Some time ago (October 2009) I replaced my trusty 1st gen Nokia Linux handset with a 3rd gen handset due to a cracked/illegible screen (don't keep them in your back pocket!). The old device still functioned but with no visual feedback it was essentially impossible to interact with.
However, there was a set of factory test pins under the battery cover which I figured could be used to read out what the device was doing and probably issue commands in turn. With it I could easily reconfigure its wifi connection, connect any number of USB accessories, and essentially transform the old broken handset into a headless TiVo-type media/backup device for the home network. But lacking the correct cable, the broken 1st gen just went back in the box. Waiting for a long weekend for me to tamper with it.
That weekend came this last Febuary when I saw some $0.50 picture frames at the local Goodwill.
I'd remembered this fellow's set up for home-manufacturing aluminum high-fi speakers, and figured I could similarly use fluid adhesion/cohesion to hold household aluminum foil in place against a sheet of glass & cut out very small, thin conductors in sufficient length to reach underneath the Nokia 770 battery cover.
Warm soapy water (surfactant) & a $1 rubber squeegee from Goodwill are used to adhere the foil to the glass for cutting. Alcohol is used to clean the soap from the foil.
Specialized tool kits for accessing FBUS test pads run about $50 from overseas outfits, with an additional $20 or so for shipping due to the size/quantity of the test probes they come with. For $40 I could buy an entire replacement nokia 770 or HTC g-1, considering how old these handsets were.
First attempt at cutting strips, 2nd pane of glass as a guide.
After cutting them free, the strips were picked up off the glass by adhering a strip of packing tape across them. A "cable" was fashioned by cutting the tape backing to size and attaching a second smaller strip to the opposite face, enclosing the conductors but leaving aprx 1mm square exposed on the ends, each.
The 'proof of concept' cable demonstrated that smoothing the foil out perfectly flat is actually necessary - not only to hold the foil still while cutting, but because even a very small wrinkle, and the associated slack as soon as you cut it free, really fouls up your attempts to cut even strips.
Conductivity/resistance was suitable the conductor spacing was close to the target (4x 1mm strips, 0.5mm gaps), but far enough off that I opted to make cable v0.2 out of four independently wrapped strips, rather then trying to tape up all four at once.
Unfortunately electrical contact was hard to maintain, even with a small cube of cardstock pushed into the slot to hold the conductors in contact with the Fbus pads. My friend Mike at this point suggested I just solder them directly to the handset. Unfortunately there really was nowhere adequate (with the exception of Pin J1111 which branched directly off the battery connector) to attach a heatsink here. Without a heatsink attached nearby to the trace - by the time it was brought up to to temperature for soldering a good connection - the heat would likely destroy the delicate electronics on the other side of the trace ie: inside the handset.
Met with that defeat, Strategy B was to make reliable electrical contact - so I could get on with verifying if these FBUS contacts even functioned like my second hand Nokia service schematic indicated they should. A weight (the needlenose) was used to hold four rigid/sharp pins (snipped from a few paper staples) to the pads so I could get on with life & test 'em.
With Pin 3 as ground, Pin J1111 was able to supply +3.3v directly from the battery, suitable to power an old Nokia CA-42 data transfer cable and make it show up on a PC as a USB device. Pin 1 did display a voltage drop indicative of serial data output being generated from the Nokia handset's startup procedure, but even simplifying Strategy B by powering the CA-42 data transfer cable directly with a pair of AA batteries - the setup was still unable to consistently hold the necessary 2 pins in contact.
--
All told about $4 dollars in supplies and about 8 or 9 hours invested in what I would call a partial success. Not bad for a first attempt at fabricating thin-film conductors or messing with an Fbus "connector".
Perhaps some day my dream of a sexy, superthin cable capable of plugging into the data port while the battery case is on will be realized. But for now it seems like the only (stable) way to go is going to involve fashioning a jig or frame that will hold some fairly substantial pins in contact very firmly.
No comments:
Post a Comment