Profile

Cover photo
Rod Whitby
916 followers|1,071,289 views
AboutPostsVideos+1'sReviews

Stream

Rod Whitby

Shared publicly  - 
 
Working on an even better mounting method for the charging circuit board, including pogo pins between the backplate and the main circuit board. See the captions on the latest photos in the album for details. 
6
Bill Huey's profile photoRod Whitby's profile photo
2 comments
 
Improved mounting scheme is successful.  Everything fits perfectly.  Second Nexus 4 with Touchstone charging completed and working.
Add a comment...

Rod Whitby

Shared publicly  - 
 
Nexus 4 Touchstone Mod (Part 4)
[ Start at Part 1: http://goo.gl/r7ATZ ]

The surgery begins ...

See the captions on the attached photos for more information.

[ Continue at Part 5: http://goo.gl/RF8LL ]

-- Rod
11
Add a comment...

Rod Whitby

Shared publicly  - 
 
Nexus 4 Touchstone Mod (Part 2)
[ Start at Part 1: http://goo.gl/r7ATZ ]

The next experiment needed before starting any non-reversible modifications is to determine whether the touchstone coil will actually work when located in the Nexus 4 back.

The result of that experiment is surprisingly positive.  It seems that the Qi charging coil in the backplate does not interfere with the operation of the Palm touchstone charging coil.  I was fully expecting to have to remove the Qi coil (hopefully in a replaceable fashion, and without damaging the outer NFC coil) to make this work, but now it seems this is not required.

Now attention turns to the placement of the charging circuit board.  I first looked at the speaker/antenna enclosure at the bottom of the phone (having found those to often contain empty space within the plastic), but there was not joy to be found there.

There does seem to be some room just north of the vibrator motor (in the same vertical plane as the SIM card slot).  It's possible that if a charging board sized section of the plastic shield in that area is removed, then there might just be enough space.  It looks doable in the horizontal plane, it'll be the height of the charging board compared to the thickness of the plastic section removed which will determine success on this step ...

See the captions on the attached photos for more information.

[ Continue at Part 3: http://goo.gl/uZLmZ ]

-- Rod
3
Add a comment...

Rod Whitby

Shared publicly  - 
 
Four Multi-Charging (Phone and Tablet) Touchstones

These modified touchstones are able to charge my Palm Pre, Palm Pre+, Palm Pixi, Palm Pixi+, Palm Pre 2, HP Veer, HP Pre 3, HP TouchPad, HP TouchPad Go, HP TouchPad 4G, HP WindsorNot, Samsung Nexus S, Samsung Galaxy Nexus, Nokia Lumia 800, and Asus Nexus 7.

See http://goo.gl/VgBha for details of how to build them.

-- Rod
24
5
James Sullins's profile photoNikita Bunovskyi's profile photoMarco Bresch's profile photoKeith Chen's profile photo
6 comments
 
Add a comment...

Rod Whitby

Shared publicly  - 
 
Nexus 7 Touchstone Mod (Part 12)
[ Start at Part 1: http://goo.gl/2gxd5 ]

The final result in pictures.

Nexus 7 Touchstone Mod completed successfully.

[ Continue at Part 13: http://goo.gl/VgBha ]

-- Rod
31
5
Jayson Tanasescu's profile photoArturo Araoz's profile photoRobson Porto's profile photoMichael Pedder's profile photo
4 comments
 
Just a question why did you not use a touchpad charger?
If the answer is because you cannot buy extra inductive charging covers for it, just forget the question.
Add a comment...

Rod Whitby

Shared publicly  - 
 
Nexus 7 Touchstone Mod (Part 9)
[ Start at Part 1: http://goo.gl/2gxd5 ]

I've finally worked out the best placement for the charging circuitry, and a good wiring layout for the coils.

Next step is to glue the coils and magnets onto the inside of the back cover, complete the soldering, and then put it all back together again.

[ Continue at Part 10: http://goo.gl/cIecx ]

-- Rod
4
Kyle Peterson's profile photoRod Whitby's profile photo
8 comments
Add a comment...

Rod Whitby

Shared publicly  - 
 
Fixing ClockworkMod Recovery on the Nexus Q

I got a Nexus Q in the post the other week, and found a binary of ClockworkMod Recovery that had been released (with no source code or build instructions - hmm) at http://rootzwiki.com/topic/28810-beta-cwm/ on RootzWiki.

Unfortunately, the screen output of that binary was all squished up in the top left corner.  Not very useable.

So I set out to fix it.

First, since there was no source code and no build instructions accompanying the binary, I had to work out how to build ClockworkMod Recovery from source.  That's not too hard - it's simply building the "recoveryimage" target in CyanogenMod.

However, you need a kernel and board description file for the target device.  Google (as is required by the GPL) released the source code for the kernel, and I have put it up on GitHub at https://github.com/NexusInternals/android_kernel_google_steelhead for easy access.

Google also seems to have been working on an AOSP build for the Nexus Q, and I have put this work up on GitHub at https://github.com/NexusInternals/android_device_google_phantasm as well, and used this as the basis for the "phantasm" device.

By adding some integration files for the CyanogenMod build system (https://github.com/NexusInternals/android_device_google_phantasm/commit/71fd86541614b171113c09a8882417e9b374ae3f) and correcting the /sdcard entry in the recovery.fstab file (https://github.com/NexusInternals/android_device_google_phantasm/commit/141287be48120e6327fac6f2fb7292b35e1add7f), I was able to get a build which booted on the Nexus Q and seems to be fully functional (apart from the squished display).

I have recorded the build steps at http://www.nexusinternals.org/wiki/Build_ClockworkMod_for_Nexus_Q for any other Nexus Q developers out there (I wonder if there actually are any other people still developing for the Nexus Q).

Next was the question of how to diagnose and fix this squished screen output problem.

The ClockworkMod recovery source code (CM9/bootable/recovery/minui/graphics.c) uses /dev/graphics/fb0 for it's output. On the Nexus Q, /dev/graphics is controlled by the omapfb kernel driver, which in turn uses the omapdss graphics driver.

I found some information about the OMAP Display Subsystem (DSS) in the kernel documentation at https://github.com/NexusInternals/android_kernel_google_steelhead/blob/ics/Documentation/arm/OMAP/DSS and in an Embedded Linux Conference presentation at http://elinux.org/images/8/83/Elc2011_semwal.pdf (this presentation describes both the hardware architecture and the software driver architecture).

Looking around in the /sys/devices/platform/omapdss filesystem on the Nexus Q, I found that the input_size parameter for the enabled overlay was reporting the correct "1920,1080" resolution of my monitor, but the output_size parameter was reporting "640,480", which matched the squished screen resolution I was seeing.  So I determined that the ClockworkMod Recovery graphics code and the omapfb drivers were doing the right thing, but there was incorrect scaling down from 1920x1080 to 640x480 happening in the overlay.

By turning on CONFIG_OMAP2_DSS_DEBUG_SUPPORT in the kernel, and forcing the kernel module debug parameter, I was then able to get debug messages from the DSS from dmesg after boot.  Here's the smoking gun:

 omapdss: no timings yet, returning some safe defaults
 omapdss OVERLAY: check_overlay 1: (0,0 1920x1080 -> 640x480) disp (640x480)

The HDMI panel driver in the kernel was returning 640x480 as the screen resolution when there were no timings available from the panel (i.e. the EDID had not yet been read).  Later in the boot log (after the EDID was read), I saw the following:

 omapdss OVERLAY: check_overlay 1: (0,0 1920x1080 -> 640x480) disp (1920x1080)

It seemed that even though the correct panel resolution was now known, then dummy 640x480 resolution was "stuck" in the overlay out_width and out_height parameters.

https://github.com/NexusInternals/android_kernel_google_steelhead/commit/c9b4c391018a34753d3ad9aa80092e94cf7200f0 is the fix I devised for this problem.

After applying this fix to the kernel, and recompiling the recovery image, I was pleased to see a full-size screen output.  It's fairly important to be able to read the text on the screen when the commands are things like "wipe data/factory reset" and "Yes -- delete all user data" ..

If you're a Nexus Q developer, and need any assistance in compiling this version, please don't hesitate to contact me.  After I've tested it some more, I'll put a binary (with pointers and source code and build instructions of course) in the original thread on RootzWiki.

-- Rod
17
5
Jason Parker's profile photoJames Sullins's profile photoMathias Djärv's profile photoJohn Triolo's profile photo
18 comments
 
There is a newer version of CWM recovery available as well, 6.0.3.1. I've installed it successfully on my Q. One thing I learned was my 2408WFP Dell, while it has an HDMI port, absolutely SUCKS for displaying certain types of content, it basically refused to see the CWM video, while moving into the living room and hooking to the LG TV, worked perfectly the first time.
Add a comment...
Have him in circles
916 people
Byron Todd's profile photo
Anthony Winter's profile photo
Chris Mason's profile photo
ryan webster's profile photo
Corey Scienza's profile photo
Synodroid's profile photo
Luciano Rozza's profile photo
Zenón Monge's profile photo
Christian Reimann's profile photo

Rod Whitby

Shared publicly  - 
 
Nexus 4 Touchstone Mod (Part 5)
[ Start at Part 1: http://goo.gl/r7ATZ ]

You're looking at the world's first multi-wireless-charging phone, a Nexus 4 that has been modified to support both Qi and Palm Touchstone wireless charging technologies.

Everything is back together, and fully operational (including NFC).  This Nexus 4 phone can now charge on both Qi wireless charging pads and Palm Touchstones, without swapping charging backs or any other inconveniences like that.

Interestingly, the Palm Touchstone seems to charge the Nexus 4 significantly faster than the Qi charger.  I will have to test it on an official Google Orb (not available locally in Australia) to see if that is a limitation of the Qi charger I have, or a limitation of the particular Qi charging circuitry in the Nexus 4.

When placed on the Qi charging pad, the phone shows "Charging (wireless)".  When placed on the Palm Touchstone, the phone shows "Charging (USB)" (because I hooked the Palm Touchstone charging circuit board output internally to the same place to which the USB port VBUS power signal is connected).

Next step is to determine whether I should add magnets inside the back of the phone (like I did on the Galaxy Nexus), or just metal discs.  I will need to test whether the magnetometer in the Nexus 4 is able to automatically compensate (after recalibration) by subtracting the constant magnetic offset caused by the fixed magnets.  Some clipping of the sensor output will obviously occur, but I charge my phone two orders of magnitude more often than I use the compass, so it's a worthwhile trade-off for guaranteed overnight charging without the device slipping off the charging pad.

Guided alignment and guaranteed stability with the magnets in the Palm Touchstone is the key feature that is missing from all current Qi chargers, and is the main thing that makes Qi wireless charging inferior (in my opinion) to using Palm Touchstones.  It seems the Palm Touchstone also has the upper hand in charging time as well (compared to my current Qi charger).

-- Rod
26
4
Massimiliano D'Ambrosio's profile photoSam Burney's profile photoHelmut Fuchs's profile photoAlex Siuda's profile photo
60 comments
 
+Andrew Gordon Thanks for the link.
Add a comment...

Rod Whitby

Shared publicly  - 
 
Nexus 4 Touchstone Mod (Part 3)
[ Start at Part 1: http://goo.gl/r7ATZ ]

As expected, clearance of the charging board in the vertical dimension is the killer here.

Not sure how I'm going to solve this one ...

The good news is that the Qi charging still works (electrically speaking, my issue with Qi charging is the lack of physical alignment support on current devices, not the electrical protocol itself) even with the Palm Touchstone coil (and it's Mu-Metal shielding) in place and the cover put back on the phone.  This also confirms that the additional height of the Palm Touchstone coil (as opposed to the charging board) is not a vertical clearance issue.

This may well end up as the only device in existence which is both Palm Touchstone and Qi charging compatible.

See the captions on the attached photos for more information.

[ Continue at Part 4: http://goo.gl/ObRQv ]

-- Rod
9
Josh Hurt's profile photoRod Whitby's profile photoKhurram Khan's profile photoJ Miller's profile photo
4 comments
 
+Khurram Khan If my GNex is hot, it will not charge.
Add a comment...

Rod Whitby

Shared publicly  - 
 
Nexus 4 Touchstone Mod (Part 1)

I've decided to add Palm Touchstone charging to the Nexus 4. Qi charging (or at least the physical implementation of it on current devices) is (IMHO) inferior to Palm Touchstone charging. +Derek Kessler seems to agree (http://www.androidcentral.com/i-hate-nexus-4-wireless-charger).

The fact that I have 5 or 6 Palm Touchstones already sitting around my home, car and office, and all the other phones and tablets (see the list at the end of this post) my family uses have already been modified to charge on the Palm Touchstones probably has some influence on my motivations.

I've already done the experiment to ensure that the coil and charging circuit board from Palm Touchstone back has enough voltage and current capacity to charge the Nexus 4.  Only one coil is required.

I have also found a document on the Internet which has enough of the Nexus 4 circuit diagram and PCB component placement details to determine where to inject the 5V into the Nexus 4 (as usual, it's easiest to inject at a point which is electrically equivalent to the standard USB port power pins). Excerpts from that document are attached.

So the big problem is where to situate the small Palm Pre charging board.

There's not a whole lot of spare space inside the Nexus 4, so I'm going to need to get creative ...

[ Continue at Part 2: http://goo.gl/y3WLm ]

-- Rod

Related Links
Nexus S Touchstone Mod: http://www.flickr.com/photos/rwhitby/sets/72157628557489531/
Lumia 800 Touchstone Mod: http://twitpic.com/9jevyp
Galaxy Nexus Touchstone Mod: http://forums.androidcentral.com/verizon-galaxy-nexus-accessories/147840-guide-galaxy-nexus-mod-palm-touchstone-no-soldering-phone.html
Nexus 7 Touchstone Mod: http://goo.gl/2gxd5 
Multi-Charging (Phone and Tablet) Touchstones: http://goo.gl/9c2SL
Other Touchstone capable devices: http://goo.gl/hNyBB
37
5
Björn Adelberg's profile photoAlexandru Tudor's profile photoAhmad Kayyali's profile photoRick Perez's profile photo
22 comments
 
Yep,  I think I may have better luck if I reroute the Touchstone charging circuit to the USB power on the phone instead of connecting to the QI input.  That should cut the current demand to the standard USB maximum charge rate of 0.5A.
Add a comment...

Rod Whitby

Shared publicly  - 
 
Nexus 7 Touchstone Mod (Part 13)
[ Start at Part 1: http://goo.gl/2gxd5 ]

Sitting the Nexus 7 directly on the Palm touchstone with no support is not that stable (touching notifications at the top of the screen causes it to tip).

So I came up with a better system.

Now the same HP TouchPad charging dock can charge both the TouchPad and the Nexus 7 (and phones for that matter too).

-- Rod
17
3
Richard Frost's profile photoChan Qian's profile photoRobson Porto's profile photoJames Sullins's profile photo
 
Hey +Rod Whitby welcome to Google Plus.. !! Good to see you here now ;)  Love the Nexus 7.
Add a comment...

Rod Whitby

Shared publicly  - 
 
Nexus 7 Touchstone Mod (Part 11)
[ Start at Part 1: http://goo.gl/2gxd5 ]

I added electrical tape to cover the connections to the coil, and set up for the final soldering of the 5V output from the charging circuitry to the motherboard.

The soldering points are the main GND plane and the main VUSB plane (5V from the USB connector), both of which can be easily exposed near CON 1.  You can test (using a multimeter) that you have the correct points by ensuring that they are connected to 5V and GND on the micro-usb connector.

I scraped the tracks to expose copper, carefully covered the surrounding components, and did the final deed.  This bit is the dangerous part - you need to make sure you don't overheat the motherboard, but those copper planes will suck up the heat and you need to get them hot enough for the solder to make a good connection.

Next is the final assembly and test ...

[ Continue at Part 12: http://goo.gl/KMy8n ]

-- Rod
5
Add a comment...
People
Have him in circles
916 people
Byron Todd's profile photo
Anthony Winter's profile photo
Chris Mason's profile photo
ryan webster's profile photo
Corey Scienza's profile photo
Synodroid's profile photo
Luciano Rozza's profile photo
Zenón Monge's profile photo
Christian Reimann's profile photo
Links
YouTube
Contributor to
Basic Information
Gender
Male
Rod Whitby's +1's are the things they like, agree with, or want to recommend.
Ski Safari
market.android.com

Ride penguins, yetis and more in an endless downhill avalanche escape!★ ★ ★ ★ ★ "Great! - Love everything about this game. The graphics are

ExpenseClam (Free)
market.android.com

Make sure you get fully reimbursed for all your business expenses!Tracker for expenses, mileage and claims.Record Expenses Quickly. The home