[Update: See post #29]
[Update #2: See links to videos in post #35]
Hello.
I am returning my iPhone 4 tomorrow, and thought I'd share my experience with those who are interested.
First some disclaimers/background.
1) Please don't expect a bunch of pictures/videos of what I'm describing. I didn't have the time or patience. Those of you who have experienced similar issues will know exactly what I'm referring to.
2) I am not an antenna or wireless systems engineer. However, I was once a EE with experience in wireless communications (2-way radios, cell phones). So, I have a modest understanding of the issues I discuss below.
3) I have been an Apple customer for the last seven years, and have only purchased computers from Apple. I say this to dispel any notions that I'm trying to drum up various negative emotions about Apple.
With that out of the way, here's what happened....
This led me to do a few crude tests. Before doing this, I followed AnandTech's trick of carrying over the field test mode from my 3G. This allowed me to know the signal level in dBm (decibels referred to 1 mW). Note that any given reading seemed to fluctuate by +/- 2 dBm; in rare cases the signal fluctuates as much +/- 4 dBm. Regardless, a change greater than or equal to 10 dBm is significant. Each 3 dBm change represents a signal power change by a factor of two, while a 10 dBm change represents a change by an order of magnitude. Without holding the iP4 the signal level in my house was typically -91 dBm, and max 3G data rates are around 1.5 Mbps down and 60 kbps up (uplink sucks out here).
At Home. First, with no case I was able to repeat and confirm the usual death grip effect on the phone by palming it, as well as simply bridging a single finger across the gap on the bottom left. The signal went from -91 dBm to -113 dBm, a change greater than 20 dB, indicating signal loss by two orders of magnitude. The same effect can be had by placing the phone on it's right edge (looking from front) and placing a piece of metal over the gap (in my case, a wedding ring). This has been shown elsewhere with a key.
Next I tried various holding arrangements to see if I could hold the phone in a way that did not lead to a reduction in signal. The only way was to NOT hold any part of the metal band, e.g., by squeezing the front and back glass with two fingers. Not exactly a comfortable position. You see, I could not avoid a significant signal loss in ANY of my normal, non-death-grip, hand positions. Of course the loss depended on where I touched the band, but even the smallest loss was more than 8 dB. I also tested data rates via the FCC app and time for loading Macrumors.com. Data rates did not seem to correlate strongly with signal level when I did not hold the phone. However, any holding position affected these rates, regardless of the starting signal level. These effects are summarized as follows:
The last two tests indicate that the issue is not a matter of DC conduction between the cellular and WiFi antennas. A plausible explanation is capacitive coupling, which then results in a significant detuning of the cell antenna. Imagine a small capacitance between each antenna and the water or your hand. This also agrees with the recent discovery that electrical tape over the gap does not prevent the death grip effect.
Around Town. I also spent some time testing the phone in various places around the Denver area. Signals ranged from -51 dBm (though I could not see any towers) and -107 dBm in a parking garage, all without holding the phone. Interestingly, when the signal was greater than -87 dBm, I could not get the signal to drop appreciably no matter how I held it! Granted, I did not test this aspect exhaustively, but this may be the most surprising outcome of all. It would also explain, in addition to Apple's weird bar-to-signal mapping, why some people never observe a reception issue. The detuning/attenuation only becomes a problem for signal levels lower than about -87 dBm.
In the end, the inability to use the phone reliably in my house is why I'm returning it. A case does not help. I also went to AT&T and got a new SIM card, but that failed to help either. Maybe I got an awesome 3G. Perhaps there is a manufacturing defect in my iP4. Or perhaps more plausibly, this guy hit the nail on the head. Regardless, I'm returning to my trusty 3G until the smoke clears, and it seems like the iP4 (or iP5) is actually a good phone, in addition to being an awesome "everything else". I don't think Apple will drastically change the hardware in response to this, but I do believe that they will try to make minor software (as well as hardware!) mods to help mitigate the worst scenarios. It may mean a slight PCB redesign or adding some small mods to the antennas (on the inside of course
). Regardless, I'm a wait this one out.
Besides, there are simply no nice cases for the iP4 yet!
[Update #2: See links to videos in post #35]
Hello.
I am returning my iPhone 4 tomorrow, and thought I'd share my experience with those who are interested.
First some disclaimers/background.
1) Please don't expect a bunch of pictures/videos of what I'm describing. I didn't have the time or patience. Those of you who have experienced similar issues will know exactly what I'm referring to.
2) I am not an antenna or wireless systems engineer. However, I was once a EE with experience in wireless communications (2-way radios, cell phones). So, I have a modest understanding of the issues I discuss below.
3) I have been an Apple customer for the last seven years, and have only purchased computers from Apple. I say this to dispel any notions that I'm trying to drum up various negative emotions about Apple.
With that out of the way, here's what happened....
- I bought my (refurbished) iPhone 3G January 2009. Despite its slowness lately, it makes for a pretty reliable phone. It is my only phone (no landline), so the phone function of the device is very imporant to me.
- My house has plaster walls with a metal mesh backing. In other words, it's sort of like living in a Faraday cage. My wife's T-Mobile phone cannot make calls from inside. My 3G had no problem in this regard, and could even initiate and maintain calls from down in the basement! (Important since our office is down there). This was true before, but especially so after AT&T switched to 850 MHz in CO. This was one of my favorite aspects of my 3G. Believe what you may about cell phone "bars", but more often than not, I would get 5 bars in the basement, whether I was on iOS3 or iOS4.
- Ordered iPhone 4 online and received it last week.
- The iP4 never showed more than 4 bars in the house on the main level. In the basement, it struggled to get 2 bars. More importantly, I could not initiate calls. I could receive calls, but they would quickly drop after I picked up. This was true no matter how gingerly I held the phone. I then fitted my iSkin case from the 3G on the iP4, but the problem did not go away. In fact, even with the case on I could still kill the bars and stop data transmission with the usual "death grip".
This led me to do a few crude tests. Before doing this, I followed AnandTech's trick of carrying over the field test mode from my 3G. This allowed me to know the signal level in dBm (decibels referred to 1 mW). Note that any given reading seemed to fluctuate by +/- 2 dBm; in rare cases the signal fluctuates as much +/- 4 dBm. Regardless, a change greater than or equal to 10 dBm is significant. Each 3 dBm change represents a signal power change by a factor of two, while a 10 dBm change represents a change by an order of magnitude. Without holding the iP4 the signal level in my house was typically -91 dBm, and max 3G data rates are around 1.5 Mbps down and 60 kbps up (uplink sucks out here).
At Home. First, with no case I was able to repeat and confirm the usual death grip effect on the phone by palming it, as well as simply bridging a single finger across the gap on the bottom left. The signal went from -91 dBm to -113 dBm, a change greater than 20 dB, indicating signal loss by two orders of magnitude. The same effect can be had by placing the phone on it's right edge (looking from front) and placing a piece of metal over the gap (in my case, a wedding ring). This has been shown elsewhere with a key.
Next I tried various holding arrangements to see if I could hold the phone in a way that did not lead to a reduction in signal. The only way was to NOT hold any part of the metal band, e.g., by squeezing the front and back glass with two fingers. Not exactly a comfortable position. You see, I could not avoid a significant signal loss in ANY of my normal, non-death-grip, hand positions. Of course the loss depended on where I touched the band, but even the smallest loss was more than 8 dB. I also tested data rates via the FCC app and time for loading Macrumors.com. Data rates did not seem to correlate strongly with signal level when I did not hold the phone. However, any holding position affected these rates, regardless of the starting signal level. These effects are summarized as follows:
- Death Grip and Bridging Gap with Finger: True 0/0 kbps DL/UL on the FCC app and page loading ground to a halt. Signal level = -113 dBm to -109 dBm.
- Normal Hold Bridging Left and Right Edges with Hand (away from gap): Signal varied between -107 dBm and -99 dBm. DL data rates varied between 100 kbps to 700 kbps.
- Bag of Water: I filled a ziploc bag with tap water and (carefully!) laid it near the left edge of the phone. The bag's exterior was completely dry. Signal and data rates were similar to death grip. Placing bag anywhere else around the cell antenna did not have such a dramatic effect.
- Death Grip with iSkin Case: Signal reduced to -113 dBm. Zero data rate outcomes seemed to occur as often as those with DL rates around 200 kbps.
The last two tests indicate that the issue is not a matter of DC conduction between the cellular and WiFi antennas. A plausible explanation is capacitive coupling, which then results in a significant detuning of the cell antenna. Imagine a small capacitance between each antenna and the water or your hand. This also agrees with the recent discovery that electrical tape over the gap does not prevent the death grip effect.
Around Town. I also spent some time testing the phone in various places around the Denver area. Signals ranged from -51 dBm (though I could not see any towers) and -107 dBm in a parking garage, all without holding the phone. Interestingly, when the signal was greater than -87 dBm, I could not get the signal to drop appreciably no matter how I held it! Granted, I did not test this aspect exhaustively, but this may be the most surprising outcome of all. It would also explain, in addition to Apple's weird bar-to-signal mapping, why some people never observe a reception issue. The detuning/attenuation only becomes a problem for signal levels lower than about -87 dBm.
In the end, the inability to use the phone reliably in my house is why I'm returning it. A case does not help. I also went to AT&T and got a new SIM card, but that failed to help either. Maybe I got an awesome 3G. Perhaps there is a manufacturing defect in my iP4. Or perhaps more plausibly, this guy hit the nail on the head. Regardless, I'm returning to my trusty 3G until the smoke clears, and it seems like the iP4 (or iP5) is actually a good phone, in addition to being an awesome "everything else". I don't think Apple will drastically change the hardware in response to this, but I do believe that they will try to make minor software (as well as hardware!) mods to help mitigate the worst scenarios. It may mean a slight PCB redesign or adding some small mods to the antennas (on the inside of course
). Regardless, I'm a wait this one out. Besides, there are simply no nice cases for the iP4 yet!
