: Nokia 770 has its own strangeness as host in this setup. After the N800 was plugged in, the 770's dmesg offered all three partitions of the N800's internal card as candidates for mounting (sda1, sda2, and sda3). From past experience with multiple Mass Storage drives plugged into a hub, I expected to see the external card offered as sdb1, but nothing came up.

P.S. 770 will also run the 0xFFFF flasher to update N8x0's rootfs without disturbing a customized initfs.

Re: N900 will not allow USB OTG!
 

Hi, I have not read the entire 61 page thread. Is the USB OTG crippled on this device and I have fall back to a n810?

The n810 can use the OTG for a 100/10 RJ45 Ethernet connector and use a USB mouse/kbd, and even better I once found USB to serial port, which I use alot to connect to servers with. Sadly, the n810 lacked the TV-out.

Re: N900 will not allow USB OTG!
 

The USB OTG is non-existent on the N900 - there's something in the USB specs which prohibits USB OTG with charging, or without supplying power, or something like that. Anyway, the USB chipset in the N900 is (apparently) not wired up to allow enabling host mode at all.

Re: N900 will not allow USB OTG!
 

Just read this article from the maemo mailing list. I have added it here in case it was missed:-

http://www.gossamer-threads.com/list...velopers/55703

Re: N900 will not allow USB OTG!
 

Are we dealing here with a portable-computing question of how to bring the N900's powers of processing and display to bear on data that it must communicate by USB connection with a relatively dumb gadget like a hard drive or a printer?

Or do we have a pride-of-ownership question in which new owners seek affirmation that their expensive purchase doesn't need any help from the last-year-but-one model, thank you? (Much less the 2005 model.)

Re: N900 will not allow USB OTG!
 

clearly what i have read on here n900 cannot do USB OTG(hardware limitation?) so why is this thread still going on and on?

Re: N900 will not allow USB OTG!
 

people likes talking, and this is talk.maemo.org so its fine :)

seriously, officially its a no-go, and here we are searching for a workaround. the best efforts (of kernel debug) appears to be in a dead end, sadly


then, usb over ip efforts are in their own thread now

Re: N900 will not allow USB OTG!
 

Well, alot of people don't give a damn about USB OTG, but would very much like USB host mode.

Re: N900 will not allow USB OTG!
 

N900 uses an integrated Mentor Graphics OTG controller * (musb
mhdrc ) but hardware is acording wiki nxp isp 1707 **.
When a hacked cable ( a mode) is plugged into usb port on vbus it can detect a voltage pulse coming from n900. This is part of usb otg signaling protocol. Also on /sys/devices/platform/musb_hdrc/mode we can read a_idle
until now we have :
1) ID pin state is transmited to kernel
2) voltage can be applied on vbus
so in my opinion the limitation is only at software level
why nokia coose NXP ( former Philips ) and Mentor Graphics usb kernel Nucleus instead of full TI asics line wich can be suported in desktop like linux I can't understand, meanwhile apple continue to sell crap ifone.
Why Nokia screw usb host mode also i can't understand. Battery charging when usb host mode is selected is possible and with aid an external device host mode can be achieved ( I will ad a link later )
I start digging on datasheets but without nokia support community can't do more.
BTW in otg mode host and peripheral mode can be swaped . Can n900 perform this shift ?

* http://www.mentor.com/products/embed...s/nucleus_usb/
** http://wiki.maemo.org/N900_Hardware_isp1707

Re: N900 will not allow USB OTG!
 

Damn you for screwing this up, Nokia. Even the f'n Droid can do this while my "mobile computer" can't.

Re: N900 will not allow USB OTG!
 

moto
" Maybe devil is not so black "
aprox google translation from romanian

The Idea
n900 at least can / must perform usb otg. I initiated otg SRP by software as peripherial. With help of an aditional hardware maybe is possible to switch host mode using HNP.
last days I digging on datasheets and try to imagine how to isolate DP / DN signaling from usb stick, hub during HNP & SRP without compromise data lines.
I will dig my old colection of IC and will try to make a functional device
keep fingers crossed

Later edit
SRP session request protocol
HNP host negotiation protonol

Re: N900 will not allow USB OTG!
 

Thanks, but what do all those abbreviations mean?

Re: N900 will not allow USB OTG!
 

This is an interesting thread :D I can't wait to see someone actually hack the USB OTG to work :D

Re: N900 will not allow USB OTG!
 

" The devil is not so black as he is painted "


blue_led, thanks for your continued work on this. If you want a set of N900 stands from my store, let me know, I'll mail them to you. If you want something small laser - cut, let me know, and I'll cut it and include that, too.

Re: N900 will not allow USB OTG!
 

I have been able to get the device to report the same power and devctl registers as the n810 and n800. What happens is that as an A device the musb detects an HNR attempt which times out and then the device switches "back" to host mode. I didn't expect this. It happens after a somewhat reasonable sequence of actions though.

With a POS hub I got what looked like a connection. There was an urb which looked like it had some kind of updates. When I connected a keyboard that didn't work. There was a warning from musb_host:128 about not being able to flush a host fifo so I assume that's related. This is with a power injecting cable and a heavily hacked kernel, off spec connectors etc. I probably haven't hacked things correctly, just got it so that it "looked" right. Patch is attached. I do *not* recommend flashing this kernel, instead follow the instructions for loading and booting the kernel from the wiki page w/out flashing.

I can't get the switch to host mode (bit 2 of devctl set) to happen reliably. On the other hand I have seen the usb on the n800 get screwed up such that a reboot was required, so maybe that's not a notable difference. If it is abnormal I think this is related to setting the phy resistors manually. That often times out. I suspect it could be because of some concurrent access and it gets screwed up. Perhaps the resistors don't need to be set manually, or at least don't need to be set manually where I'm doing it currently.

I'm less concerned about the unreliable connection and more concerned about the failure of this fifo to flush. I think this might be at the limit of what I can do unless there are patches out there to fix that issue.

FYI the state being set to a_idle when grounding the id pin doesn't effectively do anything. The id pin is picked up by the twl4030 and that module sets the state to a_idle if the id pin is what generates the interrupt. 'find /sys -name linkstat' to find the file which tells you the interrupt source. However this doesn't seem to be connected to the actual phy, the isp1704. If the id pin was connected there then I think bit 5 of the interrupt status register for the isp1704 should have been set to 0 when the id pin is grounded. It isn't as far as I can tell.

Re: N900 will not allow USB OTG!
 

Doing my experiments ....
Until now i put n900 in a stable state of powering the vbus few seconds.
according standard there is a time frame for connection 4.9 s . In this state ( a_vbus_vld ) n900 drive the vbus to 2.7 V ( 0.8< vbus_otg < 4.0 V ) and driver expect a_wait_bcon event. I could not imagine why bcon don't happen, leading switching the bus to reset state, instead i got unhandled disconnect interrupt
I found some limitations of devices offending last usb 20 standard and a lot of variants of usb 20 standard itself and a lot of obscure notes regarding functional implementations.
I think to write some code in a PIC microcontroller to inject a fake B_device connection signal into N900 but it is a little bit hard to switch to a real usb device hub or memstick fooling them they are already connected.

Re: N900 will not allow USB OTG!
 

Yeah dude this gonna be amazing!! I want it on my n900:p

Re: N900 will not allow USB OTG!
 

I'm confused, are you trying to really drive vbus? AFAIK the isp1704 doesn't have an integrated charge pump - bit 5 of the otg ctrl register is reaerved and that bit is drive vbus for the isp1301. I suppose I haven't tried setting bit 6 which is enabling an external charge pump but this is outside my comfort zone.

Also - what interrupts are you getting? On the n800 it's a connect interrupt which goes from wait_bcon to host. Forcing host I get a session req. interrupt which I am pretending is a connect int. - not sure if that actually works.

Re: N900 will not allow USB OTG!
 

I don't drive vbus( this is an offending standard situation ) . 2.7 volts came from n900 itself and this is a very weak voltage source .
when an A device make a connection to an B device the A device must drive vbus so N900 can drive vbus somehow and that is a good piece of news. 2.7 V can be easily obtained from 3.3 V inside of isp1707 without charge pump.
An external power source can sense the 2.7v on vbus and drive with 5V hubs, sticks, et c.

i got int 941 when doing **** things on datalines

Re: N900 will not allow USB OTG!
 

I tried reading as much of this as I could but I wasn't sure what would happen in my scenario. I own the following HDD:

http://www.hypershop.com/HyperDrive-COLORSPACE-s/23.htm

This drive has an internal battery to drive operation because of the screen and etc. There is a connector to take power from a battery or wall jack if you want extra juice to recharge while in use.

My question is I know the N900 doesn't supply power in its USB connection, so since this drive doesn't need power from the USB, can I get it to work with the N900?

I am a photographer that works on the road. With this I could send files from anywhere in the world that has cell coverage without the need for bringing a laptop. This would be HUGE. Any suggestions? Thanks.

Re: N900 will not allow USB OTG!
 

geohsia, to my knowledge nothing works yet. I will post if I have any updates. If someone could review updates to musb and otg that might help.

Re: N900 will not allow USB OTG!
 

As far as the recently leaked schematics are correct, VBUS is hooked up to GAIA TWL4030 PMU chargepump to deliver a correct 5V supply - rather than to the PHY chip.

Alas I've not found any matching node in sysfs to enable the GAIA vbus chargepump. Access to the related register via I2C should be possible nebertheless

jOERG

Re: N900 will not allow USB OTG!
 

Can you point me to schematics?

Re: N900 will not allow USB OTG!
 

Use a camera that uses SD cards. Buy a micro-SD card with an SD adapter. Put the microSD with adapter into the camera, take lots of pictures. Take the microSD out of the camera, put it into the N900. Send the pictures over the Internet.

Re: N900 will not allow USB OTG!
 

Re: N900 will not allow USB OTG!
 

joerg_rw, are we certain thats the rx51 and not the rx71? It seems rather odd to have power available and connect id to twl4030 and not support usb host.

The PHY can't supply power but it can enable an external charge pump. I hadn't seriously considered and put the probability of that doing anything rather low, that but it's easy to try.

Re: N900 will not allow USB OTG!
 

Well, all the pages read "Version: 2.0 | 07.10.2009 | N900/RX-51 | Board version: 2JH_21a "

It's quite possible the board rev of MP devices is a little different, though 2009-10-07 (if that's the correct way to read this) on schematics makes me think it isn't completely different from what shipped 6 weeks later.

Anyway for switching the port to hostmode it's not a big thing if ID is connected to PHY or anything else, and also it doesn't matter you have to enable VBUS via register config. You need to disable bat charging via USB same time you're enabling VBUS charge pump (this is no perpetuum mobile ;-). All the issues arising from OTG protocol complexity are neglegtable for a simple manually enabled host mode.

jOERG

Re: N900 will not allow USB OTG!
 

From TWL4030__TPS65950 OMAP™ Power Management and System Companion Device ES 1.0 TRM (Rev. D)__swcu050d (TI publically available docs):

(page 869)
**Table 15-40. OTG_CTRL**
Address Offset 0x0A
Description Controls UTMI+ OTG functions of the PHY
Type RW

Bit: 5
Field Name: DRVVBUS
Description:
. Signals the internal charge pump to drive 5 V on VBUS
. 0b: Do not drive VBUS.
. 1b: Drive 5 V on VBUS.
Type: RW
Reset 0

Re: N900 will not allow USB OTG!
 

Yes, the schematic has the isp170x, see ref D4380. The pinouts match for that part and the schematic's USB diagram matches my understanding where the ID pin goes to the twl4030 but not the isp170x.

The schematic answers one of my questions - how is CHGR_DET_EN_N wired? The schematic shows it as being permanently grounded. "If CHGR_DET_EN_N is LOW, and when the ISP1704A is in power down-mode, on detecting VBUS > Vth(r)(VBUS), the ISP1704A will perform automatic USB battery charger detection." The datasheet also says "When used as a host or an OTG A-device transceiver, it is recommended that the CHGR_DET_EN_N pin is deasserted." So I don't know if USB host will work as long as the PHY doesn't get powered down, or if that signal being permanently enabled means it's pretty much impossible to negotiate a connection as host.

Re: N900 will not allow USB OTG!
 

7.10.2009 could theoretically be July 10, 2009 depending on your perspective.

Have you done it already then?

Re: N900 will not allow USB OTG!
 

Looking at the properties of that PDF, I see:

Created: Mon 31 Aug 2009 14:27:59 BST
Modified: Wed 07 Oct 2009 10:59:49 BST

Re: N900 will not allow USB OTG!
 

Having not been able to see the schematic before posting that, I thought perhaps it showed the twl4030 as the phy. That's not the case.

So the twl4030 (cribbing from the tps65950 datasheet) can drive vbus using an integrated charge pump. In order to drive vbus, cp.in needs to be connected to vbat. It is. r6 (cp.gnd) is grounded which is also good. However, I don't see any indication of the CP flying capacitor being present. According to wikipedia, the source of all knowledge, a charge pump uses a capacitor to operate. AFAIK that would not be integrated with the twl4030, so the charge pump can't operate. I also don't see vbus going to any other power supply.

So the question is, why not have the charge pump? The 4 things I can think of is a) it added too much noise b) it doesn't actually work c) it can't supply enough power to meet spec d) it's stupid and tries to charge itself off its own charge pump.

Or perhaps there is no integrated charge pump on the chip they used. As I understand charge pumps are expensive and they may have decided the demand didn't justify the additional cost.

Also, even though the ID pin is connected to the twl4030, I suspect that normally this interrupt needs to be signaled through the ulpi interface (I think that's what normally happens) which the twl4030 can't do.