Active Topics

 


Reply
Thread Tools
Mara's Avatar
Posts: 1,310 | Thanked: 820 times | Joined on Mar 2006 @ Irving, TX
#231
I'm a bit sceptical how they are going to fix this in software. The antenna design is pure HW, and no amount of SW fixing can overcome HW limitations.

The only thing that I can think of is that if the iPhone 4 does have multiple antennas they could have receiver diverse function, which basically is a switch that looks signal on different antennas and pass the one that has the best signal to the receiver. However, I'm not aware any actual handsets have this feature. (The basestations do...) If indeed the iPhone 4 does have this then it is possible there is SW glitch that the diverse path is not working.

The diverse can also be applied to transmit side, but that is much more rare and difficult to do. Remember that the cell phone signaling goes both ways: From tower to cell and cell to tower. Both ways have to work. If they somehow fix the receiver (tower to cell), they still have a problem with transmit (cell to tower). This basically will be a seen first as the other end not hearing you. (If I recall there has been reports that the other end is loosing the voice first, indicating that the phone transmit signal is too weak for basestation to "hear" you.)

Let's stay tuned and see what the final solution is. Maybe they just use the accelerometer and proximity sensor and if they detect that someone is holding the phone they "add three bars" to the antenna signal meter, so no users complain no more about signal strength. That will work for 99.x% of typical iPhone users, and then they will go complain to AT&T about crappy network.
 

The Following 2 Users Say Thank You to Mara For This Useful Post:
ysss's Avatar
Posts: 4,384 | Thanked: 5,524 times | Joined on Jul 2007 @ ˙ǝɹǝɥʍou
#232
I wonder if how much settings you can tweak on the 3G radio?
Is there different channels to manually switch to, to optimize the range/power transmission? How about different algorithm of error handling that can be employed to reduce this?

I've read wildly differing report that I think there's more than just the antenna that contributes to this issue. One report says they can make the iPhone 4 lose signal even when it's in the same room as an AT&T repeater, but some also say that they can grip the iPhone 4 til the signal bar disappear but phone calls can still go on clear as day.
__________________
Class .. : Power User
Humor .. : [#####-----] | Alignment: Pragmatist
Patience : [###-------] | Weapon(s): Galaxy Note + BB Bold Touch 9900
Agro ... : [###-------] | Relic(s) : iPhone 4S, Atrix, Milestone, N900, N800, N95, HTC G1, Treos, Zauri, BB 9000, BB 9700, etc

Follow the MeeGo Coding Competition!
 
Posts: 670 | Thanked: 747 times | Joined on Aug 2009 @ Kansas City, Missouri, USA
#233
Originally Posted by Mara View Post
I'm a bit sceptical how they are going to fix this in software. The antenna design is pure HW, and no amount of SW fixing can overcome HW limitations...Let's stay tuned and see what the final solution is.
You're right that this is strictly a HW problem, stupidity pure and simple. Anyone who's ever tried to adjust a set of rabbit ears knows a hand touching an antenna does strange things. This is a textbook case of designing for form over function, the 'cool' factor overruling how well it will actually work. Despite a (somewhat) deserved rep for ease of use, etc. Apple's made a lot of poor decisions like this, since before the friggin' 1-button mouse.

It's already out that Apple's idea of a 'fix' (other than a band) is a software patch which basically does nothing but make the phone search more for a stronger tower signal. That may help little, but of course at a cost of battery power.
__________________
Registered Linux user #266531.
 
ysss's Avatar
Posts: 4,384 | Thanked: 5,524 times | Joined on Jul 2007 @ ˙ǝɹǝɥʍou
#234
Originally Posted by Crashdamage View Post
You're right that this is strictly a HW problem, stupidity pure and simple. Anyone who's ever tried to adjust a set of rabbit ears knows a hand touching an antenna does strange things. This is a textbook case of designing for form over function, the 'cool' factor overruling how well it will actually work. Despite a (somewhat) deserved rep for ease of use, etc. Apple's made a lot of poor decisions like this, since before the friggin' 1-button mouse.

It's already out that Apple's idea of a 'fix' (other than a band) is a software patch which basically does nothing but make the phone search more for a stronger tower signal. That may help little, but of course at a cost of battery power.
Care to provide a factual source to your claim? Especially that last part.
__________________
Class .. : Power User
Humor .. : [#####-----] | Alignment: Pragmatist
Patience : [###-------] | Weapon(s): Galaxy Note + BB Bold Touch 9900
Agro ... : [###-------] | Relic(s) : iPhone 4S, Atrix, Milestone, N900, N800, N95, HTC G1, Treos, Zauri, BB 9000, BB 9700, etc

Follow the MeeGo Coding Competition!
 
Mara's Avatar
Posts: 1,310 | Thanked: 820 times | Joined on Mar 2006 @ Irving, TX
#235
Originally Posted by ysss View Post
I wonder if how much settings you can tweak on the 3G radio?
Is there different channels to manually switch to, to optimize the range/power transmission? How about different algorithm of error handling that can be employed to reduce this?

I've read wildly differing report that I think there's more than just the antenna that contributes to this issue. One report says they can make the iPhone 4 lose signal even when it's in the same room as an AT&T repeater, but some also say that they can grip the iPhone 4 til the signal bar disappear but phone calls can still go on clear as day.
Yes, there are diffferent channels but they are pretty close to each others. I don't think that can solve the issue, though may help few dB... In borderline reception conditions few dB improvement can make day and night difference in BER (Bit Error Rate) and dropped call or "good call". When reception gets weaker the phone supposed to change to lower modulation schemes that are more tolerant to RF signal errors, thought you will lose data throughput. I thnk the best way to test this is to actually have data link up and monitor the throughput speed. (like some speedtest web site...) The voice can still be transferred in pretty crappy network conditions but data will slow down to a crawl much earlier.

The "bars" do not really mean anything. They are just a quick visual indicator to a non-tech people to judge if they have reasonable signal strength. Even with no bars nornally means there is still signal, althought weak. In digital world weak siignal is still perfect as long as it is clean but just "weak". At opposite example the signal can be strong but reception can be bad? How is that possible? Let's say an example that you have multipath reflection (nearby building?) where direct path and reflected path mix. There you have strong RF level but the signal is still "bad" The receiver can not de-modulate the signal due to too high EVM etc... where the constellation points are all over the map. For more info this page may shed some light that it is not only the "bars"...
http://en.wikipedia.org/wiki/Quadrat...ude_modulation

EDIT: This document is actually easier to understand:
http://chapters.scte.org/cascade/QPS...d%2016-QAM.pdf

Last edited by Mara; 2010-06-27 at 15:13. Reason: Better example...
 

The Following 5 Users Say Thank You to Mara For This Useful Post:
ysss's Avatar
Posts: 4,384 | Thanked: 5,524 times | Joined on Jul 2007 @ ˙ǝɹǝɥʍou
#236
Originally Posted by Mara View Post
The "bars" do not really mean anything. They are just a quick visual indicator to a non-tech people to judge if they have reasonable signal strength. Even with no bars nornally means there is still signal, althought weak. In digital world weak siignal is still perfect as long as it is clean but just "weak". At opposite example the signal can be strong but reception can be bad? How is that possible? Let's say an example that you have multipath reflection (nearby building?) where direct path and reflcted path mix. There you have strong RF level but the signal is still "bad" The receiver can not de-modulate the signal due to too high EVM etc... where the constellation points are all over the map. For more info this page may shed some light that it is not only the "bars"...
http://en.wikipedia.org/wiki/Quadrat...ude_modulation
Thanks...

But didn't you just diminish the importance of iPhone 4's issue by this passage? The consistent theme throughout this issue is "grip bottom-left corner" and "lose signal bars"; but the bit about interruption toward actual calls/data transfer activity is a bit mixed. "No bars" don't always mean "No call/data" from the reports...

That's the only thing keeping me from dismissing Apple at this point, heh.
__________________
Class .. : Power User
Humor .. : [#####-----] | Alignment: Pragmatist
Patience : [###-------] | Weapon(s): Galaxy Note + BB Bold Touch 9900
Agro ... : [###-------] | Relic(s) : iPhone 4S, Atrix, Milestone, N900, N800, N95, HTC G1, Treos, Zauri, BB 9000, BB 9700, etc

Follow the MeeGo Coding Competition!
 
Mara's Avatar
Posts: 1,310 | Thanked: 820 times | Joined on Mar 2006 @ Irving, TX
#237
Originally Posted by ysss View Post
Thanks...

But didn't you just diminish the importance of iPhone 4's issue by this passage? The consistent theme throughout this issue is "grip bottom-left corner" and "lose signal bars"; but the bit about interruption toward actual calls/data transfer activity is a bit mixed. "No bars" don't always mean "No call/data" from the reports...

That's the only thing keeping me from dismissing Apple at this point, heh.
Not really? I think the antenna issue is still significant whenever you are not at "perfect" signal strength areas. Let's say that you are in area you see 2-3 bars when phone is not at hand... and if you now drop 4 bars by holding it, the signal gets too weak to detect it from the noise floor. Of course if AT&T signal is always full 5 bars anywhere you use it the antenna issue may not be a factor at all.
 

The Following 2 Users Say Thank You to Mara For This Useful Post:
Posts: 193 | Thanked: 92 times | Joined on May 2010 @ galveston, tx
#238
Originally Posted by Mara View Post
Yes, there are diffferent channels but they are pretty close to each others. I don't think that can solve the issue, though may help few dB... In borderline reception conditions few dB improvement can make day and night difference in BER (Bit Error Rate) and dropped call or "good call". When reception gets weaker the phone supposed to change to lower modulation schemes that are more tolerant to RF signal errors, thought you will lose data throughput. I thnk the best way to test this is to actually have data link up and monitor the throughput speed. (like some speedtest web site...) The voice can still be transferred in pretty crappy network conditions but data will slow down to a crawl much earlier.

The "bars" do not really mean anything. They are just a quick visual indicator to a non-tech people to judge if they have reasonable signal strength. Even with no bars nornally means there is still signal, althought weak. In digital world weak siignal is still perfect as long as it is clean but just "weak". At opposite example the signal can be strong but reception can be bad? How is that possible? Let's say an example that you have multipath reflection (nearby building?) where direct path and reflected path mix. There you have strong RF level but the signal is still "bad" The receiver can not de-modulate the signal due to too high EVM etc... where the constellation points are all over the map. For more info this page may shed some light that it is not only the "bars"...
http://en.wikipedia.org/wiki/Quadrat...ude_modulation

EDIT: This document is actually easier to understand:
http://chapters.scte.org/cascade/QPS...d%2016-QAM.pdf
Just to add to your informative note.. basic 2G GSM uses GMSK, which is simple FSK followed by a Gaussian filter. http://en.wikipedia.org/wiki/Minimum-shift_keying
This modulation is pretty robust toward lower SNR when compared to 3G schemes like QAM / QPSK.

But you are right that it looks like the antenna design is so sensitive to touch that it fails to act like one altogether... as indicated by dropped calls.
 

The Following 5 Users Say Thank You to noipv4 For This Useful Post:
Posts: 692 | Thanked: 264 times | Joined on Dec 2009
#239
Hey guys here are some fixes that don't require any hardware changes

http://www.youtube.com/watch?v=mSyiAD_ruCA

http://www.youtube.com/watch?v=V-rPLjPsstY
__________________
"Impossible is not in the Maemo vocabulary" - Caballero
 
Mara's Avatar
Posts: 1,310 | Thanked: 820 times | Joined on Mar 2006 @ Irving, TX
#240
There is special field test mode that you can enter by dialing *3001#12345# and "call". (Supposed to work at least on older iPhones.) This will show you the real signal strength in numbers. (Likely in dBm.) This may work on new iPhone too and would tell in much more details how much the signal is affectd by touching different spots on the antenna. Anyone with iPhone can check it out...
 

The Following 2 Users Say Thank You to Mara For This Useful Post:
Reply

Tags
bestthreadever, form > function, h8'ers gon' h8, isheeps, worstthreadever


 
Forum Jump


All times are GMT. The time now is 19:52.