From johan.soderberg at sts.gu.se Mon Sep 1 10:01:24 2008 From: johan.soderberg at sts.gu.se (=?ISO-8859-1?Q?Johan_S=F6derberg?=) Date: Mon, 1 Sep 2008 11:01:24 +0200 (CEST) Subject: [Ronja] research on Ronja Message-ID: <19381742.1220259684217.JavaMail.oracle@midtier-vas-2.it.gu.se> Hi, My name is Johan S?derberg and I'm a PhD researcher at the Science & Technology Studies department at G?teborg University, Sweden. Im writing my thesis on hardware hacking and Ronja will be one of my case studies. I have previously written briefly about Ronja in a book I published this spring (it is in swedish, unfortunately, but available here: http://www.bokforlagetatlas.se/index.php?sid=2&pid=16&showtitle=332), and now I want to do more in-depth research about the project. The reason for my interest in Ronja is that it is one of the very few hardware hacking projects Im aware of that has gone from a good idea or a design to an actually working technology, and been sustained for a number of years. In my thesis Im exploring the possibility that advanced users can take a lead in hardware development in the same way they have in free software development. Im curious about how Ronja is developed, how it is used, where it has been disseminated, who supports it, etc. I have already been in touch with Karel. Currently Im in Prague and I will be staying here for the whole autumn. I would like to talk to everyone even remotely associated with the project. Developers, users, supporters, lurkers... (also those of you not living in czech republic). If someone is about to build a Ronja sometime soon I would greatly appreciate to be part of it and see work-in-progress (Mluvim trochu cesky). My mail adress is johan.soderberg at sts.gu.se thanks Johan? ? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://pointless.net/pipermail/ronja/attachments/20080901/96fef2b3/attachment.html From twibright at hispeed.ch Mon Sep 1 15:18:21 2008 From: twibright at hispeed.ch (Karel Kulhavy) Date: Mon, 1 Sep 2008 16:18:21 +0200 Subject: [Ronja] Bug in the Ronja author Message-ID: <3429199.1220278702107.JavaMail.root@viefep12> The Ronja project, apart from the design, involves also it's author(s). Therefore it makes sense to ask not only about bugs in the design, but also in the designers. Recently I have discovered a bug in myself called Obsessive Compulsive Personality Disorder: http://en.wikipedia.org/wiki/Ocpd The symptoms are: *1* Preoccupation with details, rules, lists, order, organization, or schedules to the extent that the major point of the activity is lost *2* Showing perfectionism that interferes with task completion (e.g., is unable to complete a project because his or her own overly strict standards are not met) *3* Excessive devotion to work and productivity to the exclusion of leisure activities and friendships (not accounted for by obvious economic necessity) *4* Being overconscientious, scrupulous, and inflexible about matters of morality, ethics, or values (not accounted for by cultural or religious identification) *5* Inability to discard worn-out or worthless objects even when they have no sentimental value *6* Reluctance to delegate tasks or to work with others unless they submit to exactly his or her way of doing things *7* Adopting a miserly spending style toward both self and others; money is viewed as something to be hoarded for future catastrophes *8* Shows rigidity and stubbornness I think Ronja has been definitely affected by 2,6,7,8 and maybe by 1 and 4. 3 was beneficial for Ronja, without 3, Ronja wouldn't get as far as it did. Now when the bug has been found it's time to fix it. OCPD is the easiest to fix personality disorder and can be fixed by DIY. I wonder about your comments and feedback regarding effects of OCPD on Ronja project you have observed and how you would suggest to fix them. Cheers, Karel From antitron at web.de Tue Sep 2 12:18:15 2008 From: antitron at web.de (Thomas Egenhofer) Date: Tue, 02 Sep 2008 13:18:15 +0200 Subject: [Ronja] Bug in the Ronja author In-Reply-To: References: Message-ID: <1220354295.6061.2.camel@Nekomimi-host> > Recently I have discovered a bug in myself called Obsessive Compulsive > Personality Disorder: > http://en.wikipedia.org/wiki/Ocpd > HELP! i'm fully affected by this bug too! if there is a known bugfix let me know. greetings, thomas From discover at fly.srk.fer.hr Tue Sep 2 15:36:32 2008 From: discover at fly.srk.fer.hr (Silvije) Date: Tue, 2 Sep 2008 16:36:32 +0200 (CEST) Subject: [Ronja] Bug in the Ronja author Message-ID: Clock, lot of this things everybody could recognize in themself, me also. Although to some let's say acceptable extent, it does not have to be bad thing. Just as you said, without 3) project would not be where it is. Your, let's say "inflexibility" about RONJA design, gave me a push to find my own solutions, try to improve parts of design I thought could be better, learn some new stuff and so on. So again it was not totaly a bad thing, for me personaly it was maybe frustrating but good thing in the end :) And I am not the only case, lot of people developed something based on your RONJA design. Is it better or not that is for discussion of course. However, I wish to thank you! If nothing else your idea saved me great amount of money (that I did not have, and sadly still dont have) I could pay to local ISP for link that was established at begining with RONJA and after that with RONJA based (and some other people ideas based) device. Thank you all, and keep up the good work! From kubajz at kbx.cz Tue Sep 2 20:22:27 2008 From: kubajz at kbx.cz (Jakub Sykora) Date: Tue, 02 Sep 2008 21:22:27 +0200 Subject: [Ronja] Bug in the Ronja author In-Reply-To: <1220354295.6061.2.camel@Nekomimi-host> References: <1220354295.6061.2.camel@Nekomimi-host> Message-ID: <48BD9273.7000201@kbx.cz> Is there any open source involved person, who is not affected? I think it goes hand in hand... K Thomas Egenhofer wrote: >> Recently I have discovered a bug in myself called Obsessive Compulsive >> Personality Disorder: >> http://en.wikipedia.org/wiki/Ocpd >> > HELP! i'm fully affected by this bug too! > if there is a known bugfix let me know. > > greetings, > thomas > > -- Jakub S?kora email: kubajz na kbx.cz <') ICQ: 68976632 ( =- mobil: +420 777 594 201 '' From ronjalist at vitalit.co.uk Sat Sep 6 20:54:48 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Sat, 6 Sep 2008 20:54:48 +0100 Subject: [Ronja] TX Head design question... Message-ID: <48C2E008.5080101@vitalit.co.uk> Hi All, Just a quick question - Is there any reason why instead of forcing the TX led to produce the pulses, a "shutter" cannot be employed instead ? By "shutter" I mean a piece of high speed lcd or similar placed between the led and the optics ? Like we see in ship to ship visual communications in old films. Thanks in advance. Matt From jdb at lartmaker.nl Sat Sep 6 21:10:09 2008 From: jdb at lartmaker.nl (J.D. Bakker) Date: Sat, 6 Sep 2008 22:10:09 +0200 Subject: [Ronja] TX Head design question... In-Reply-To: <48C2E008.5080101@vitalit.co.uk> References: <48C2E008.5080101@vitalit.co.uk> Message-ID: >Hi All, > Just a quick question - Is there any reason why instead of forcing >the TX led to produce the pulses, a "shutter" cannot be employed instead ? No fundamental reason, just practical ones. Sourcing shutter film/foil which can reliably switch at 10MHz is likely challenging, especially in the small quantities that the average Ronja builder would need. Common LCD shutters max out at 100Hz-1kHz, so we'd need four to five orders of magnitude better performance. If the shutter were placed between LED and lens viewing angle and diffraction must be accounted for, if the shutter is placed after the lens it needs to be physically large (and diffraction is still an issue). More here: http://www.liquidcrystaltechnologies.com/tech_support/LCDShutterConsiderations.htm Have you any suggestions for such a shutter? JDB. -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lartmaker.nl/ From asteri_x at freemail.hu Sat Sep 6 23:20:29 2008 From: asteri_x at freemail.hu (=?UTF-8?B?R3l1cmvDsyBNYXJ0aW4=?=) Date: Sun, 07 Sep 2008 00:20:29 +0200 Subject: [Ronja] TX Head design question... In-Reply-To: References: <48C2E008.5080101@vitalit.co.uk> Message-ID: <48C3022D.3000306@freemail.hu> Hi! in the early years, when Laser printers were born they had a quarz crystal, in which acustic wawes were induced. these wawes acted as a diffraction pattern to the laser, and this diffraction was steering the laser on different points on the paper, or was switching the laser when the LD was too slow. This is only working cool if the light is monocromatic. LEDs are unfortunately not. Also these crystals are extremely expensive. But if you are good, then you can use a DVD, focus a beam on it, and steer the reflected light to the TX lens. This would work like a shutter :D But you would need servo control of the head, which is not that easy. bye, Martin J.D. Bakker ?rta: >> Hi All, >> Just a quick question - Is there any reason why instead of forcing >> the TX led to produce the pulses, a "shutter" cannot be employed instead ? > > No fundamental reason, just practical ones. Sourcing shutter > film/foil which can reliably switch at 10MHz is likely challenging, > especially in the small quantities that the average Ronja builder > would need. Common LCD shutters max out at 100Hz-1kHz, so we'd need > four to five orders of magnitude better performance. If the shutter > were placed between LED and lens viewing angle and diffraction must > be accounted for, if the shutter is placed after the lens it needs to > be physically large (and diffraction is still an issue). > > More here: > > http://www.liquidcrystaltechnologies.com/tech_support/LCDShutterConsiderations.htm > > Have you any suggestions for such a shutter? > > JDB. From ronjalist at vitalit.co.uk Sun Sep 7 01:08:06 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Sun, 7 Sep 2008 01:08:06 +0100 Subject: [Ronja] TX Head design question... In-Reply-To: <"p062405b8c4e89245c570(a)(091)130.161.115.44(093)*"@MHS> References: <48C2E008.5080101@vitalit.co.uk> Message-ID: <48C31B66.9030205@vitalit.co.uk> I honestly had no firm idea, but from a scrap & salvage point of view - perhaps after the optics, and wouldn't a 15 inch tft from a scrap laptop or similar do? J.D. Bakker wrote: >> Hi All, >> Just a quick question - Is there any reason why instead of forcing >> the TX led to produce the pulses, a "shutter" cannot be employed instead ? > > No fundamental reason, just practical ones. Sourcing shutter > film/foil which can reliably switch at 10MHz is likely challenging, > especially in the small quantities that the average Ronja builder > would need. Common LCD shutters max out at 100Hz-1kHz, so we'd need > four to five orders of magnitude better performance. If the shutter > were placed between LED and lens viewing angle and diffraction must > be accounted for, if the shutter is placed after the lens it needs to > be physically large (and diffraction is still an issue). > > More here: > > http://www.liquidcrystaltechnologies.com/tech_support/LCDShutterConsiderations.htm > > Have you any suggestions for such a shutter? > > JDB. From jdb at lartmaker.nl Sun Sep 7 01:25:06 2008 From: jdb at lartmaker.nl (J.D. Bakker) Date: Sun, 7 Sep 2008 02:25:06 +0200 Subject: [Ronja] TX Head design question... In-Reply-To: <48C31B66.9030205@vitalit.co.uk> References: <48C2E008.5080101@vitalit.co.uk> <48C31B66.9030205@vitalit.co.uk> Message-ID: At 01:08 +0100 07-09-2008, ronjalist wrote: >I honestly had no firm idea, but from a scrap & salvage point of view - >perhaps after the optics, and wouldn't a 15 inch tft from a scrap laptop >or similar do? -> >J.D. Bakker wrote: > > [...] Common LCD shutters max out at 100Hz-1kHz [...] JDB. -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lartmaker.nl/ From ronjalist at vitalit.co.uk Sun Sep 7 01:28:24 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Sun, 7 Sep 2008 01:28:24 +0100 Subject: [Ronja] TX Head design question... In-Reply-To: <"p062405b9c4e8cfa82895(a)(091)130.161.115.44(093)*"@MHS> References: <48C2E008.5080101@vitalit.co.uk> Message-ID: <48C32028.6060901@vitalit.co.uk> Err since this is not a shutter by design and is a monitor - I'd be surprised if you could display vga with those timings ? J.D. Bakker wrote: > At 01:08 +0100 07-09-2008, ronjalist wrote: >> I honestly had no firm idea, but from a scrap & salvage point of view - >> perhaps after the optics, and wouldn't a 15 inch tft from a scrap laptop >> or similar do? > > -> > >> J.D. Bakker wrote: >> > [...] Common LCD shutters max out at 100Hz-1kHz [...] > > JDB. From ronjalist at vitalit.co.uk Sun Sep 7 01:33:33 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Sun, 7 Sep 2008 01:33:33 +0100 Subject: [Ronja] TX Head design question... In-Reply-To: <48C32028.6060901@vitalit.co.uk> References: <48C2E008.5080101@vitalit.co.uk> Message-ID: <48C3215D.8000803@vitalit.co.uk> Scrap that - I apologise I was thinking in Mhz but reading Hz - that and a quick check of my monitors refresh rate... Sorry ronjalist wrote: > Err since this is not a shutter by design and is a monitor - I'd be > surprised if you could display vga with those timings ? > > J.D. Bakker wrote: >> At 01:08 +0100 07-09-2008, ronjalist wrote: >>> I honestly had no firm idea, but from a scrap & salvage point of view - >>> perhaps after the optics, and wouldn't a 15 inch tft from a scrap laptop >>> or similar do? >> -> >> >>> J.D. Bakker wrote: >>> > [...] Common LCD shutters max out at 100Hz-1kHz [...] >> JDB. > From gmaxwell at gmail.com Sun Sep 7 01:51:07 2008 From: gmaxwell at gmail.com (Gregory Maxwell) Date: Sat, 6 Sep 2008 20:51:07 -0400 Subject: [Ronja] TX Head design question... In-Reply-To: <48C3022D.3000306@freemail.hu> References: <48C2E008.5080101@vitalit.co.uk> <48C3022D.3000306@freemail.hu> Message-ID: On Sat, Sep 6, 2008 at 6:20 PM, Gyurk? Martin wrote: > Hi! > > in the early years, when Laser printers were born they had a quarz > crystal, in which acustic wawes were induced. these wawes acted as a > diffraction pattern to the laser, and this diffraction was steering the > laser on different points on the paper, or was switching the laser when > the LD was too slow. > This is only working cool if the light is monocromatic. > LEDs are unfortunately not. > Also these crystals are extremely expensive. AOM - Acoustic Optical Modulator. And yes, you can get ones that operate at 10MHz. But they are expensive, and need to be used with narrow band light. They were, at one time, used for communications too. Not really realistic for RONJA. From ronjalist at vitalit.co.uk Sun Sep 7 02:11:14 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Sun, 7 Sep 2008 02:11:14 +0100 Subject: [Ronja] Further lcd thoughs... Message-ID: <48C32A32.3090601@vitalit.co.uk> OK 50Hz is not enough, but you do have 1024x768 (or better) resolution and you could forgo the main optics, use a fresnel lens and subdivide the panel into 1000 segments... With a telescopic lens arrangement and an lcd webcam as the rx. Thoughts ? From jdb at lartmaker.nl Sun Sep 7 02:45:20 2008 From: jdb at lartmaker.nl (J.D. Bakker) Date: Sun, 7 Sep 2008 03:45:20 +0200 Subject: [Ronja] Further lcd thoughs... In-Reply-To: <48C32A32.3090601@vitalit.co.uk> References: <48C32A32.3090601@vitalit.co.uk> Message-ID: At 02:11 +0100 07-09-2008, ronjalist wrote: >OK 50Hz is not enough, but you do have 1024x768 (or better) resolution >and you could forgo the main optics, use a fresnel lens and subdivide >the panel into 1000 segments... > >With a telescopic lens arrangement and an lcd webcam as the rx. > >Thoughts ? It'll work, but at the expense of link range. Assuming equal transmit power, and comparing two systems with comparable optics (no fair giving a telescope to the webcam and not to the regular Ronja), each of those 1000 segments will have 1/1000th of the total power. Achievable link distance is roughly proportional to the square root of power, so those 1000 segments reduce the max link distance by a factor of ~30, to just over 40 meter for a standard Ronja config. Now, in the real world, - LCD panels transmit less than 50% of all incident light - on top of that, a significant portion of the light (>60%) will be diffused/diffracted off-axis. This leaves only 10-20% of all transmitted energy to hit the receiver - you *will* need some separation between segments, or accept that light from one segment will bleed into the next one. This, too, costs TX energy. - alignment is going to be harder + on longer ranges, in-air absorption will eat more energy than the square law suggests (so this would work in favor of a short-range segmented system) More on link budgets here: http://www.pointless.net/pipermail/ronja/2006-August/009130.html JD 'no free lunch' B. -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lartmaker.nl/ From ronjalist at vitalit.co.uk Sun Sep 7 03:42:08 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Sun, 7 Sep 2008 03:42:08 +0100 Subject: [Ronja] Further lcd thoughs... In-Reply-To: <"p062405bac4e8de2b8f44(a)(091)130.161.115.44(093)*"@MHS> References: <48C32A32.3090601@vitalit.co.uk> Message-ID: <48C33F80.9050309@vitalit.co.uk> Cluestick gratefully received :-) What if the optics on the rx were focused to fill the ccd of a webcam (which has had it's ir filter removed), Or is it that optics for a 100 meter zoom shot are too bulky, need or too high a precision or are simply too costly ? If the optical zoom is high enough surely the output power only need to be above ambient to illuminate ? I just thought that as we are now looking for an image - not a light level we could get funky with the rx lenses ? J.D. Bakker wrote: > At 02:11 +0100 07-09-2008, ronjalist wrote: >> OK 50Hz is not enough, but you do have 1024x768 (or better) resolution >> and you could forgo the main optics, use a fresnel lens and subdivide >> the panel into 1000 segments... >> >> With a telescopic lens arrangement and an lcd webcam as the rx. >> >> Thoughts ? > > It'll work, but at the expense of link range. Assuming equal transmit > power, and comparing two systems with comparable optics (no fair > giving a telescope to the webcam and not to the regular Ronja), each > of those 1000 segments will have 1/1000th of the total power. > Achievable link distance is roughly proportional to the square root > of power, so those 1000 segments reduce the max link distance by a > factor of ~30, to just over 40 meter for a standard Ronja config. > > Now, in the real world, > > - LCD panels transmit less than 50% of all incident light > - on top of that, a significant portion of the light (>60%) will be > diffused/diffracted off-axis. This leaves only 10-20% of all > transmitted energy to hit the receiver > - you *will* need some separation between segments, or accept that > light from one segment will bleed into the next one. This, too, costs > TX energy. > - alignment is going to be harder > + on longer ranges, in-air absorption will eat more energy than the > square law suggests (so this would work in favor of a short-range > segmented system) > > More on link budgets here: > > http://www.pointless.net/pipermail/ronja/2006-August/009130.html > > JD 'no free lunch' B. From ronjalist at vitalit.co.uk Sun Sep 7 05:18:17 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Sun, 7 Sep 2008 05:18:17 +0100 Subject: [Ronja] Further lcd thoughs... In-Reply-To: <48C33F80.9050309@vitalit.co.uk> References: <48C32A32.3090601@vitalit.co.uk> Message-ID: <48C35609.5000008@vitalit.co.uk> In reply to self : Most webcams are at best 640x480 at 30fps. And as we would running monochrome (sans ir filter) there is simply not enough bandwidth to make it sensible. ronjalist wrote: > Cluestick gratefully received :-) > > What if the optics on the rx were focused to fill the ccd of a webcam > (which has had it's ir filter removed), Or is it that optics for a 100 > meter zoom shot are too bulky, need or too high a precision or are > simply too costly ? > > If the optical zoom is high enough surely the output power only need to > be above ambient to illuminate ? > > I just thought that as we are now looking for an image - not a light > level we could get funky with the rx lenses ? > > > J.D. Bakker wrote: >> At 02:11 +0100 07-09-2008, ronjalist wrote: >>> OK 50Hz is not enough, but you do have 1024x768 (or better) resolution >>> and you could forgo the main optics, use a fresnel lens and subdivide >>> the panel into 1000 segments... >>> >>> With a telescopic lens arrangement and an lcd webcam as the rx. >>> >>> Thoughts ? >> It'll work, but at the expense of link range. Assuming equal transmit >> power, and comparing two systems with comparable optics (no fair >> giving a telescope to the webcam and not to the regular Ronja), each >> of those 1000 segments will have 1/1000th of the total power. >> Achievable link distance is roughly proportional to the square root >> of power, so those 1000 segments reduce the max link distance by a >> factor of ~30, to just over 40 meter for a standard Ronja config. >> >> Now, in the real world, >> >> - LCD panels transmit less than 50% of all incident light >> - on top of that, a significant portion of the light (>60%) will be >> diffused/diffracted off-axis. This leaves only 10-20% of all >> transmitted energy to hit the receiver >> - you *will* need some separation between segments, or accept that >> light from one segment will bleed into the next one. This, too, costs >> TX energy. >> - alignment is going to be harder >> + on longer ranges, in-air absorption will eat more energy than the >> square law suggests (so this would work in favor of a short-range >> segmented system) >> >> More on link budgets here: >> >> http://www.pointless.net/pipermail/ronja/2006-August/009130.html >> >> JD 'no free lunch' B. > From antitron at web.de Sun Sep 7 13:08:48 2008 From: antitron at web.de (Thomas Egenhofer) Date: Sun, 07 Sep 2008 14:08:48 +0200 Subject: [Ronja] Further lcd thoughs... In-Reply-To: References: Message-ID: <1220789328.6047.11.camel@Nekomimi-host> if you want to imporve speed without breaking range or ending up with an expensive and impossible to use system. you have basically 3 choices: 1st. increase switching speeds of your transmitter/receiver. which is most likely only possible with laser-diodes without further trouble. -would require quite some changes to the existing hardware. 2nd. you you several different colors to transmit several bits paralell. -requires advanced optical system which i would prefer not to use. 3rd. and probably the most reasonable choice. use several light-levels for encoding the signal. 3-level encoding like the standart 100mbit network does sounds possible to me. requires a completely new transmitter and receiver. but has potential to make even autoneg working to some extend (since the idle signal would be missing on 10mbit connections its not fully operational, but 100mbit link should be ok just like that). > Content-Type: text/plain; charset="us-ascii"; format="flowed" > > In reply to self : > > Most webcams are at best 640x480 at 30fps. > And as we would running monochrome (sans ir filter) there is simply not > enough bandwidth to make it sensible. webcams or cemras in general. even if you manage to project your image over several hundret meter (try with a beamer if you have one) , allign it would be very difficult, even if you use 3x3 or 4x4 pixels for each bit. syncronising the refresh rates of both would be another issue. guess the resulting link would be extremly unstable.using a tv-remote is easyer, and in the end you might get more data transmitted with it. From ronjalist at vitalit.co.uk Sun Sep 7 23:25:31 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Sun, 7 Sep 2008 23:25:31 +0100 Subject: [Ronja] Further lcd thoughs... In-Reply-To: <1220789328.6047.11.camel@Nekomimi-host> References: Message-ID: <48C454DB.50803@vitalit.co.uk> The 3rd option is by far the best but to reach 25Mhz usually involves some mix of fpga/dsp/??? which are expensive and complicated to play with.... I remain convinced that there must be another less expensive more practical io route. A vga port for example can be used (depending on video card support) to generate a hdtv signal - if it can do that I suspect that it may be an outbound route - now need and inbound ..... Thomas Egenhofer wrote: > if you want to imporve speed without breaking range or ending up with an > expensive and impossible to use system. you have basically 3 choices: > 1st. increase switching speeds of your transmitter/receiver. which is > most likely only possible with laser-diodes without further trouble. > -would require quite some changes to the existing hardware. > > 2nd. you you several different colors to transmit several bits paralell. > -requires advanced optical system which i would prefer not to use. > > 3rd. and probably the most reasonable choice. use several light-levels > for encoding the signal. 3-level encoding like the standart 100mbit > network does sounds possible to me. requires a completely new > transmitter and receiver. but has potential to make even autoneg working > to some extend (since the idle signal would be missing on 10mbit > connections its not fully operational, but 100mbit link should be ok > just like that). > >> Content-Type: text/plain; charset="us-ascii"; format="flowed" >> >> In reply to self : >> >> Most webcams are at best 640x480 at 30fps. >> And as we would running monochrome (sans ir filter) there is simply not >> enough bandwidth to make it sensible. > webcams or cemras in general. even if you manage to project your image > over several hundret meter (try with a beamer if you have one) , allign > it would be very difficult, even if you use 3x3 or 4x4 pixels for each > bit. syncronising the refresh rates of both would be another issue. > guess the resulting link would be extremly unstable.using a tv-remote is > easyer, and in the end you might get more data transmitted with it. > > From gmaxwell at gmail.com Sun Sep 7 23:29:49 2008 From: gmaxwell at gmail.com (Gregory Maxwell) Date: Sun, 7 Sep 2008 18:29:49 -0400 Subject: [Ronja] Further lcd thoughs... In-Reply-To: <48C454DB.50803@vitalit.co.uk> References: <1220789328.6047.11.camel@Nekomimi-host> <48C454DB.50803@vitalit.co.uk> Message-ID: On Sun, Sep 7, 2008 at 6:25 PM, ronjalist wrote: > The 3rd option is by far the best but to reach 25Mhz usually involves > some mix of fpga/dsp/??? which are expensive and complicated to play > with.... > > I remain convinced that there must be another less expensive more > practical io route. > > A vga port for example can be used (depending on video card support) to > generate a hdtv signal - if it can do that I suspect that it may be an > outbound route - now need and inbound ..... In fact, a VGA port could be used as a very inexpensive high speed digital to analog converter to achieve a multi-level encoding purely in software... but then you do not have a matching recieve side. The GNUradio board with the LF TX and LF RX board are pretty much exactly what you'd need for building a multi-level encoded RONJA, but it is not inexpensive. From jdb at lartmaker.nl Sun Sep 7 23:49:06 2008 From: jdb at lartmaker.nl (J.D. Bakker) Date: Mon, 8 Sep 2008 00:49:06 +0200 Subject: [Ronja] Further lcd thoughs... In-Reply-To: <48C454DB.50803@vitalit.co.uk> References: <48C454DB.50803@vitalit.co.uk> Message-ID: At 23:25 +0100 07-09-2008, ronjalist wrote: >A vga port for example can be used (depending on video card support) to >generate a hdtv signal - if it can do that I suspect that it may be an >outbound route - now need and inbound ..... Unfortunately the transmitter isn't the hard bit (apart from modulating a light source fast enough, but even that's comparatively simple). JDB. -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lartmaker.nl/ From ronjalist at vitalit.co.uk Mon Sep 8 03:30:22 2008 From: ronjalist at vitalit.co.uk (ronjalist) Date: Mon, 8 Sep 2008 03:30:22 +0100 Subject: [Ronja] Further lcd thoughs... In-Reply-To: <"p062405bdc4ea0a7c9739(a)(091)130.161.115.44(093)*"@MHS> References: Message-ID: <48C48E3E.7000101@vitalit.co.uk> RX - old 52x cdrom drive ? Given that hacking the firmware of cd / dvd drives is common place (al la xbox360) is this not a possible source for an rx ? J.D. Bakker wrote: > At 23:25 +0100 07-09-2008, ronjalist wrote: >> A vga port for example can be used (depending on video card support) to >> generate a hdtv signal - if it can do that I suspect that it may be an >> outbound route - now need and inbound ..... > > Unfortunately the transmitter isn't the hard bit (apart from > modulating a light source fast enough, but even that's comparatively > simple). > > JDB. From asteri_x at freemail.hu Tue Sep 9 23:51:29 2008 From: asteri_x at freemail.hu (=?UTF-8?B?R3l1cmvDsyBNYXJ0aW4=?=) Date: Wed, 10 Sep 2008 00:51:29 +0200 Subject: [Ronja] Further lcd thoughs... In-Reply-To: <48C48E3E.7000101@vitalit.co.uk> References: <48C48E3E.7000101@vitalit.co.uk> Message-ID: <48C6FDF1.8030604@freemail.hu> I have some experience in the dvd and cd industry. I worked for Datarius for 3 years. First you need to figure out, which connections of the pickuphead (PUH) are the photodiode pins and the LD pins. Also the area of those photodiodes is so small, that it is almost impossible to reasonably collect ligth with these built in diodes. Manufactureres never give out specifications. I hate them already. I also were thinking of using the complete PUH tx and rx without even disassembling the optics. CD and DVD uses polarization shift for being able to send and receive coaxially. This means, the rx is 90 degrees turned respective to tx. If 2 identical PUH are used, and are 90 degrees turned, they should even be able to work in full duplex, because of the orthogonal polarization. Also the voice-coil lens of the PUH is able to compensate for scintillation in the air. This whole thing would make a PHD thesis. Especially the compensation thing. Maybe I someday do it when I have time. The MLT-3 of the 100Mbit ethernet could be almost directly used. You need linear amplification for the LD tx and the rx. Some optical power regulation in the TX too, thats all. Maybe the autonegotiation spikes must be suppressed to avoid damages of the LD. The Idle signal is always transmitted in a scrambled form, so you are always syncronized. But for the record: Use red light! not IR! "Never stare into the laser with your remaining good eye..." IR os no fun, you hardly see where it goes. Bye, Martin ronjalist ?rta: > RX - old 52x cdrom drive ? > > Given that hacking the firmware of cd / dvd drives is common place (al > la xbox360) is this not a possible source for an rx ? > > J.D. Bakker wrote: >> At 23:25 +0100 07-09-2008, ronjalist wrote: >>> A vga port for example can be used (depending on video card support) to >>> generate a hdtv signal - if it can do that I suspect that it may be an >>> outbound route - now need and inbound ..... >> Unfortunately the transmitter isn't the hard bit (apart from >> modulating a light source fast enough, but even that's comparatively >> simple). >> >> JDB. > From jakub.michnik at centrum.cz Mon Sep 15 09:53:18 2008 From: jakub.michnik at centrum.cz (=?UTF-8?B?SmFrdWIgTWljaG7DrWs=?=) Date: Mon, 15 Sep 2008 10:53:18 +0200 Subject: [Ronja] Dosah ronji s laserem In-Reply-To: <48C6FDF1.8030604@freemail.hu> References: <48C48E3E.7000101@vitalit.co.uk> <48C6FDF1.8030604@freemail.hu> Message-ID: <48CE227E.6080203@centrum.cz> Chci se zeptat tech, co s tim neco zkouseli, na jakou vzdalenost se dostali s pouzitim laseru (jakeho) a jak velke problemy meli s mirenim, nebo spis s casovou stalosti namireni. Proste jestli to casem neuhlo. btw mam tu 2x Rx a 2x Tx v SMD, takze kdyby nekdo chtel, dejte vedet. Kubas From twibright at hispeed.ch Sat Sep 20 17:50:36 2008 From: twibright at hispeed.ch (Karel Kulhavy) Date: Sat, 20 Sep 2008 18:50:36 +0200 Subject: [Ronja] Multiple receivers and transmitters Message-ID: <13250471.1221929436089.JavaMail.root@viefep12> I have been thinking how to divide multiple heads between receivers and transmitters optimally, i. e. if we want to build N heads, how to get the maximum range. Let?s assume there is no absorption in the atmosphere here. The answer is make 2/3 transmitters and 1/3 receivers. Example: 1 receiver and 2 transmitters, 2 receivers and 4 tx, 3 rx and 6 tx etc. Making 4 TX instead of 1 increases the signal nd the noise stays the same, the range is double. Making 4 RX increases the signal 4x and the noise 2x (each RX generates statistically independent noise) with resulting sqrt(2)=1.4x increase in distance. If we try to maximize 4th power of the distance (which is the same as maximizing the distance) and mark t the percentage of transmitter heads (between 0 and 1), the maximized function is t^2*(1-t)=t^2-t^3. Derivative is 2t-3t^2=0. 2-3t=0, t=2/3. Karel From jdb at lartmaker.nl Sat Sep 20 18:09:12 2008 From: jdb at lartmaker.nl (J.D. Bakker) Date: Sat, 20 Sep 2008 19:09:12 +0200 Subject: [Ronja] Multiple receivers and transmitters In-Reply-To: <13250471.1221929436089.JavaMail.root@viefep12> References: <13250471.1221929436089.JavaMail.root@viefep12> Message-ID: >I have been thinking how to divide multiple heads between receivers >and transmitters optimally, i. e. if we want to build N heads, how >to get the maximum range. Given N transmitters and M receivers, it might be better (but more complicated) to apply MIMO techniques. JDB. -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lartmaker.nl/ From twibright at hispeed.ch Sat Sep 20 18:26:20 2008 From: twibright at hispeed.ch (Karel Kulhavy) Date: Sat, 20 Sep 2008 19:26:20 +0200 Subject: [Ronja] Multiple receivers and transmitters Message-ID: <4306667.1221931580950.JavaMail.root@viefep12> ---- "J.D. Bakker" schrieb: > >I have been thinking how to divide multiple heads between receivers > >and transmitters optimally, i. e. if we want to build N heads, how > >to get the maximum range. > > Given N transmitters and M receivers, it might be better (but more > complicated) to apply MIMO techniques. Once I was running with a friend and a friend of him who was running there too turned out to do MIMO for his doctoral thesis. From what he said I understood that MIMO is that you have N RX and N TX and a matrix and if you invert this matrix it behaves like N wires parallel together, ie. unit matrix. Did I undersdtood it right? Karel > > JDB. > -- > LART. 250 MIPS under one Watt. Free hardware design files. > http://www.lartmaker.nl/ > > -- > Twibright Ronja mailing list http://ronja.twibright.com > Ronja at lists.pointless.net > http://pointless.net/mailman/listinfo/ronja From gmaxwell at gmail.com Sat Sep 20 18:53:25 2008 From: gmaxwell at gmail.com (Gregory Maxwell) Date: Sat, 20 Sep 2008 13:53:25 -0400 Subject: [Ronja] Multiple receivers and transmitters In-Reply-To: References: <13250471.1221929436089.JavaMail.root@viefep12> Message-ID: On Sat, Sep 20, 2008 at 1:09 PM, J.D. Bakker wrote: >>I have been thinking how to divide multiple heads between receivers >>and transmitters optimally, i. e. if we want to build N heads, how >>to get the maximum range. > > Given N transmitters and M receivers, it might be better (but more > complicated) to apply MIMO techniques. In radio the problem that MMIO solves is multipath: Between your TX and RX there are obstructions which alter the signal, they shift phase in a frequency dependent manner, and creates nulls and peaks. MMIO uses spatially separated antenna which get different views of the multi-path environment. You can then define a collection of filters which take advantage of the multipath to form separate channels and gain capacity. For a fairly slowly amplitude modulated optical path like RONJA I would not expect there to be significant multipath. I do not believe MMIO would help. But MMIO systems typically use efficient modulation systems which would probably help ronja very much, but supporting them would be very costly. Karel where does the noise come from in the receive head are we talking about optical background noise which would increase with increased collection area or is it primarily noise from the photodetector and amplifier? If it is primary electronic noise, creating a larger passive collector may allow you to increase receiver area without seeing much of a nosie increase. So long as your transmitter area is strictly smaller than the reciever's field of view, however, you would always be as well or better off to increase the transmitter area. though perhaps you may wish to make the receiver more selective (for example, by placing it at the back of a long black tube). From pkyaduvanshi at gmail.com Sat Sep 20 19:25:53 2008 From: pkyaduvanshi at gmail.com (PRAMOD YADUVANSHI) Date: Sat, 20 Sep 2008 23:55:53 +0530 Subject: [Ronja] Multiple receivers and transmitters In-Reply-To: References: <13250471.1221929436089.JavaMail.root@viefep12> Message-ID: <50d3ad760809201125w339c6ba4i51183cb3a6240e9e@mail.gmail.com> I feel for a laser based link MIMO will be a better option to improve SNR but for LED based Tx/Rx scintillation effects will be almost negligible and MIMO will hardly help. 2008/9/20 Gregory Maxwell > On Sat, Sep 20, 2008 at 1:09 PM, J.D. Bakker wrote: > >>I have been thinking how to divide multiple heads between receivers > >>and transmitters optimally, i. e. if we want to build N heads, how > >>to get the maximum range. > > > > Given N transmitters and M receivers, it might be better (but more > > complicated) to apply MIMO techniques. > > > In radio the problem that MMIO solves is multipath: Between your TX > and RX there are obstructions which alter the signal, they shift > phase in a frequency dependent manner, and creates nulls and peaks. > MMIO uses spatially separated antenna which get different views of the > multi-path environment. You can then define a collection of filters > which take advantage of the multipath to form separate channels and > gain capacity. > > For a fairly slowly amplitude modulated optical path like RONJA I > would not expect there to be significant multipath. I do not believe > MMIO would help. But MMIO systems typically use efficient modulation > systems which would probably help ronja very much, but supporting > them would be very costly. > > Karel where does the noise come from in the receive head are we > talking about optical background noise which would increase with > increased collection area or is it primarily noise from the > photodetector and amplifier? If it is primary electronic noise, > creating a larger passive collector may allow you to increase receiver > area without seeing much of a nosie increase. > > So long as your transmitter area is strictly smaller than the > reciever's field of view, however, you would always be as well or > better off to increase the transmitter area. though perhaps you may > wish to make the receiver more selective (for example, by placing it > at the back of a long black tube). > > -- > Twibright Ronja mailing list http://ronja.twibright.com > Ronja at lists.pointless.net > http://pointless.net/mailman/listinfo/ronja > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://pointless.net/pipermail/ronja/attachments/20080920/92b44b8e/attachment.html From cd930 at centrum.cz Sun Sep 21 06:07:03 2008 From: cd930 at centrum.cz (CD930) Date: Sun, 21 Sep 2008 07:07:03 +0200 Subject: [Ronja] Multiple receivers and transmitters References: <13250471.1221929436089.JavaMail.root@viefep12> <50d3ad760809201125w339c6ba4i51183cb3a6240e9e@mail.gmail.com> Message-ID: <5D3A10597883401C8704FE5950D69AD1@Martindoma> MIMO and FSO? ehm... noooo. This is only for RF with COFDM and guard interval pilots. Martin ----- Original Message ----- From: PRAMOD YADUVANSHI To: Twibright Ronja Sent: Saturday, September 20, 2008 8:25 PM Subject: Re: [Ronja] Multiple receivers and transmitters I feel for a laser based link MIMO will be a better option to improve SNR but for LED based Tx/Rx scintillation effects will be almost negligible and MIMO will hardly help. 2008/9/20 Gregory Maxwell On Sat, Sep 20, 2008 at 1:09 PM, J.D. Bakker wrote: >>I have been thinking how to divide multiple heads between receivers >>and transmitters optimally, i. e. if we want to build N heads, how >>to get the maximum range. > > Given N transmitters and M receivers, it might be better (but more > complicated) to apply MIMO techniques. In radio the problem that MMIO solves is multipath: Between your TX and RX there are obstructions which alter the signal, they shift phase in a frequency dependent manner, and creates nulls and peaks. MMIO uses spatially separated antenna which get different views of the multi-path environment. You can then define a collection of filters which take advantage of the multipath to form separate channels and gain capacity. For a fairly slowly amplitude modulated optical path like RONJA I would not expect there to be significant multipath. I do not believe MMIO would help. But MMIO systems typically use efficient modulation systems which would probably help ronja very much, but supporting them would be very costly. Karel where does the noise come from in the receive head are we talking about optical background noise which would increase with increased collection area or is it primarily noise from the photodetector and amplifier? If it is primary electronic noise, creating a larger passive collector may allow you to increase receiver area without seeing much of a nosie increase. So long as your transmitter area is strictly smaller than the reciever's field of view, however, you would always be as well or better off to increase the transmitter area. though perhaps you may wish to make the receiver more selective (for example, by placing it at the back of a long black tube). -- Twibright Ronja mailing list http://ronja.twibright.com Ronja na lists.pointless.net http://pointless.net/mailman/listinfo/ronja ------------------------------------------------------------------------------ -- Twibright Ronja mailing list http://ronja.twibright.com Ronja na lists.pointless.net http://pointless.net/mailman/listinfo/ronja ------------- dal?? ??st --------------- HTML p??loha byla odstran?na... URL: http://pointless.net/pipermail/ronja/attachments/20080921/2f19745f/attachment.html From sr6220305 at yahoo.com Tue Sep 23 20:23:03 2008 From: sr6220305 at yahoo.com (Sean Ramdeholl) Date: Tue, 23 Sep 2008 12:23:03 -0700 (PDT) Subject: [Ronja] Multiple receivers and transmitters In-Reply-To: <5D3A10597883401C8704FE5950D69AD1@Martindoma> Message-ID: <644721.82159.qm@web55505.mail.re4.yahoo.com> Back to skool boys? Project title: Characterization of MIMO Free-Space Optical Channels In Free-Space Optical (FSO) communication, information data is transmitted wirelessly using laser beams. Such links have the potential to provide virtually unlimited bandwidth without the expense of buried fibre optic cabling. Unfortunately, FSO link reliability is highly dependent on atmospheric conditions such as turbulence, rain, cloud, dust and fog. These effects cause fading of the received laser beam. In typical FSO channels, a deep fade can cause the loss of millions of data bits. Improvement in reliability can be gained by using multiple lasers and multiple apertures to create a multiple-input multiple output (MIMO) FSO channel. A comprehensive characterisation of this channel is crucial in the future development of reliable and efficient FSO communication systems. In this project, the student will be involved in a 1.5-2 km MIMO FSO laser range experiment. This will include assisting with the setup and calibration of laser experiment equipment, and the collection of channel measurement data. The student will characterise the channel by fitting known statistical distributions to the measurement data, and analysing the temporal and spatial correlation of the channel www.unisa.edu.au/resdegrees/scholarships/unisaSRSprojects.asp Sean. Sean Ramdeholl --- On Sun, 9/21/08, CD930 wrote: From: CD930 Subject: Re: [Ronja] Multiple receivers and transmitters To: "Twibright Ronja" Date: Sunday, September 21, 2008, 2:07 AM MIMO and FSO?? ehm...? noooo. This is only for RF with COFDM and guard interval pilots. Martin -------------- next part -------------- An HTML attachment was scrubbed... URL: http://pointless.net/pipermail/ronja/attachments/20080923/79665940/attachment.html From twibright at hispeed.ch Tue Sep 23 22:41:30 2008 From: twibright at hispeed.ch (Karel Kulhavy) Date: Tue, 23 Sep 2008 23:41:30 +0200 Subject: [Ronja] Multiple receivers and transmitters Message-ID: <25921703.1222206090944.JavaMail.root@viefep15> ---- Gregory Maxwell schrieb: > On Sat, Sep 20, 2008 at 1:09 PM, J.D. Bakker wrote: > >>I have been thinking how to divide multiple heads between receivers > >>and transmitters optimally, i. e. if we want to build N heads, how > >>to get the maximum range. > > > > Given N transmitters and M receivers, it might be better (but more > > complicated) to apply MIMO techniques. > > > In radio the problem that MMIO solves is multipath: Between your TX > and RX there are obstructions which alter the signal, they shift > phase in a frequency dependent manner, and creates nulls and peaks. > MMIO uses spatially separated antenna which get different views of the > multi-path environment. You can then define a collection of filters > which take advantage of the multipath to form separate channels and > gain capacity. > > For a fairly slowly amplitude modulated optical path like RONJA I > would not expect there to be significant multipath. I do not believe > MMIO would help. But MMIO systems typically use efficient modulation > systems which would probably help ronja very much, but supporting > them would be very costly. > > Karel where does the noise come from in the receive head are we > talking about optical background noise which would increase with > increased collection area or is it primarily noise from the > photodetector and amplifier? If it is primary electronic noise, In the night the noise is made mostly by the amplifier. In the daylight the noise is mostly created by the quantization of the ambient light. Karel From jdb at lartmaker.nl Tue Sep 23 23:17:37 2008 From: jdb at lartmaker.nl (J.D. Bakker) Date: Wed, 24 Sep 2008 00:17:37 +0200 Subject: [Ronja] Multiple receivers and transmitters In-Reply-To: <25921703.1222206090944.JavaMail.root@viefep15> References: <25921703.1222206090944.JavaMail.root@viefep15> Message-ID: At 23:41 +0200 23-09-2008, Karel Kulhavy wrote: > In the >daylight the noise is mostly created by the quantization of the ambient light. Strictly speaking that's not what happens (or at best it's a major oversimplification). The photodiode is a semiconductor in which incident light causes an electrical current to flow. In daylight a large part of that current is effectively DC (compared to the signal bandwidth). Any DC current flowing through a semiconductor junction induces shot noise as individual electrons cross the barrier, which can be likened to the noise of raindrops falling on a thin roof. This noise current is indistinguishable from the wanted signal. Shot noise increases with the square root of current flow, and thus with the square root of incident light power. On a sunny day the receiver gets 10^5 more light energy than on a moonlit night, but the difference in noise level is 'only' 300x. JDB. -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lartmaker.nl/ From twibright at hispeed.ch Wed Sep 24 00:24:45 2008 From: twibright at hispeed.ch (Karel Kulhavy) Date: Wed, 24 Sep 2008 1:24:45 +0200 Subject: [Ronja] Multiple receivers and transmitters Message-ID: <7058832.1222212285196.JavaMail.root@viefep15> ---- "J.D. Bakker" schrieb: > At 23:41 +0200 23-09-2008, Karel Kulhavy wrote: > > In the > >daylight the noise is mostly created by the quantization of the ambient light. > > Strictly speaking that's not what happens (or at best it's a major > oversimplification). > > The photodiode is a semiconductor in which incident light causes an > electrical current to flow. In daylight a large part of that current > is effectively DC (compared to the signal bandwidth). Any DC current > flowing through a semiconductor junction induces shot noise as > individual electrons cross the barrier, which can be likened to the > noise of raindrops falling on a thin roof. This noise current is > indistinguishable from the wanted signal. This is not true. If you irradiate the photodiode with some special laser light (mode-squeezed or how it's called) you get less noise than shot noise. The noise comes from the random times of arrival of the photons. But we cannot control the type of light the Sun produces so it cannot be exploited for increasing the range anyway. Karel > > Shot noise increases with the square root of current flow, and thus > with the square root of incident light power. On a sunny day the > receiver gets 10^5 more light energy than on a moonlit night, but the > difference in noise level is 'only' 300x. > > JDB. > -- > LART. 250 MIPS under one Watt. Free hardware design files. > http://www.lartmaker.nl/ > > -- > Twibright Ronja mailing list http://ronja.twibright.com > Ronja at lists.pointless.net > http://pointless.net/mailman/listinfo/ronja From twibright at hispeed.ch Sun Sep 28 10:27:24 2008 From: twibright at hispeed.ch (Karel Kulhavy) Date: Sun, 28 Sep 2008 11:27:24 +0200 Subject: [Ronja] Problem with opening a gerber file *grb by a berbv program :( Message-ID: <13970990.1222594044216.JavaMail.root@viefep16> ---- Pawel Plociennik schrieb: > Hi, > > I have trying in a gerbv program to open a gerber file (for example a twister2.pcb.output_back.gbr from a twister2 packet) it is failed by this below error : > "Error: eval: unbound variable: g04" I have just tried with gerbv 2.1.0 and it works. Karel From twibright at hispeed.ch Sun Sep 28 12:52:11 2008 From: twibright at hispeed.ch (Karel Kulhavy) Date: Sun, 28 Sep 2008 13:52:11 +0200 Subject: [Ronja] Measuring transistor with a diode multimeter Message-ID: <5092895.1222602731035.JavaMail.root@viefep16> I have an idea that maybe transistor amplification h21e could be measured with a diode-only multimeter. Unfortunately I don't have a transistor at home. Could you please someone measure some transistor B against E B connected with C against E and paste the exact numbers? I'll try to magically guess the h21e from it and then we'll see if it works ;-) Karel