[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

Re: P3D information




Jon:

Since GSM uses GMSK you need the amplifier to be linear over an
admittedly smaller range than those emissions whose envelope to
collapse to zero, otherwise you would get spectral regrowth .   The
Gausian filter advantage would totally disappear.   This is what I
meant in my earlier posting to you about filtering always implying
envelope variation.

You simply cannot take a theoretically infinite bandwidth signal and
throw away much of its (admittedly low power) high frequency components
and not get amplitude variation.      Admittedly,  PSK and QPSK (not
staggered) have places where they have  phase changes of pi (180 degrees)
and the envelope goes through zero.   You need a linear transmitter
of pretty high quality to faithfully reproduce the input waveform.

However, GSM is a combination of FDMA and TDMA.   The systems I
know about combine the individual signals before transmission.   If  the
wideband transmitter did not have a lot of work go into it to do 
linearization, you
would get some pretty nasty mixing products between channels.

One of the great theoretical accomplishments of Qualcomm is the understanding
that you greatly increase capacity of a CDMA system by controlling power in
the handheld.  You are correct that you need the amplifier to be able to 
operate
over this range of powers in order to reproduce the input waveform faithfully.


As to feedforward LPA's, you are preaching to the choir.  HELAPS, possibly
the first concept of feedforward amplifiers that has been used since Oscar 
7, was
invented by Karl Meinzer, DJ4ZC.   I am certain that the many HELAPS amplifiers
built by Werner Haas for various reasons over the years would impress you.
The transmitters on Phase III-D,  if like the previous incarnations,  all 
use Class
E amplifiers and totally seperate the phase information from the amplitude 
information,
and do all of the heavy lifting to high power in a very nonlinear chain and 
then
"plate modulate" the final amplifier with the amplitude information.  This 
is OLD
OLD OLD.

As for increases in capacity,  and other various calculations, while 
complicated,
all pretty much lead one to the same basic conclusions.  The wireless explosion
we are currently undergoing simply could not have happened without a major
increase (3:1 or 4:1) in the channel carrying capacity .    I believe the 
theoretical
results were that Qualcomm-like CDMA under ideal circumstances could increase
capacity upwards of 4:1 but lets say it is 3:1- 4:1.  This is the range of 
numbers
that have been needed to make the massive growth in cellular technology
use economically feasible.    There simply was not enough room to undergo the
growth needed with analog systems as you well know and without major increases
in efficiency, the digital systems, with their massive changes in 
infrastructure,
would not be viable.    If some sacrifice in voice quality is needed for this,
so be it.

At this time, I pay less for two Qualcomm dual mode phones and their use than I
did a single AMPS boat anchor ten years ago.  It is worth the loss of voice
quality over a 25 Khz FM voice channel.    I can't speak from anything but
experience on "road noise", but it has been my observation, from use, that
DSP engine in the Qualcomm phone is doing active suppression.  I might
be wrong on this, but it is my impression that it is doing so.

Bob




At 09:31 PM 10/09/2000 Monday, you wrote:
>on 10/9/00 1:19 PM, Robert McGwier at rwmcgwier@home.com wrote:
>
> > You are giving us a lot of qualitative information and very little meat.
> > What do you mean by your capacity measurements and your comparisons
> > of quality.    Bell Atlantic, now Verizon,  better than quadrupled its
> > capacity
> > using the same bandwidth as AMPS.  They have one of the best 
> infrastructures
> > around and this has contributed greatly to its success on the east coast.
> > Maybe we need some critical details to understand exactly what you're 
> saying.
>
>I think saying that using a 3:1 channel loading is QUITE quantitative.
>
>Here are reasons why capacity can be increased in newer systems over older
>systems:
>
>1.) Digital provides typically a 3:1 channel capacity over analog.  Instead
>of 1 analog voice channel in a 25 kHz segment, we can now multiplex 3
>channels of TDMA voice in 25 KHz.  The analogy breaks down a little for CDMA
>as you are transmitting all of the multiplexed digital channels over about
>1.5 MHz channel (I believe that's the current CDMA standard in use it may be
>off slightly).  With CDMA, you can theoretically get more than a 3:1
>improvement, but by the time you add all of the error correction and slow
>the bit rate down for better voice quality you are stuck with a 3:1 or
>perhaps a 4:1 improvement in capacity (as you stated).
>
>2.) Most of the old analog systems used Class C amplifiers with large cavity
>combiners to filter out the out of band harmonics.  These old beasts were
>manually tuned.  The newer systems use auto-tuned combiners, but they still
>have the same limitations.  And that is that you have to have a certain
>channel spacing to keep each channel out of the passband of its neighbors
>combiner.  However, with a lot of the newer analog and digital systems,
>multi-carrier linear power amplifiers are replacing the older single
>channel, cavity combined systems.  In a feed forward linear PA system, all
>of the channel combining is done at low power rather than at high power.
>All these combined channels are then amplified by the same amplifier
>together.  The result is that you eliminate the high power combiners and you
>can have tighter spacing between channels.  This will also give you more
>channels in a given bandwidth (869 to 894 MHz downlink in the USA to be
>precise).  Feed forward linear PAs are near and dear to my heart as I used
>to design them and now I sell them.  I know how they work backwards and
>forwards.
>
>So combining the added capacity of digital modulation with the improved
>channel spacing of Feed Forward LPAs, we get our improved capacity.
>
>Now, it IS possible to get much higher capacity improvements, but you lose
>your voice quality.  In Motorola's iDEN system that Nextel uses, the 2 way
>radio feature has a 6:1 channel ratio.  6 conversations are multiplexed on
>to 1 25 kHz channel.  However, the voice quality goes way down.  In the
>first generation of the iDEN system, Motorola had the cellular phone feature
>also at the 6:1 ratio.  They found that customers despised the audio quality
>as people are used to a certain telephone quality.  However, for the 2 way
>radio and dispatch end of things, they found the users didn't mind as they
>were used to less than perfect voice quality.
>
>I think that covers it pretty well.....
>
>Need any more?
>
>73,
>
>Jon
>NA9D
>
>-------------------------------------
>Jon Ogden
>NA9D (ex: KE9NA)
>
>Member:  ARRL, AMSAT, DXCC, NRA
>
>http://www.qsl.net/ke9na
>
>"A life lived in fear is a life half lived."


----
Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org



AMSAT Top AMSAT Home