Talk:Buzz or not

From Openmoko

(Difference between revisions)
Jump to: navigation, search
(+Context)
(no sw-fix)
Line 32: Line 32:
 
My understanding of the affair is that we are talking a potential product recall. Finding a software solution is still much preferable (refer to the GSM TTFF bug case). I understand that they want to give it one more try for a reasonable amount of time, but of course I am not sitting with a buggy phone^H^H^H^H^H^H brick waiting for resolution. I would probably be less patient !
 
My understanding of the affair is that we are talking a potential product recall. Finding a software solution is still much preferable (refer to the GSM TTFF bug case). I understand that they want to give it one more try for a reasonable amount of time, but of course I am not sitting with a buggy phone^H^H^H^H^H^H brick waiting for resolution. I would probably be less patient !
 
[[User:MinhHaDuong|MinhHaDuong]] 09:26, 15 December 2008 (UTC)
 
[[User:MinhHaDuong|MinhHaDuong]] 09:26, 15 December 2008 (UTC)
 +
 +
 +
I don't know of any plans inside OM to fix the buzz issue via SW. Buzz is considered resistant against all SW-fixes. The way it works is basically well understood by now, and there's just no possible vector to reasonably come up with a SW-fix for that. (of course there's always a chance we overseen a small detail that actually can be changed by some register-setting or something like that, so I had to say "we don't see any possible vector...". Filtering the buzz out of the audio-signal by some fancy algorithms running on main CPU probably isn't possible at all, due to missing DSP-grunt at least)   
 +
[[User:jOERG|jOERG]] 2008-12-15

Revision as of 09:43, 15 December 2008

Who is "WE"? On OM-HW-EE side this poll isn't considered helpful, as it contains no info whatsoever about actual network-situation (which is widely assumed to be the cause of buzz appearing in a particular situation). Please note the original poster seems not affiliated to OM technical staff, and we (OM) might eventually ask our customers to support a different poll. Also note there's a rework paper to fix buzz, a non-verified draft for now [jOERG]


We, I like to think, are the people who are having a problem using the FreeRunner as a phone. If you want to put up an official OpenMoko poll/questionnaire please do so, but can you make it quick? Tel us how people who can't handle a soldering tool in a professional way can fix their phones. [Michel an OpenMoko customer and believer that a phone can be open source]


Finding a way to fix by starting a poll? Hmm, maybe like this:

Country ZIP code skills and willing to help (Y), need help(N) contact (mail, IRC)
D 90429 Y max. 10 devices per week. Take beer ;-) joerg@openmoko.org

Hi, you said "and we (OM) might eventually ask our customers to support a different poll" if the above is an offical OM poll then there maybe some humor left at openmoko.org :) Allthough I think the poll I started (after Steve Mosher from openmoko.com sugested it) is a bit more helpfull in solving it, or at least give some sort of overview, then in providing you with beer :) [Michel]


I prefer the beer though, as I don't see the "solving it" coming from the original poll. Overview is relative, you will find most people don't even know what band they're ACTUALLY on, and all this has been discussed thoroughly on mailing list.

Well it's up to you, I'm really curious about the results we will see from this.

In some wikipages, the history really is the best :-) /jOERG


This will certainly help to document if applying a fix helped to solve a problem or not. Will also help to contact people to duplicate problems so CND becomes less of an issue. [Rakshat]


Newcomers to this discussion should know that Openmoko has worked out a hardware solution to the "buzz" issue. The technical draft describing the procedure is being circulated on the mailing lists. It involves soldering a capacitor and changing a resistor. The poll that Joerg suggests above is about volunteering to fix other people's units, provided the said volunteers has the hardware skills and equipment of course. The alternative would be for owners to bring their unit and the procedure description to a cellphone repair shop (or an electronics club). The delicate question behind all this is of course liability. I cannot imagine anybody would be too happy to foot the repair bill and endorse the surgery risk (although that part would probably remain with the repair shop). Plus there are practical and legal complexities to apply the harware fix to all units: they are all over the Earth now, some have been sold direct, some through retailers. Just managing that would requires several week-persons ! My understanding of the affair is that we are talking a potential product recall. Finding a software solution is still much preferable (refer to the GSM TTFF bug case). I understand that they want to give it one more try for a reasonable amount of time, but of course I am not sitting with a buggy phone^H^H^H^H^H^H brick waiting for resolution. I would probably be less patient ! MinhHaDuong 09:26, 15 December 2008 (UTC)


I don't know of any plans inside OM to fix the buzz issue via SW. Buzz is considered resistant against all SW-fixes. The way it works is basically well understood by now, and there's just no possible vector to reasonably come up with a SW-fix for that. (of course there's always a chance we overseen a small detail that actually can be changed by some register-setting or something like that, so I had to say "we don't see any possible vector...". Filtering the buzz out of the audio-signal by some fancy algorithms running on main CPU probably isn't possible at all, due to missing DSP-grunt at least) jOERG 2008-12-15

Personal tools

Who is "WE"? On OM-HW-EE side this poll isn't considered helpful, as it contains no info whatsoever about actual network-situation (which is widely assumed to be the cause of buzz appearing in a particular situation). Please note the original poster seems not affiliated to OM technical staff, and we (OM) might eventually ask our customers to support a different poll. Also note there's a rework paper to fix buzz, a non-verified draft for now [jOERG]


We, I like to think, are the people who are having a problem using the FreeRunner as a phone. If you want to put up an official OpenMoko poll/questionnaire please do so, but can you make it quick? Tel us how people who can't handle a soldering tool in a professional way can fix their phones. [Michel an OpenMoko customer and believer that a phone can be open source]


Finding a way to fix by starting a poll? Hmm, maybe like this:

Country ZIP code skills and willing to help (Y), need help(N) contact (mail, IRC)
D 90429 Y max. 10 devices per week. Take beer ;-) joerg@openmoko.org

Hi, you said "and we (OM) might eventually ask our customers to support a different poll" if the above is an offical OM poll then there maybe some humor left at openmoko.org :) Allthough I think the poll I started (after Steve Mosher from openmoko.com sugested it) is a bit more helpfull in solving it, or at least give some sort of overview, then in providing you with beer :) [Michel]


I prefer the beer though, as I don't see the "solving it" coming from the original poll. Overview is relative, you will find most people don't even know what band they're ACTUALLY on, and all this has been discussed thoroughly on mailing list.

Well it's up to you, I'm really curious about the results we will see from this.

In some wikipages, the history really is the best :-) /jOERG


This will certainly help to document if applying a fix helped to solve a problem or not. Will also help to contact people to duplicate problems so CND becomes less of an issue. [Rakshat]


Newcomers to this discussion should know that Openmoko has worked out a hardware solution to the "buzz" issue. The technical draft describing the procedure is being circulated on the mailing lists. It involves soldering a capacitor and changing a resistor. The poll that Joerg suggests above is about volunteering to fix other people's units, provided the said volunteers has the hardware skills and equipment of course. The alternative would be for owners to bring their unit and the procedure description to a cellphone repair shop (or an electronics club). The delicate question behind all this is of course liability. I cannot imagine anybody would be too happy to foot the repair bill and endorse the surgery risk (although that part would probably remain with the repair shop). Plus there are practical and legal complexities to apply the harware fix to all units: they are all over the Earth now, some have been sold direct, some through retailers. Just managing that would requires several week-persons ! My understanding of the affair is that we are talking a potential product recall. Finding a software solution is still much preferable (refer to the GSM TTFF bug case). I understand that they want to give it one more try for a reasonable amount of time, but of course I am not sitting with a buggy phone^H^H^H^H^H^H brick waiting for resolution. I would probably be less patient ! MinhHaDuong 09:26, 15 December 2008 (UTC)


I don't know of any plans inside OM to fix the buzz issue via SW. Buzz is considered resistant against all SW-fixes. The way it works is basically well understood by now, and there's just no possible vector to reasonably come up with a SW-fix for that. (of course there's always a chance we overseen a small detail that actually can be changed by some register-setting or something like that, so I had to say "we don't see any possible vector...". Filtering the buzz out of the audio-signal by some fancy algorithms running on main CPU probably isn't possible at all, due to missing DSP-grunt at least) jOERG 2008-12-15