Re: Nandblasting not working in one xo
Thank you very much John for the explanation and all the help. :) G'day. On Tue, Jul 10, 2012 at 9:33 AM, John Watlington wrote: > > NANDBlaster uses a fixed transmit speed (modulation). > If the signal budget for a laptop isn't sufficient to support > that speed, it will fail to receive many packets. > > When using normal WiFi, the transmit speed (modulation) > is decreased until reliable communication can be obtained --- > therefore a laptop with decreased signal budget (e.g. bad antenna) > may still work, although with degraded performance. > > Regards, > John > > On Jul 9, 2012, at 11:10 PM, James Cameron wrote: > > > That the antenna change did not work shows the problem is in the > > wireless card. > > > > You asked why not use the same mechanism as Sugar? > > > > Consider the transmitter performance. > > > > Your network used by Sugar probably has an access point with higher > > transmit power and better antenna than the laptop being used as > > NANDblaster sender. > > > > So it is perhaps the combination of small damage to one laptop and > > large damage to another laptop, that causes NANDblaster to fail. But > > the combination of good access point and large damage causes Sugar > > networking to be successful. > > > > See http://wiki.laptop.org/go/Antenna_testing#Link_Budget for a > > calculation of wireless success, to see what variables are important. > > > > Does Sugar in ad-hoc wireless mode work between the two laptops? Or > > Sugar in mesh wireless mode with no other laptops nearby? > > > > If so, that's very interesting. > > > > Open Firmware and Linux use different commands sent to the wireless card. > > > > I've checked, and we are using the same wireless firmware 5.110.22.p23 > > in both Open Firmware and Linux (build 883). > > > > Daniel, do you know of any commands that the Linux kernel may have > > sent to the card that may improve signal, even by accident? > > > > On Mon, Jul 09, 2012 at 04:06:16PM +0545, Roshan Karki wrote: > >> I tried with antenna change but as you told, didn't work. So I think > this is > >> the dead end. Thank you for your help. But one question I wonder is in > Sugar I > >> can use very poor network very well. Why not use the same mechanism in > OFW as > >> well? > >> > >> On Fri, Jul 6, 2012 at 10:41 AM, James Cameron > wrote: > >> > >>G'day, > >> > >>Thanks for the photographs. There's nothing wrong that I can see > >>either. > >> > >>Repair may attempt antenna change, but it is unlikely to be fixed > with > >>only antenna change. > >> > >>Perhaps the radio module has been damaged. On XO-1 the module is > >>soldered down and is impractical to replace. In later models > (XO-1.5, > >>XO-1.75) the module is in a socket. > >> > >>-- > >>James Cameron > >>http://quozl.linux.org.au/ > >> > >> > > > > -- > > James Cameron > > http://quozl.linux.org.au/ > > ___ > > Devel mailing list > > Devel@lists.laptop.org > > http://lists.laptop.org/listinfo/devel > > ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
NANDBlaster uses a fixed transmit speed (modulation). If the signal budget for a laptop isn't sufficient to support that speed, it will fail to receive many packets. When using normal WiFi, the transmit speed (modulation) is decreased until reliable communication can be obtained --- therefore a laptop with decreased signal budget (e.g. bad antenna) may still work, although with degraded performance. Regards, John On Jul 9, 2012, at 11:10 PM, James Cameron wrote: > That the antenna change did not work shows the problem is in the > wireless card. > > You asked why not use the same mechanism as Sugar? > > Consider the transmitter performance. > > Your network used by Sugar probably has an access point with higher > transmit power and better antenna than the laptop being used as > NANDblaster sender. > > So it is perhaps the combination of small damage to one laptop and > large damage to another laptop, that causes NANDblaster to fail. But > the combination of good access point and large damage causes Sugar > networking to be successful. > > See http://wiki.laptop.org/go/Antenna_testing#Link_Budget for a > calculation of wireless success, to see what variables are important. > > Does Sugar in ad-hoc wireless mode work between the two laptops? Or > Sugar in mesh wireless mode with no other laptops nearby? > > If so, that's very interesting. > > Open Firmware and Linux use different commands sent to the wireless card. > > I've checked, and we are using the same wireless firmware 5.110.22.p23 > in both Open Firmware and Linux (build 883). > > Daniel, do you know of any commands that the Linux kernel may have > sent to the card that may improve signal, even by accident? > > On Mon, Jul 09, 2012 at 04:06:16PM +0545, Roshan Karki wrote: >> I tried with antenna change but as you told, didn't work. So I think this is >> the dead end. Thank you for your help. But one question I wonder is in Sugar >> I >> can use very poor network very well. Why not use the same mechanism in OFW as >> well? >> >> On Fri, Jul 6, 2012 at 10:41 AM, James Cameron wrote: >> >>G'day, >> >>Thanks for the photographs. There's nothing wrong that I can see >>either. >> >>Repair may attempt antenna change, but it is unlikely to be fixed with >>only antenna change. >> >>Perhaps the radio module has been damaged. On XO-1 the module is >>soldered down and is impractical to replace. In later models (XO-1.5, >>XO-1.75) the module is in a socket. >> >>-- >>James Cameron >>http://quozl.linux.org.au/ >> >> > > -- > James Cameron > http://quozl.linux.org.au/ > ___ > Devel mailing list > Devel@lists.laptop.org > http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
That the antenna change did not work shows the problem is in the wireless card. You asked why not use the same mechanism as Sugar? Consider the transmitter performance. Your network used by Sugar probably has an access point with higher transmit power and better antenna than the laptop being used as NANDblaster sender. So it is perhaps the combination of small damage to one laptop and large damage to another laptop, that causes NANDblaster to fail. But the combination of good access point and large damage causes Sugar networking to be successful. See http://wiki.laptop.org/go/Antenna_testing#Link_Budget for a calculation of wireless success, to see what variables are important. Does Sugar in ad-hoc wireless mode work between the two laptops? Or Sugar in mesh wireless mode with no other laptops nearby? If so, that's very interesting. Open Firmware and Linux use different commands sent to the wireless card. I've checked, and we are using the same wireless firmware 5.110.22.p23 in both Open Firmware and Linux (build 883). Daniel, do you know of any commands that the Linux kernel may have sent to the card that may improve signal, even by accident? On Mon, Jul 09, 2012 at 04:06:16PM +0545, Roshan Karki wrote: > I tried with antenna change but as you told, didn't work. So I think this is > the dead end. Thank you for your help. But one question I wonder is in Sugar I > can use very poor network very well. Why not use the same mechanism in OFW as > well? > > On Fri, Jul 6, 2012 at 10:41 AM, James Cameron wrote: > > G'day, > > Thanks for the photographs. There's nothing wrong that I can see > either. > > Repair may attempt antenna change, but it is unlikely to be fixed with > only antenna change. > > Perhaps the radio module has been damaged. On XO-1 the module is > soldered down and is impractical to replace. In later models (XO-1.5, > XO-1.75) the module is in a socket. > > -- > James Cameron > http://quozl.linux.org.au/ > > -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
I tried with antenna change but as you told, didn't work. So I think this is the dead end. Thank you for your help. But one question I wonder is in Sugar I can use very poor network very well. Why not use the same mechanism in OFW as well? On Fri, Jul 6, 2012 at 10:41 AM, James Cameron wrote: > G'day, > > Thanks for the photographs. There's nothing wrong that I can see > either. > > Repair may attempt antenna change, but it is unlikely to be fixed with > only antenna change. > > Perhaps the radio module has been damaged. On XO-1 the module is > soldered down and is impractical to replace. In later models (XO-1.5, > XO-1.75) the module is in a socket. > > -- > James Cameron > http://quozl.linux.org.au/ > ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
On Thu, Jul 5, 2012 at 3:21 AM, James Cameron wrote: > On Fri, Jun 29, 2012 at 02:56:38PM -0400, Chris Leonard wrote: >> That is an awesome bit of troubleshooting software for repair >> centers to know about. is this all written up on the wiki >> somewhere? > > http://wiki.laptop.org/go/Antenna_testing Thank you James. Really nice bit of functionality and well documented too :-) . cjl ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
G'day, Thanks for the photographs. There's nothing wrong that I can see either. Repair may attempt antenna change, but it is unlikely to be fixed with only antenna change. Perhaps the radio module has been damaged. On XO-1 the module is soldered down and is impractical to replace. In later models (XO-1.5, XO-1.75) the module is in a socket. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
Sorry for the late reply, I was out on fields. In the mean time I opened up the XO but couldn't find any thing wrong with it. I've uploaded the pictures here, https://docs.google.com/folder/d/0BwLO_x-v4VE4eXpuOTFyVlZsQ3M/edit . Can you please kindly check once and see if you can spot any fault or some other part that I missed to look. However I also compared by opening another working xo but didn't find any differences. On Tue, Jul 3, 2012 at 11:25 AM, James Cameron wrote: > On Tue, Jul 03, 2012 at 11:04:07AM +0545, Roshan Karki wrote: > > I'm sorry I never noticed the period. Actually I was typing one > > command per line and thought . was for period. With dot I'm getting > > 0 as result. > > Sorry about that. The Forth word . pops an item from the stack and > prints it as a number in the current radix. > > Okay, this means that during 300ms of listening on the specified > channel, the wireless device was unable to receive any of the > multicast packets from the sender. > > The transmit power used by the NANDblaster sender is likely to be much > lower than the transmit power used by your access point. The result > of the RF link budget [1] may not allow successful receive. > > > You are right. There is around -20 difference between working and > > this xo. > > Try placing this XO with the antennas 25cm from the sender, with the > antennas of both XO set vertical. > > I'm now convinced you have an antenna cable problem that needs > maintenance. There is likely to be about 20 dB extra loss in the > cables, compared to the working XO. > > As an optimisation, you can also use test-antenna to find which XO has > the best RSSI, and use that XO as the NANDblaster sender. A high RSSI > also means the XO will transmit better, in most cases. > > References: > > 1. http://en.wikipedia.org/wiki/Link_budget > > -- > James Cameron > http://quozl.linux.org.au/ > ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
On Fri, Jun 29, 2012 at 02:56:38PM -0400, Chris Leonard wrote: > That is an awesome bit of troubleshooting software for repair > centers to know about. is this all written up on the wiki > somewhere? http://wiki.laptop.org/go/Antenna_testing -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
On Tue, Jul 03, 2012 at 11:04:07AM +0545, Roshan Karki wrote: > I'm sorry I never noticed the period. Actually I was typing one > command per line and thought . was for period. With dot I'm getting > 0 as result. Sorry about that. The Forth word . pops an item from the stack and prints it as a number in the current radix. Okay, this means that during 300ms of listening on the specified channel, the wireless device was unable to receive any of the multicast packets from the sender. The transmit power used by the NANDblaster sender is likely to be much lower than the transmit power used by your access point. The result of the RF link budget [1] may not allow successful receive. > You are right. There is around -20 difference between working and > this xo. Try placing this XO with the antennas 25cm from the sender, with the antennas of both XO set vertical. I'm now convinced you have an antenna cable problem that needs maintenance. There is likely to be about 20 dB extra loss in the cables, compared to the working XO. As an optimisation, you can also use test-antenna to find which XO has the best RSSI, and use that XO as the NANDblaster sender. A high RSSI also means the XO will transmit better, in most cases. References: 1. http://en.wikipedia.org/wiki/Link_budget -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
On Tue, Jul 3, 2012 at 10:16 AM, James Cameron wrote: > On Tue, Jul 03, 2012 at 09:45:56AM +0545, Roshan Karki wrote: > > On Tue, Jul 3, 2012 at 8:35 AM, James Cameron wrote: > > > Is the message "No multicast NAND server" displayed? > > > > Yes, I get this message. > > Thanks. > > Re: the multinand-traffic? command > > > It returns to OK prompt. No 0 or any thing. > > That is a surprise. It may mean an abort happened. But if an abort > happened, then you should not get message "No multicast NAND server". > Confusing. > > Let us debug further together. > > Test 1: check that you typed in the command exactly as shown: > > ok open-wlan 6 multinand-traffic? . close-wlan > > The . (dot, period) is important. It prints the result. > > If dot was typed, proceed to tests below, as these tests are for > finding if an abort happened: > I'm sorry I never noticed the period. Actually I was typing one command per line and thought . was for period. With dot I'm getting 0 as result. > > Test 2: > > ok open-wlan close-wlan ." done" > > Tell me the output? Will be either ok or done ok. > > Test 3: > > ok open-wlan ." opened" > ok 6 multinand-traffic? . ." done" > > Tell me the output? > > Test 4: > > ok open-wlan 6 multinand-traffic? . close-wlan > ok ftrace > > Tell me the output? > > > I also have access point in another room 10m away. I averaged around > > -75 with antenna uncovered. > > Your -75 feels like low signal for me. Try another XO? > You are right. There is around -20 difference between working and this xo. > > > No worries :) . I think I should once send this XO for > > maintenance. Any other checks you want me to do before that? > > I'm not yet convinced you have an antenna problem or a software > problem. Something wrong is happening, but I don't know what it is > yet. > > Do you have another NANDblaster nearby, for different XO model? > We only have one XO model. :) > > For maintenance: check to see if the antenna cable is twisted. > -- > James Cameron > http://quozl.linux.org.au/ > ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
On Tue, Jul 03, 2012 at 09:45:56AM +0545, Roshan Karki wrote: > On Tue, Jul 3, 2012 at 8:35 AM, James Cameron wrote: > > Is the message "No multicast NAND server" displayed? > > Yes, I get this message. Thanks. Re: the multinand-traffic? command > It returns to OK prompt. No 0 or any thing. That is a surprise. It may mean an abort happened. But if an abort happened, then you should not get message "No multicast NAND server". Confusing. Let us debug further together. Test 1: check that you typed in the command exactly as shown: ok open-wlan 6 multinand-traffic? . close-wlan The . (dot, period) is important. It prints the result. If dot was typed, proceed to tests below, as these tests are for finding if an abort happened: Test 2: ok open-wlan close-wlan ." done" Tell me the output? Will be either ok or done ok. Test 3: ok open-wlan ." opened" ok 6 multinand-traffic? . ." done" Tell me the output? Test 4: ok open-wlan 6 multinand-traffic? . close-wlan ok ftrace Tell me the output? > I also have access point in another room 10m away. I averaged around > -75 with antenna uncovered. Your -75 feels like low signal for me. Try another XO? > No worries :) . I think I should once send this XO for > maintenance. Any other checks you want me to do before that? I'm not yet convinced you have an antenna problem or a software problem. Something wrong is happening, but I don't know what it is yet. Do you have another NANDblaster nearby, for different XO model? For maintenance: check to see if the antenna cable is twisted. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
On Tue, Jul 3, 2012 at 8:35 AM, James Cameron wrote: > On Fri, Jun 29, 2012 at 02:19:34PM +0545, Roshan Karki wrote: > > Nandbalsting simply doesn't start in that problematic xo. > > Could you please give more detail on this? > > Please try the nb command manually: > > ok nb > > Is the message "No multicast NAND server" displayed? > Yes, I get this message. > > > no error and I also can connect and use Internet from sugar. Though > > all of the xo in sets have same firmware I even tried updating the > > firmware to latest version to see if it would help. > > > > "open-wlan 6 multinand-traffic? . close-wlan" doesn't return > > . I tried changing server, starting them in different > > channel. What could be the problem here? > > This is a good test, only if your sender is operating on channel 6. > If it is not, the test is not good. > Yes the sender is operating on channel 6. > > You say it doesn't return . What does this test do instead? > > I have tested here. If the multinand-traffic? command is not > returning then it should return 0 (zero). > > If it returns zero, then it is because there was no multicast traffic > received with a valid multicast address. > > Do you see zero, or does it not return to ok prompt at all? > It returns to OK prompt. No 0 or any thing. > > On Mon, Jul 02, 2012 at 10:36:57AM +0545, Roshan Karki wrote: > > I tried doing as you've mentioned. My results are no matter which > > antenna is selected covering the right antenna caused > > decrease(number increased) in average RSSI. Left antenna didn't > > react. There's only around -1 difference in average RSSI between > > antenna 1 and 2. In diversity mode covering any or both antenna > > didn't cause any change in average RSSI. > > > > BTW pressing 'd' allowed me to go back to diversity mode. > > On Mon, Jul 02, 2012 at 10:53:54AM +0545, Roshan Karki wrote: > > I performed the diagnostic in one of the working XO as well. The > > outcome for single antenna selected mode was exact i.e. covering > > left antenna didn't change > > > > anything while covering right did in both antennas. While in > > diversity mode covering any caused average RSSI to decrease (number > > increase). > > Thanks. I have tested with two XO-1 and see identical trends. > > What are your results in average RSSI dB, with antenna uncovered? > I also have access point in another room 10m away. I averaged around -75 with antenna uncovered. > > Here are my results. Numbers are average RSSI dB after 10 seconds. > Access point is in another room 10m away. > > Test start, SKU1 -59, SKU39 -56. > > SKU1 covered left antenna -69, covered right antenna -68, > SKU39 covered left antenna -66, covered right antenna -70. > Conclusion: both antennas are in use. > > press 1, SKU1 -62, SKU39 -54 > SKU1 covered left antenna -62, covered right antenna -82, > SKU39 covered left antenna -54, covered right antenna -71, > Conclusion: right antenna is in exclusive use. > > press d, SKU1 -59, SKU39 -56 > SKU1 covered left antenna -76, covered right antenna -75, > SKU39 covered left antenna -62, covered right antenna -66, > Conclusion: both antennas are in use. > > press 2, SKU1 -62, SKU39 -56 > SKU1 covered left antenna -64, covered right antenna -77, > SKU39 covered left antenna -55, covered right antenna -70, > Conclusion: right antenna is in exclusive use. > > press d, SKU1 -61, SKU39 -57 > SKU1 covered left antenna -74, covered right antenna -66, > SKU39 covered left antenna -63, covered right antenna -69, > Conclusion: both antennas are in use. > > press 0, SKU1 -60, SKU39 -57 > SKU1 covered left antenna -75, covered right antenna -70, > SKU39 covered left antenna -66, covered right antenna -67, > Conclusion: both antennas are in use. > > press d, SKU1 -61, SKU39 -57 > > press l, SKU1 -62, SKU39 -55 > SKU1 covered left antenna -62, covered right antenna -77, > SKU39 covered left antenna -55, covered right antenna -69, > Conclusion: right antenna is in exclusive use. > > press d, SKU1 -60, SKU39 -57 > > press r, SKU1 -62, SKU39 -54 > SKU1 covered left antenna -62, covered right antenna -82, > SKU39 covered left antenna -55, covered right antenna -67, > Conclusion: right antenna is in exclusive use. > > press d, SKU1 -59, SKU39 -56 > > Overall conclusion: XO-1 SKU1 and SKU39 have diversity receive and can > also be configure for right antenna only. The test cannot configure > for left antenna only. > Therefore this test is not very useful to you, sorry! > No worries :) . I think I should once send this XO for maintenance. Any other checks you want me to do before that? > > -- > James Cameron > http://quozl.linux.org.au/ > ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
On Fri, Jun 29, 2012 at 02:19:34PM +0545, Roshan Karki wrote: > Nandbalsting simply doesn't start in that problematic xo. Could you please give more detail on this? Please try the nb command manually: ok nb Is the message "No multicast NAND server" displayed? > no error and I also can connect and use Internet from sugar. Though > all of the xo in sets have same firmware I even tried updating the > firmware to latest version to see if it would help. > > "open-wlan 6 multinand-traffic? . close-wlan" doesn't return > . I tried changing server, starting them in different > channel. What could be the problem here? This is a good test, only if your sender is operating on channel 6. If it is not, the test is not good. You say it doesn't return . What does this test do instead? I have tested here. If the multinand-traffic? command is not returning then it should return 0 (zero). If it returns zero, then it is because there was no multicast traffic received with a valid multicast address. Do you see zero, or does it not return to ok prompt at all? On Mon, Jul 02, 2012 at 10:36:57AM +0545, Roshan Karki wrote: > I tried doing as you've mentioned. My results are no matter which > antenna is selected covering the right antenna caused > decrease(number increased) in average RSSI. Left antenna didn't > react. There's only around -1 difference in average RSSI between > antenna 1 and 2. In diversity mode covering any or both antenna > didn't cause any change in average RSSI. > > BTW pressing 'd' allowed me to go back to diversity mode. On Mon, Jul 02, 2012 at 10:53:54AM +0545, Roshan Karki wrote: > I performed the diagnostic in one of the working XO as well. The > outcome for single antenna selected mode was exact i.e. covering > left antenna didn't change > > anything while covering right did in both antennas. While in > diversity mode covering any caused average RSSI to decrease (number > increase). Thanks. I have tested with two XO-1 and see identical trends. What are your results in average RSSI dB, with antenna uncovered? Here are my results. Numbers are average RSSI dB after 10 seconds. Access point is in another room 10m away. Test start, SKU1 -59, SKU39 -56. SKU1 covered left antenna -69, covered right antenna -68, SKU39 covered left antenna -66, covered right antenna -70. Conclusion: both antennas are in use. press 1, SKU1 -62, SKU39 -54 SKU1 covered left antenna -62, covered right antenna -82, SKU39 covered left antenna -54, covered right antenna -71, Conclusion: right antenna is in exclusive use. press d, SKU1 -59, SKU39 -56 SKU1 covered left antenna -76, covered right antenna -75, SKU39 covered left antenna -62, covered right antenna -66, Conclusion: both antennas are in use. press 2, SKU1 -62, SKU39 -56 SKU1 covered left antenna -64, covered right antenna -77, SKU39 covered left antenna -55, covered right antenna -70, Conclusion: right antenna is in exclusive use. press d, SKU1 -61, SKU39 -57 SKU1 covered left antenna -74, covered right antenna -66, SKU39 covered left antenna -63, covered right antenna -69, Conclusion: both antennas are in use. press 0, SKU1 -60, SKU39 -57 SKU1 covered left antenna -75, covered right antenna -70, SKU39 covered left antenna -66, covered right antenna -67, Conclusion: both antennas are in use. press d, SKU1 -61, SKU39 -57 press l, SKU1 -62, SKU39 -55 SKU1 covered left antenna -62, covered right antenna -77, SKU39 covered left antenna -55, covered right antenna -69, Conclusion: right antenna is in exclusive use. press d, SKU1 -60, SKU39 -57 press r, SKU1 -62, SKU39 -54 SKU1 covered left antenna -62, covered right antenna -82, SKU39 covered left antenna -55, covered right antenna -67, Conclusion: right antenna is in exclusive use. press d, SKU1 -59, SKU39 -56 Overall conclusion: XO-1 SKU1 and SKU39 have diversity receive and can also be configure for right antenna only. The test cannot configure for left antenna only. Therefore this test is not very useful to you, sorry! -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
I performed the diagnostic in one of the working XO as well. The outcome for single antenna selected mode was exact i.e. covering left antenna didn't change anything while covering right did in both antennas. While in diversity mode covering any caused average RSSI to decrease (number increase). On Mon, Jul 2, 2012 at 10:36 AM, Roshan Karki wrote: > Hi James, > > First of all thank you very much for looking at this problem. > > I tried doing as you've mentioned. My results are no matter which antenna > is selected covering the right antenna caused decrease(number increased) in > average RSSI. Left antenna didn't react. There's only around -1 difference > in average RSSI between antenna 1 and 2. In diversity mode covering any or > both antenna didn't cause any change in average RSSI. > > BTW pressing 'd' allowed me to go back to diversity mode. > > > > > On Mon, Jul 2, 2012 at 3:46 AM, James Cameron wrote: > >> On Fri, Jun 29, 2012 at 02:56:38PM -0400, Chris Leonard wrote: >> > That is an awesome bit of troubleshooting software for repair centers >> > to know about. is this all written up on the wiki somewhere? >> >> No, mainly because it isn't in a formal release yet. >> >> -- >> James Cameron >> http://quozl.linux.org.au/ >> > > ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
Hi James, First of all thank you very much for looking at this problem. I tried doing as you've mentioned. My results are no matter which antenna is selected covering the right antenna caused decrease(number increased) in average RSSI. Left antenna didn't react. There's only around -1 difference in average RSSI between antenna 1 and 2. In diversity mode covering any or both antenna didn't cause any change in average RSSI. BTW pressing 'd' allowed me to go back to diversity mode. On Mon, Jul 2, 2012 at 3:46 AM, James Cameron wrote: > On Fri, Jun 29, 2012 at 02:56:38PM -0400, Chris Leonard wrote: > > That is an awesome bit of troubleshooting software for repair centers > > to know about. is this all written up on the wiki somewhere? > > No, mainly because it isn't in a formal release yet. > > -- > James Cameron > http://quozl.linux.org.au/ > ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
On Fri, Jun 29, 2012 at 02:56:38PM -0400, Chris Leonard wrote: > That is an awesome bit of troubleshooting software for repair centers > to know about. is this all written up on the wiki somewhere? No, mainly because it isn't in a formal release yet. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
That is an awesome bit of troubleshooting software for repair centers to know about. is this all written up on the wiki somewhere? cjl On Fri, Jun 29, 2012 at 5:58 AM, James Cameron wrote: > G'day Roshan, > > I understand NANDblaster is not starting on a specific XO-1. > > This might happen if an antenna is damaged; even if you can use > internet from Sugar. Open Firmware configures the wireless card > somewhat differently to Linux. > > We have a new antenna test function in development. This firmware > update contains it: > > http://dev.laptop.org/~quozl/q2f12jf.rom > > You need an open wireless access point to run the test, but there does > not need to be any IP network behind the access point. > > The test is used thus: > > ok essid NETWORK > ok test-antenna > > Where NETWORK is the wireless access point name. > > During the test the following data are displayed: > > - beacon received signal strength indicator, in dB, > > - beacon signal to noise ratio, in dB, > > - receiver noise floor, in dBm, > > - average values of the above three data, > > - receive and transmit antenna allocation. > > For your problem, the "avg rssi" value is the most indicative. You > should find it will react by decreasing (larger negative numbers) as a > result of stowing the antenna and placing a hand over the whole > antenna area. Usually one antenna will be in use, and you can > find which one it is by changing which side of the laptop you cover. > > During the test there are keyboard keys with special meaning: > > - key 1 to select antenna 1, > > - key 2 to select antenna 2, > > - key s to perform network scan, > > - key a to deassociate and reassociate with access point. > > By default on power up reset, both antennas are selected, and this is > called diversity mode, shown as "d". (However, the "d" key does not > effectively restore this mode after keys "1" or "2". A wireless card > reset is probably needed.) > > The test can be used to indicate the performance of antenna, > transmission coax, termination, socket, antenna switch, and receiver. > > But there are sources of unreliability in the test: > > a. the radio noise environment, > > b. the precise position of the antenna or laptop, such as in a null, > > These sources of unreliability make comparisons difficult, but if you > see substantially different values greater than 20 dB between the two > antennas, I suggest one may be damaged. > > The test also does not test transmit by the laptop, only receive. > Since NANDblaster should only receive, this is okay. > > If you identify an antenna as damaged, schedule the laptop for a > detailed repair. Look carefully at the coax cable leading from the > laptop to the axle of the antenna. > > If both antennas are performing fine in this test, then it is time for > me to dig deeper. I'll next be available to do that on Tuesday your > time. > > On Fri, Jun 29, 2012 at 02:19:34PM +0545, Roshan Karki wrote: >> Hi, >> >> Similar to this problem, >> http://lists.laptop.org/pipermail/devel/2009-October/ >> 025769.html nandblasting is not working in one xo while it works flawlessly >> in >> 10 others. >> >> Nandbalsting simply doesn't start in that problematic xo. OFW test-all >> reports >> no error and I also can connect and use Internet from sugar. Though all of >> the >> xo in sets have same firmware I even tried updating the firmware to latest >> version to see if it would help. >> >> "open-wlan 6 multinand-traffic? . close-wlan" doesn't return . I >> tried >> changing server, starting them in different channel. What could be the >> problem >> here? >> >> Thanks. > > -- > James Cameron > http://quozl.linux.org.au/ > ___ > Devel mailing list > Devel@lists.laptop.org > http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Nandblasting not working in one xo
G'day Roshan, I understand NANDblaster is not starting on a specific XO-1. This might happen if an antenna is damaged; even if you can use internet from Sugar. Open Firmware configures the wireless card somewhat differently to Linux. We have a new antenna test function in development. This firmware update contains it: http://dev.laptop.org/~quozl/q2f12jf.rom You need an open wireless access point to run the test, but there does not need to be any IP network behind the access point. The test is used thus: ok essid NETWORK ok test-antenna Where NETWORK is the wireless access point name. During the test the following data are displayed: - beacon received signal strength indicator, in dB, - beacon signal to noise ratio, in dB, - receiver noise floor, in dBm, - average values of the above three data, - receive and transmit antenna allocation. For your problem, the "avg rssi" value is the most indicative. You should find it will react by decreasing (larger negative numbers) as a result of stowing the antenna and placing a hand over the whole antenna area. Usually one antenna will be in use, and you can find which one it is by changing which side of the laptop you cover. During the test there are keyboard keys with special meaning: - key 1 to select antenna 1, - key 2 to select antenna 2, - key s to perform network scan, - key a to deassociate and reassociate with access point. By default on power up reset, both antennas are selected, and this is called diversity mode, shown as "d". (However, the "d" key does not effectively restore this mode after keys "1" or "2". A wireless card reset is probably needed.) The test can be used to indicate the performance of antenna, transmission coax, termination, socket, antenna switch, and receiver. But there are sources of unreliability in the test: a. the radio noise environment, b. the precise position of the antenna or laptop, such as in a null, These sources of unreliability make comparisons difficult, but if you see substantially different values greater than 20 dB between the two antennas, I suggest one may be damaged. The test also does not test transmit by the laptop, only receive. Since NANDblaster should only receive, this is okay. If you identify an antenna as damaged, schedule the laptop for a detailed repair. Look carefully at the coax cable leading from the laptop to the axle of the antenna. If both antennas are performing fine in this test, then it is time for me to dig deeper. I'll next be available to do that on Tuesday your time. On Fri, Jun 29, 2012 at 02:19:34PM +0545, Roshan Karki wrote: > Hi, > > Similar to this problem, http://lists.laptop.org/pipermail/devel/2009-October/ > 025769.html nandblasting is not working in one xo while it works flawlessly in > 10 others. > > Nandbalsting simply doesn't start in that problematic xo. OFW test-all reports > no error and I also can connect and use Internet from sugar. Though all of the > xo in sets have same firmware I even tried updating the firmware to latest > version to see if it would help. > > "open-wlan 6 multinand-traffic? . close-wlan" doesn't return . I tried > changing server, starting them in different channel. What could be the problem > here? > > Thanks. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Nandblasting not working in one xo
Hi, Similar to this problem, http://lists.laptop.org/pipermail/devel/2009-October/025769.html nandblasting is not working in one xo while it works flawlessly in 10 others. Nandbalsting simply doesn't start in that problematic xo. OFW test-all reports no error and I also can connect and use Internet from sugar. Though all of the xo in sets have same firmware I even tried updating the firmware to latest version to see if it would help. "open-wlan 6 multinand-traffic? . close-wlan" doesn't return . I tried changing server, starting them in different channel. What could be the problem here? Thanks. ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
