RF Performance WTF?

I do not understand why there seems to be such as performance difference
between phones as folks on this newsgroup have stated numerous times. I
would think that Sprint would provide specifications for phone RF
performance to all phone manufacturers and that the specifications would not
be so wide that you have phones that provide excellent performance while
other phone manufacturer/model is unusable. What is the story here? Doesn't
make sense to me.

"Mij Adyaw" <mijadyaw@nospam.net> wrote in message news:kx8gf.1254$pF.1203@fed1read04...
>I do not understand why there seems to be such as performance difference
> between phones as folks on this newsgroup have stated numerous times. I
> would think that Sprint would provide specifications for phone RF
> performance to all phone manufacturers and that the specifications would not
> be so wide that you have phones that provide excellent performance while
> other phone manufacturer/model is unusable. What is the story here? Doesn't
> make sense to me.
>

I suppose only that Sprint tests prospective new phone models for acceptability,
but probably they do not have sufficient market control to issue RF mandates.
Maybe "acceptability" involves compromises among many different features.
A phone with lots of cool features might squeak through with an inferior antenna.

"Mij Adyaw" <mijadyaw@nospam.net> wrote in message
news:kx8gf.1254$pF.1203@fed1read04...
>I do not understand why there seems to be such as performance difference
>between phones as folks on this newsgroup have stated numerous times. I would
>think that Sprint would provide specifications for phone RF performance to
>all phone manufacturers and that the specifications would not be so wide that
>you have phones that provide excellent performance while other phone
>manufacturer/model is unusable. What is the story here? Doesn't make sense to
>me.

Search dejanews (Google news) for some of the long discussions of this...

It boils down to the following: Sprint does provide specifications to phone
manufacturers and (presumably) all of the phones Sprint sells meet those
specs. However, there are a couple of issues: (1) some phones _significantly_
exceed the minimum specs (as time goes by and fab processes improve, it often
becomes easier and easier for parts to do so), and once those are out on the
market they tend to get a reputation (even to the point of people saying other
phones, that still meet the specs, suck!), (2) a CDMA radio system is complex
enough that -- like computer benchmarks -- there's no single performance
specification you can come up with that covers every conceivable usage
scenario and lets you boil down one phone's performance vs. another's to a
single number, and (3) natural variations in the processes used to make RF ICs
vary over time, and it's at all uncommon to find, say, a batch of RF
amplifiers that are, say, 1dB quieter today than they were yesterday (even
though they both meet the spec). Unlike, say, a CPU where you have to
specifically overclock it to gain a performance advantage (and a processor
sold to you for a certain clock frequency is guaranteed to work at its
specified frequency, and may or may not work faster), in the analog front-end
of the phone you always get an advantage if there is one to be had.

In other words, Sprint can run a test on your particular phone and tell you
whether or not it's in spec. Beyond that, people should just consider
themselves fortunate if then end up with something better!

Also keep in mind that, for the end user, "performance" is much more than just
the "RF performance" of the phone itself. Things like dropped calls can occur
due to software bugs, network infrastructure overload, and even the behavior
of other people moving into and out of the same cell site that you're on. (A
not-so-well known fact about power controlled CDMA systems is that the
system's capacity is still a significant function of how close you are to the
tower -- for a user at the fringe's of a cell site, other users who initially
were close to the tower but who then start moving out towards the fringes will
drop the site's capacity, so sooner or later someone's going to get dropped if
the site was already close to capacity -- even though everyone would have been
able to continue their calls just fine if they simply remained still! The
hand-off algorithms try hard to prevent this situation from occurring, but the
point again is that the system is complex and behaves in a 'statistical'
manner -- the _average_ number of simultaneous calls a cell site can support
is _significantly_ higher than the _worst case_ number of simultaneous calls
it can support.)

Well said Joel ... . If someone does take the time to put an FAQ for this
specific newsgroup, your info should be included.

Bob

"Joel Kolstad" <JKolstad71HatesSpam@yahoo.com> wrote in message
news:11o4kjneqngr73@corp.supernews.com...
> "Mij Adyaw" <mijadyaw@nospam.net> wrote in message
> news:kx8gf.1254$pF.1203@fed1read04...
>>I do not understand why there seems to be such as performance difference
>>between phones as folks on this newsgroup have stated numerous times. I
>>would think that Sprint would provide specifications for phone RF
>>performance to all phone manufacturers and that the specifications would
>>not be so wide that you have phones that provide excellent performance
>>while other phone manufacturer/model is unusable. What is the story here?
>>Doesn't make sense to me.
>
> Search dejanews (Google news) for some of the long discussions of this...
>
> It boils down to the following: Sprint does provide specifications to
> phone manufacturers and (presumably) all of the phones Sprint sells meet
> those specs. However, there are a couple of issues: (1) some phones
> _significantly_ exceed the minimum specs (as time goes by and fab
> processes improve, it often becomes easier and easier for parts to do so),
> and once those are out on the market they tend to get a reputation (even
> to the point of people saying other phones, that still meet the specs,
> suck!), (2) a CDMA radio system is complex enough that -- like computer
> benchmarks -- there's no single performance specification you can come up
> with that covers every conceivable usage scenario and lets you boil down
> one phone's performance vs. another's to a single number, and (3) natural
> variations in the processes used to make RF ICs vary over time, and it's
> at all uncommon to find, say, a batch of RF amplifiers that are, say, 1dB
> quieter today than they were yesterday (even though they both meet the
> spec). Unlike, say, a CPU where you have to specifically overclock it to
> gain a performance advantage (and a processor sold to you for a certain
> clock frequency is guaranteed to work at its specified frequency, and may
> or may not work faster), in the analog front-end of the phone you always
> get an advantage if there is one to be had.
>
> In other words, Sprint can run a test on your particular phone and tell
> you whether or not it's in spec. Beyond that, people should just consider
> themselves fortunate if then end up with something better!
>
> Also keep in mind that, for the end user, "performance" is much more than
> just the "RF performance" of the phone itself. Things like dropped calls
> can occur due to software bugs, network infrastructure overload, and even
> the behavior of other people moving into and out of the same cell site
> that you're on. (A not-so-well known fact about power controlled CDMA
> systems is that the system's capacity is still a significant function of
> how close you are to the tower -- for a user at the fringe's of a cell
> site, other users who initially were close to the tower but who then start
> moving out towards the fringes will drop the site's capacity, so sooner or
> later someone's going to get dropped if the site was already close to
> capacity -- even though everyone would have been able to continue their
> calls just fine if they simply remained still! The hand-off algorithms
> try hard to prevent this situation from occurring, but the point again is
> that the system is complex and behaves in a 'statistical' manner -- the
> _average_ number of simultaneous calls a cell site can support is
> _significantly_ higher than the _worst case_ number of simultaneous calls
> it can support.)
>
>
>

Mij Adyaw wrote:
> I do not understand why there seems to be such as performance difference
> between phones as folks on this newsgroup have stated numerous times.

More correctly, various user are reporting their specific experiences
and performance in their specific environment.

For many users, they may have a specific area that is problematic - it
may work fine at work, but not at home, ergo its a bad phone.

Most users don't have the opportunity to compare multiple phones side by
side to to determine if its the location or handset that is the problem.

More objective observations may be:

* My older Motorola flip phone would rarely work in the bedroom, but
both my Sanyos had no problem. Here I could say the Sanyos were better
than the Motorola.

* My partner's Toshiba (AudioVox) would consistently drop calls on the
west side of our front porch patio table, my Sanyo had no problem. Here
I could say the Sanyo was better than the Toshiba.

* My Sanyo 8100 would consistently beep out of range three times in the
same places along a 30 mile stretch of road, my 4900 only beeped once.
Here I could say the 4900 is better than the 8100.

* My Sanyos would consistently drop calls along a major freeway during
rush hour traffic. Here I could say it was cell overloading and not a
handset problem.