NEWS: Apple sued over hot iPad shutdowns
in [Wireless Internet]
|
Prev: NEWS: Security shortcomings in WPA2 that threaten security of wireless networks
Next: NEWS: Security shortcomings in WPA2 that threaten security ofwireless networks
From: John Navas on 6 Aug 2010 11:34 On Thu, 05 Aug 2010 17:16:57 -0700, in <hlkm56pra2bet8aj0u9gu0tohl98n0ppui(a)4ax.com>, Jeff Liebermann <jeffl(a)cruzio.com> wrote: >On Thu, 05 Aug 2010 16:56:11 -0700, John Navas ><spamfilter1(a)navasgroup.com> wrote: >>I think a persistent setting is changed in the phone when activated. > >Yep. The flag to allow operation might also be on the SIM card. >Difficult to tell yet. The SIM is activated at the carrier. -- John "We have met the enemy and he is us" -Pogo
From: Jeff Liebermann on 6 Aug 2010 11:53 On Fri, 6 Aug 2010 06:50:22 -0700 (PDT), -hh <recscuba_google(a)huntzinger.com> wrote: >Jeff Liebermann <je...(a)cruzio.com> wrote: >> [...] >> >> See: >> <http://802.11junk.com/jeffl/cellular/cell-test.htm> > >A nice start at some objective data; thanks. Y'er welcome. >> Nope. �The iPhone 4 is at least between 6 and 18 times more affected >> by touching the antenna than the worst conventional cell phone I >> tested. �Double that again for the typical cell phone. �It's not a >> problem in strong signal areas, but might drop the call in a weak >> signal area. � >Accepting the postulation that the iPhone4 attenuates more, this fact >alone does not allow us to come to a final conclusion, >unfortunately. It does for me. 30+ years of assorted RF experience allows me to devine several guesses: 1. The iPhone is unusually worse than a "normal" cell phone in its hand on antenna effects. 2. Such unusually large effects suggests that something more than normal antenna detuning is involved. 3. My suspicions point to a regenertative (oscillatory) receiver front end. 4. Apple may be hiding something in its removal of the field test mode program. >Specifically, what is really needed is the net effective >performance. Ok, I'll admit that I tested more VZW phones than AT&T. I'll see if I can find some more AT&T phones. >I'm not saying that I'm trying to deny that there's any >problem, but from what I've seen (apologies: I may have missed seeing >it), the work to date is incomplete. Being incomplete, it is >inappropriate to make conclusionary claims on total net system >performance. I beg to differ. Throwing the entire cellular system into the testing does quite the opposite of what I was trying to demonstrate. The idea is to isolate the problem down to a specific section of the radio. The signal strength measurement tests everything between the antenna and the demodulator, which is more than I would prefer, but the best I can do without ripping an iPhone 4 apart, and probing the guts with test equipment. >Thus, there's no doubt that one antenna system ("A") which gets >degraded by -25 dB is clearly a "worse" antenna attentuation problem >than a second system ("B") that is degraded by -9dB. > >However, what that overlooks is what is the net "End of the Day" >system performance. Converting to scalar units, dB = 10*log(scalar_ratio/10) and scalar_ratio = 10^(dB/10) -25dB = 316 times drop -9dB = 8 times drop I don't think I need to look at external effects to suggest that the iPhone 4 is far worse. >For example, perhaps the reason why A gets attenuated worse is because >its design allows for an otherwise more sensitive system to start >with. Have you considered WHY the iPhone 4 is apparently more sensitive than the others? Most likely, it's the combination of a larger physical antenna (about 2/3 the circumference of the iPhone 4), and general improvements in receiver front end design. However, my guess(tm) is that there's some slight regeneration in the receiver front end. The positive feedback will produce improved sensitivity, but at the expense of stability, which is exactly what killing the regeneration by placing one's hand on the antenna demonstrates. <http://en.wikipedia.org/wiki/Regenerative_circuit> Note that the RF sensitivity of a receiver is almost totally determined by the combination of antenna and the noise figure of the receiver front end. >Notionally, its performance threshholds when not attenuated >could be something like: > >System A: -200dB >System B: -180dB > > >Thus the (still notional) net results for each System after suffering >from attenuation would be: > >System A: -200dB - (-25dB) = -175dB >System B: -180dB - (-9dB) = -172dB > >So which one is better now? I don't think you understand how decibels operate. Enlightenment follows. dB's are a ratio. By adding additional decibels to the comparision of the two above systems, you've multiplied each number by some fixed value (presumably identical in both cases). For example, -6dB is a ratio of 1:4 or 1/4th. If you throw on the rest of the system gain or whatever, presumably these factors are equal for each radio under test. So, the ratio of: -25dB = 316 times -9dB = 8 times and -25 - -9 = 16 dB is the same as just taking the scalar ratios and conveting to dB. 316/8 = 39.5 times dB = 10 log(39.5) = 16 dB Now, muliply both factor by whatever you want, and you'll still get 16dB difference. For example, if I add 100dB to each number: -125dB = 3.16 * 10^-12 -109dB = 7.94 * 10^-10 7.94*10^-10 / 3.16*10^-12 = 39.8 = 16dB results in the exact same 16dB or 40 times worse difference found in the original ratios. The added 100dB just drops out. dB = 10 log(16) = 12dB >FYI, to clarify one final time, I'm *not* trying to claim that the >iPhone4 performs in this fashion. > >All I'm pointing out is that the numbers to date ... with the caveat >of "the ones that I've seen" ... are only addressing the first half of >the question. Where's the second half? There is a 2nd half, but it's not involved in the tests. It's the ability of the cell site to recover from a loss of signal from the handset. Different carriers have different tolerance levels. My guess(tm) is that AT&T has extended the time allowed after loss of carrier before initiating a disconnect, to compensate for the antenna problem. The time could easily be several seconds, long enough for the user to realize their hand is in the wrong place and reposition. >Until we have the full picture, any "System Peformance" conclusions >drawn to date are premature, since they're failing to examine the >entire system. I beg to differ. Adding the rest of the system into the test does not allow me to isolate the antenna problem. Were the rest of the system available for end to end testing, it would merely add some fixed number of decibels to the comparisons, which would still result in the identical ratios as before. >-hh -- Jeff Liebermann jeffl(a)cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
From: Lloyd Parsons on 6 Aug 2010 13:12 In article <enao561t0lu8384i7ckhguul08bvt077j5(a)4ax.com>, John Navas <spamfilter1(a)navasgroup.com> wrote: > On Thu, 05 Aug 2010 20:36:57 -0500, in > <lloydparsons-C72AF3.20365705082010(a)idisk.mac.com>, Lloyd Parsons > <lloydparsons(a)mac.com> wrote: > > >In article <jcom56hvfoojlfl72uj3qnvdb0jom75rau(a)4ax.com>, > > Jeff Liebermann <jeffl(a)cruzio.com> wrote: > > >> Let's play sematics. Why would Lloyd ask John to provide numbers for > >> something that he didn't claim (i.e. overall ownership)? John said: > >> On the contrary -- Android has now moved past iPhone > >> into 2nd place behind RIM in smartphones. > >> which means current sales, not overall user count. John's statement > >> is correct for both the current quarter and 6 month sales. > >> > >> Also, see link and quoted numbers below. The first column is "overall > >> US ownership" which is the number of users in the US by phone > >> operating system. The numbers were extracted from the 2nd and 3rd > >> graphs from the URL below. My notation as it being the 2nd quarter, > >> 2010 is to indicate the ending date. That should be the same as "in > >> numbers in use". > >> > >> >> For overall user figures (2nd quarter, 2010), see: > >> >> <http://blog.nielsen.com/nielsenwire/online_mobile/android-soars-but-iph > >> >> one > >> >> -still-most-desired-as-smartphones-grab-25-of-u-s-mobile-market/> > >> >> > >> >> Overall US 6 months > >> >> ownership sales > >> >> Rim 35% 33% > >> >> iPhone 28% 23% > >> >> Windoze Mobile 15% 11% > >> >> Android 13% 27% > >> >> Palm 3% 1% > >> >> Linux 3% 3% > >> >> Symbian 2% 2% > > > >Jeff, you misread John. His implication was that Android is bigger in > >use than iPhone. But it is easy to misread what John means or implies > >as he has no grasp on business concepts at all. > > I implied nothing -- my post speaks for itself, and was correct. > Your post was intended to give the impression that there are more Android phones out there than iPhones, which is not the case at all. > >Go back a read a few of his posts concerning market share, revenue vs > >profit and others for examples of his lack of basic business concepts. > > Which of those basic concepts do you not understand? I understand all of them, you OTOH, have shown a complete lack of understanding. That's a shame as I'm retired while you are still trying to run a business. -- Lloyd
From: -hh on 6 Aug 2010 13:14 Jeff Liebermann <je...(a)cruzio.com> wrote: > On Fri, 6 Aug 2010 06:50:22 -0700 (PDT), -hh > > <recscuba_goo...(a)huntzinger.com> wrote: > >Jeff Liebermann <je...(a)cruzio.com> wrote: > >> [...] > > >> See: > >> <http://802.11junk.com/jeffl/cellular/cell-test.htm> > > >A nice start at some objective data; thanks. > > Y'er welcome. > > >> Nope. The iPhone 4 is at least between 6 and 18 times more affected > >> by touching the antenna than the worst conventional cell phone I > >> tested. Double that again for the typical cell phone. It's not a > >> problem in strong signal areas, but might drop the call in a weak > >> signal area. > >Accepting the postulation that the iPhone4 attenuates more, this fact > >alone does not allow us to come to a final conclusion, > >unfortunately. > > It does for me. 30+ years of assorted RF experience allows me to > devine several guesses: Understood, but my point is that no matter how good they may be, they're still nevertheless guesses, not objectively verifiable conclusions. > 1. The iPhone is unusually worse than a "normal" cell phone in its > hand on antenna effects. > 2. Such unusually large effects suggests that something more than > normal antenna detuning is involved. > 3. My suspicions point to a regenertative (oscillatory) receiver > front end. > 4. Apple may be hiding something in its removal of the field test > mode program. > > >Specifically, what is really needed is the net effective > >performance. > > Ok, I'll admit that I tested more VZW phones than AT&T. I'll see if I > can find some more AT&T phones. Not my point. My point is that until we have the effective "sensitivity", the magnitude of the dB drop doesn't really matter (per se). > >I'm not saying that I'm trying to deny that there's any > >problem, but from what I've seen (apologies: I may have missed seeing > >it), the work to date is incomplete. Being incomplete, it is > >inappropriate to make conclusionary claims on total net system > >performance. > > I beg to differ. Throwing the entire cellular system into the testing > does quite the opposite of what I was trying to demonstrate. The idea > is to isolate the problem down to a specific section of the radio. Understood - - and my point is that holistically, the one radio alone isn't the complete system. > >Thus, there's no doubt that one antenna system ("A") which gets > >degraded by -25 dB is clearly a "worse" antenna attentuation problem > >than a second system ("B") that is degraded by -9dB. > > >However, what that overlooks is what is the net "End of the Day" > >system performance. > > Converting to scalar units, > dB = 10*log(scalar_ratio/10) > and > scalar_ratio = 10^(dB/10) > > -25dB = 316 times drop > -9dB = 8 times drop But if the net result is that the signal can still be received & used, a "so what?" applies. > I don't think I need to look at external effects to suggest that the > iPhone 4 is far worse. Then that's where our positions differ. Granted, it would be nice (and a lot easier) if a lot of the external factors could be ignored, but reality is rarely so kind. > >For example, perhaps the reason why A gets attenuated worse is because > >its design allows for an otherwise more sensitive system to start > >with. > > Have you considered WHY the iPhone 4 is apparently more sensitive than > the others? Which "more sensitive" performance metric are you referring to? > Most likely, it's the combination of a larger physical > antenna (about 2/3 the circumference of the iPhone 4), and general > improvements in receiver front end design. Both of which would make the "base" sensitivity of the system better (ie, able to pull in a weaker signal). > However, my guess(tm) is > that there's some slight regeneration in the receiver front end. The > positive feedback will produce improved sensitivity, but at the > expense of stability, which is exactly what killing the regeneration > by placing one's hand on the antenna demonstrates. Which is the sensitivity to attenuation (or whatever we want to call it) which degrades the system's sensitivity (ability to pull in a weak signal). > <http://en.wikipedia.org/wiki/Regenerative_circuit> > > Note that the RF sensitivity of a receiver is almost totally > determined by the combination of antenna and the noise figure of the > receiver front end. That doesn't yet directly address my point, which was that a "more sensitive" system may still end up having a net better performance even if it is more easily degraded. > >Notionally, its performance threshholds when not attenuated > >could be something like: > > >System A: -200dB > >System B: -180dB > > >Thus the (still notional) net results for each System after suffering > >from attenuation would be: > > >System A: -200dB - (-25dB) = -175dB > >System B: -180dB - (-9dB) = -172dB > > >So which one is better now? > > I don't think you understand how decibels operate. For these purposes, its little more than a Log scale, where every 3dB's indicates a doubling/halving. FWIW, I'll admit that cellphones are a bit out of my lane, since my primary interest area is in high gain directional systems. > ... So, the ratio of: > -25dB = 316 times > -9dB = 8 times > and > -25 - -9 = 16 dB Agreed on the math ... but: > [...] For example, if I add 100dB to each number: ....but: my point is that this "100dB" isn't necessarily the exact same value for the two transceivers, and that's a factor that can swamp the 16dB differences in attenuation losses into irrelevance. > >FYI, to clarify one final time, I'm *not* trying to claim that the > >iPhone4 performs in this fashion. > > >All I'm pointing out is that the numbers to date ... with the caveat > >of "the ones that I've seen" ... are only addressing the first half of > >the question. Where's the second half? > > There is a 2nd half, but it's not involved in the tests. It's the > ability of the cell site to recover from a loss of signal from the > handset. Different carriers have different tolerance levels. My > guess(tm) is that AT&T has extended the time allowed after loss of > carrier before initiating a disconnect, to compensate for the antenna > problem. The time could easily be several seconds, long enough for > the user to realize their hand is in the wrong place and reposition. That's a third level beyond than what I was trying to refer to. > >Until we have the full picture, any "System Peformance" conclusions > >drawn to date are premature, since they're failing to examine the > >entire system. > > I beg to differ. Adding the rest of the system into the test does not > allow me to isolate the antenna problem. Were the rest of the system > available for end to end testing, it would merely add some fixed > number of decibels to the comparisons, which would still result in the > identical ratios as before. How is it going to be a fixed number of dB, when as you pointed out, the antenna size of one of the receptors is significantly larger than others? Its like you're suggesting that a full wave monopole antenna isn't going to have any difference in its basic reception performance versus a half-wave or a quarter-wave monopole antenna before any other reception performance factors are considered. -hh
From: Wes Groleau on 6 Aug 2010 13:29
On 08-06-2010 09:50, -hh wrote: > All I'm pointing out is that the numbers to date ... with the caveat > of "the ones that I've seen" ... are only addressing the first half of > the question. Where's the second half? Or maybe more than half. Other factors include but are not limited to time-out period, number of retries, quality of noise filtering, presence or absence of logic to re-establish connection, . . . -- Wes Groleau Don't Keep Dying Languages Alive! http://Ideas.Lang-Learn.us/russell?itemid=1436 |