Untangling the iPhone 4 Antenna Issue

It’s now well known that Apple’s new iPhone 4 has a serious antenna problem.

The underlying engineering issues have been well covered, for example herehere and here.

At first Apple said “Don’t hold it that way”.  That didn’t go down too well.

Now Apple have released a carefully worded statement, which you can read here.

We dismantle Apple’s statement to see what we might read between the lines.

What we know

Reception on every phone gets worse when you hold it.

Reception on iPhone 4 gets much worse when you hold it the “wrong” way.

Unheld, iPhone 4’s antenna performance is exquisite, perhaps the current best-in-class.

We also know that the number of bars shown on any iPhone is generally a bit more optimistic than on other phones — but that is not the issue here.  The differences are tiny.  This, as we will see, is a distraction.

What Apple said

We have discovered the cause of the drop in bars

It is simple, surprising, we were stunned!

We found a mistake in the way we calculate how many bars to show

we’ve been showing too many bars!

The big drop in bars was because the high bars we not real

We will fix this by “adopting AT&T’s formula”

What Apple did not say

Apple neither confirmed nor denied that holding the iPhone 4 in a certain way causes a dramatic drop in signal strength.  We know that it does and that it’s a serious defect.

Apple gave no details on the “mistake” they say they found.  In particular, they did not say that the mistake was in any way related to the optimistically high bar count demonstrated by the reverse-engineers.

Apple gave no details on “AT&T’s recently recommended formula”.  In fact, it seems, no details are available, anywhere.  Nada.  (someone please prove me wrong about this!)

Apple did not say that the “mistake” they found was causing the dramatic loss of signal, which of course it isn’t.  Nor did they say that the new formula would fix the problem, which of course it won’t.

Whats really going on?

Steve Jobs knows a serious problem when he sees one, and this is a serious problem.  The purely technical fix will mean re-engineering.  No doubt that is already in progress, but it will take months to move to production.

In the meantime, what to do? Apple must ride through the problem on a PR wave. Jobs can see no other choice.

So what’s the plan?

Jobs must create a distraction.  “We found a bug!  Boy, are we surprised!”

Then he must link the bug to the problem.  This is where it gets complicated, because there is no linkage.

So Job’s intends to create a linkage, by changing what the bars mean!

That will take some PR. Jobs proposes to educate the market about what happens when you hold a phone, and to re-educate us all about what the bars mean.  For this he needs backing, which is where AT&T come in.

Jobs want us to think he means, AT&T, the technical authority on wireless.  He really means, AT&T, our business/marketing partner.  In fact, the iPhone 4 antenna issue has nothing whatsoever to do with AT&T, in either capacity.  AT&T are simply playing the role of grand savior in a grand charade that will carry Apple through this “minor technical hitch”.

In the process, AT&T will gain some badly needed cred, the iPhone 4 will be rescued (by obscuring the real issue) and phone users will become a little more aware about what they can do to get their phone to work better.

To understand the details requires a bit of explanation…

What the bars mean now

The bars show signal strength, but in a dumbed down way.

Signal strength is like the water pressure in your shower.  If there’s not enough pressure, you have a miserable shower, or no shower at all.  So with your phone, if there’s not enough signal, you have a bad connection, or no connection at all.

We can un-dumb things easily, by giving signal strength a number, like everything else we measure.  The units don’t matter.  We’ll call them wits, because they’re sort of like power (watts).

There is a lowest signal level at which the signal is “lost”.  We will call that signal strength 0 (zero) wits.  At this level almost every phone will drop the connection and show no bars.

Then there is a highest needed signal level at which the connection is as close to perfect as it will ever get.  On a 3G phone, this level is about 20 wits.  At and above this level most phones show 5 bars.

What happens between 0 and 20 wits depends on whether you’re talking or downloading.  A voice call needs only a tiny signal, maybe 5 or 6 wits, before it starts to break up.  A data connection behaves differently.  Below 20 wits, the data rate is lowered.  You can still download, but it takes longer and longer, depending on how few wits you have.

In practice the signal strength can get much higher than 20.  Standing right near a cell tower, the signal might be as high as 60.  Yet there is no way the phone can take advantage of the surplus.  There is a (technical) speed limit.  So most phones continue to show full signal strength, typically 5 bars.

That’s a bit like having a speedometer that reads only up to the speed limit.

Could this be the real “mistake” that Apple is referring to?

What the bars will come to mean

Apple and AT&T have agreed (I speculate) to let the bars show true signal strength, from 0 to (say) 60!

This will change the game completely.  The change will be immediate (it’s a tiny software change) and it will be accepted because (a) it is sanctioned by AT&T, (b) the iPhone 4 is an otherwise magnificent phone and (c) lots of other social reasons that Jobs understands perfectly.

How that will “fix” (the perception of) the iPhone 4 problem

The next best thing to fixing the problem is to show the user when he may have a problem.

Holding an iPhone 3GS “naturally” causes a drop of only about 2 wits (ref).

Holding an iPhone 4 in the same (wrong!) way can easily cause a drop of 20 wits or more.

Under the current system, 5 bars may only mean 20 wits.  So while you think you have a strong signal, you may not have enough to survive the grip of death.

Under the new system, a signal of 20 wits may only show up as only one bar.  It won’t seem nearly as strange to lose that one bar by changing your grip.  Having two or more bars will mean you have more than 20 wits, hopefully enough to survive the grip of death.

The rest will be PR history

People with iPhone’s will find themselves with generally fewer bars than their friends, but their phones will work just fine.  Apple will tout how their phones work flawlessly all the way down to one bar, where other phones start to degrade at higher bar counts.  Jobs may even throw in a switch to “use the old formula”, as if to say “we could also pretend we have more signal, if we wanted to.”

All of a sudden the public debate will be about which of the two systems is “right”. Other phone manufacturers will be forced either to follow Apple’s “lead” (and seem to be playing catch up) or to defend the old system.  Either way the attention will no longer be on the iPhone 4 antenna.

The scheme doesn’t have to work.  It just has to work long enough for Apple to fix the hardware.


Bob Cringely wrote an article about this article.   In the discussion that followed, Bob reminded us that “with AT&T losing its U.S. iPhone exclusive in January, Apple doesn’t care what AT&T thinks”.  Bob is right, and I went too far supposing that Jobs had asked AT&T for their opinion.  So I retract that.

Rather it seems that the reference may have been a challenge to AT&T.  My suspicion is that Apple are indeed using something AT&T put out confidentially, but not quite the way AT&T intended it to be used.  Apple seems to be saying, we’re not playing the same old bars game.  So far, it seems, AT&T have not commented, nor has anyone been able to point to any “recently recommended formula” in an AT&T released document.

Apple also said they would make bars 1, 2 and 3 a bit taller.  This reinforced my conjecture that Apple wants users to understand that having few bars is not necessarily a bad thing.  Having more bars just means you can survive all kinds of antenna abuse, like covering the mystical black gap.


Read my self-assessment.

Post a comment or leave a trackback: Trackback URL.


  • Mac Beach  On 7 July 2010 at 1:36 pm

    Excellent article!

    Apple is so much better at software than Microsoft and so much better at hardware than the likes of Dell and HP (for the most part) that it is a totally shame that they attempt to obfuscate those few times when they have failed in some way. By doing so they diminish their own credibility and I place this “personality defect” that the company seems to have squarely on Jobs’ shoulders.

  • Jean Hominal  On 7 July 2010 at 7:46 pm

    Here is a more direct link to Apple’s letter about it:

  • pissed off  On 8 July 2010 at 2:35 am

    This is all so much wasted hot air ATT Drops Calls……. it has lousy service – why do you think it spent millions to counter Verizon ad’s, their service stinks and Job’s admits it, reference his remarks of it will get better by the end of the summer (note he did not specify the year).

    All this specious blabbering lacks a independent test for the facts in a certified lab, the citing above are not useful nor creditable

  • zahadum  On 8 July 2010 at 4:21 am


    this stringent analysis will surely go down in history as one of the most shrewd calls (and astute reads!) in annals of PR!

    well done (to both you & apple).

    too bad most other commentary is so vapid 😦

  • LAViking  On 8 July 2010 at 5:51 am

    While I agree with part of your statement I think your wrong in parts.

    I did a test at the apple store using the 20 display iPhone 4’s. What I found is that 16 of the 20 demo units DID NOT suffer from the “death grip” no matter how long I held them. Those 16 iPhones stayed at 5 bars! 4 of them did drop signal to 1 bar. These 20 phones all share the same technical, geographic and environmental influences. This could easily be something software related to how the phone fails to hand off to the stronger signal tower allowing the bars to drop. That would explain why 16 of the 20 worked perfectly even thought the antenna was bridged. Or, you could argue that there was a manufacturing or component defect in those 4 phones that did drop the signal making it a hardware issue. Either way, I argue that this does not point to a design flaw in the antenna as that would indicate an unforeseen problem that I am pretty sure Apple understood very clearly when they began designing the iPhone4. Apple simply chose to engineer past this physics issue and use a combination of software and design to mitigate the physics challenge as much as possible. It was a trade off so-to-speak, but the unintended part was most likely how the protocol handshaking would happen in the real world between the multiple cell towers or how a component reliability would play into the manufacturing side. Keep in mind, most people have not had the “death grip” problem and my test supports that even in the same conditions 80% of the phones show no problem. Apple needs to address the 20% that do, but they will and they are.

    That is why they are not saying the antenna is faulty, because it is working exactly as intended with complete clarity and understanding of the physics involved.

  • WaltFrench  On 9 July 2010 at 10:04 am

    I thought your “wits” were a decent way to translate the signal strength representation.

    Then you speculated that Apple would somehow show “true signal strength” with the bars, although you did not identify what that would be. I take it that you are NOT proposing one extra bar for every 60-fold increase in strength that would span the range of unusable to as strong as would be likely to encounter. That’d leave 3, 4 and 5 bars as useless—all would have all the call clarity and data speed possible.

    As you (seemingly) know, there is NO standard for what the difference is between any number of bars. And Apple follows competitive phones in picking a (similar) “good as it needs to be” for five bars, then stepping down a bar every time the signal weakens a certain percent from the previous.

    Not only do they now follow the same standard as everybody else to that approximation, they can still adjust their 5-bar standard to match AT&T’s general practice and still follow the standard. There has to be some arbitrary 5-bar point that’s picked. So it seems a bit BS to say they would (soon) show “true signal strength,” when they currently do follow what we understand to be industry practice.

    But you lost me and perhaps other readers that Cringely sent your way, with that snarky, “Apple will tout how their phones work flawlessly all the way down to one bar…” That’s a conjecture that calls into question the objectivity of the rest of your piece.

    By the way, if you don’t hang on Jobs’s every utterance like some of us: the Reality Distortion Field has worked for 30+ years because Jobs is careful about his reputation. I expect that Apple will make some sort of accommodation to this issue, treating it as its Tylenol scare, but will not make fools of themselves by pretending that there is no issue with an antenna that clearly works better in many circumstances (like mine), but is also subject to sharp, and currently non-user-obvious, degradation.

    • Dave  On 9 July 2010 at 11:52 am

      Thanks for the carefully considered comments.
      The “60 times increase per bar” claim is indeed conjecture on my part. I thought I made that clear. I know it’s wild to believe this, and I’ll be happy to be proved wrong by future events.
      It’s just that I can’t see any other explanation for Steve’s explicit references to showing “too many bars” and “four bars that should have been only two”. That implies more than a refinement and something other than AT&T’s “general practice”.
      I’m also very skeptical that this is a software bug in the classical sense, like when a coder wrote + when he meant & or something. I’m extremely skeptical of a bug that affects only 20% of phones, as some have claimed. It’s far more likely that the effect of the antenna detuning depends on which frequency channel the connection is assigned to.
      Your final observation, that the antenna is “subject to sharp, and currently non-user-obvious, degradation” underlies my primary conjecture, that Steve intends to make the degradation obvious to the user. He wants to enable users to take responsibility for their own signal environment. He has made it clear that he will not admit to this as an antenna design error.

  • WaltFrench  On 9 July 2010 at 5:56 pm

    “He has made it clear that he will not admit to this as an antenna design error.”

    Doing so in what I have elsewhere characterized as a smear campaign is awful PR— blood in the water for ’em. And in any case, Apple Legal makes sure that happens pretty seldom.

    And others with specific mobile antenna experience who’ve looked at the issue seem disinclined to call it an error, too. As you know, all engineering is about compromises (or, in 21st century American Business English, “tradeoffs.” Smaller size? Check. Better free-field reception? Check. Help from software that scans for “cleanliness” of signal (probably, absence of multipath) versus simply strength? Check, and now we’re more than a little proud of ourselves. Ability to hold a clean connection better with variations due to movement? Wow!

    Subject to sharp detuning if the antenna segments are bridged? So? Don’t do it!

    You can see Apple’s thinking. A major effort at dealing with the American cell system*, with some new types of user experiences, gets treated by previous antagonists as a huge flaw, as opposed to something that people can leverage to work better for them. I might call it an unwise design choice given the enthusiasm some have for proving themselves smarter than Apple and its customers, but I wouldn’t quite use the word, “error.”

    * Many of us are continually amazed at how our we see better coverage while traveling in rural China (more bars on my RAZR), even miles from a town, than in many metro USA locations. Some blame NIMBY against towers; I focus on the intentionally fragmented fiefdoms the FCC creates in its spectrum allocation process.

  • Katrina  On 19 April 2013 at 8:14 am

    I’m gone to convey my little brother, that he should also pay a visit this blog on regular basis to take updated from most up-to-date news.

    • Dave  On 19 April 2013 at 9:17 am

      Thank you for your interest. Be aware that I haven’t updated this blog in three years!


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: