Thanks, I have read whole datasheet, but I wasn't sure if it's a chip-only solution, or something that require custom digitizer. Frankly, I still don't get how it works it magic as just a tiny chip, without modifying screen module, but from your explanation, I presume it hooks up itself via pins of digitizer and deliver capacitive detection "magic" through it?

Cheers and relax, some things obvious to EE may be not so clear to us non-EE people

/Estel

The operation principle is based on clever probing on the properties of the resistive touchscreen, particularly the reduced resistivity of one plane when parts of it get "shorted" via a "bypass" through the two touchpoints to the other plane. The principle is known since ~6 or more years. No capacitive technology involved.
(actually the connects "touchpoint1" and "touchpoint2" also have some resistive component depending on pressure and area/diameter of the touchpoint - I dropped that for clarity of the schematics) Think of plane1 as upper and plane2 as lower plane. On touchpoints they contact each other.

Simplified explanation:
You got 4 "wires" X-Left, X-Right, Y-Upper, Y-Lower, and you can probe 6 distinctive resistivities from that 4 terminals. I leave it up to the reader to do the permutations. 6 input variables to your clever algorithm are good for delivering a solution for X, Y, pressure - for two touchpoints. It all depends on the math you do. Some sets of input variable values may yield more than one valid solution, in that case you can't say for sure which is the right one. This clean mathematical analysis also shows why you can't do 3-touch - your max 6 input variables are not supporting any algo that yields 9 resulting values for 3 touchpoints (for 5-wire r-ts this looks different).

Hello,

Thanks for your project, I'm impatient to update my n900

I did not saw any mention of the multimedia capacity of the neo900. I know that the n900 has hardware mp3 decoder, but no ogg support (it need to be decoded in software).

Which format (audio and video) are planned in hardware for the neo900 ?

@ Chimrod,
http://talk.maemo.org/showthread.php...87&post1391987

The decoding of multimedia formats is a software task, either done on CPU or on DSP (and/or GPU).
The DSP and GPU (PowerVR) are compatible though "stronger" on the DM3730 SoC we'll use for Neo900. It depends on the OS/distro and even the particular app which use will be made of those hardware coprocessors.

cheers
jOERG

Thanks for your answer. I've read this post, but did not understand it was related with the audio decoder.

Thanks, if I understand correctly, this is the application job to use correctly the dsp.

I just wanted to be sure that earing mp3 collection throught the fm transmitter will not empty my batterie in one hour.

http://tuomas.kulve.fi/blog/2009/11/...on-ogg-vs-mp3/ might be interesting for you

cheers
jOERG

Actualy very interesting.

I really appreciate the reactivity and the support you give here. This give to the project a very good direction

Sorry if asked before, but I couldn't find a response:

There were some talks about "how sacurely can we turn modem off", but, for change I would like some insight on different approach for the matter - how much info, realistically in selected modem(s), could we query from modem hardware?

I mean that, in most old "dumbphones" from Nokia, there was a "hidden" netmon program, that was giving accurate information about transmission details, including even exact channel that we're transmitting at. It was often used by cellular network enthusiasts to select least occupied (by transfer) base stations on neighbourhood (even if not the closest one), point their antennas toward it, and "force" connection to selected BS, by enabling only selected band *and* channel in their modems. This, practically, allowed to semi-goaround the limitation, that only base stations choose, where client device connects.
---

Now, I imagine, that netmon was directly query'ing modem via some AT commands equivalent. In N900, hoever, or AT queries are handled by middle-man (AIUI), allowing only for very limited set of even basic queries, leaving "forcing" anything (like channel) totally out of question. Heck, we're even struggling to get (real-life) readings of band we're using in practice (getting only list of bands that we *could* be theoretically using).
---

Now, the actual question - how it will look on Neo900? Does that "complete control over modem" mean we will get all info *and* be able to force some connection parameters, like with stand-alone cellular modems for PCs?

/Estel