 | | From: | dspguru_wannabe | | Subject: | A small doubt about signal constellation | | Date: | 20 Jan 2005 23:52:09 -0800 |
|
|
 | Hi all,
Why is it that it is evil to have a constellation point on the origin?
Or is it?
|
|
| | From: | Kevin Neilson | | Subject: | Re: A small doubt about signal constellation | | Date: | Fri, 21 Jan 2005 10:35:25 -0700 |
|
|
| dspguru_wannabe wrote:
> Hi all,
>
> Why is it that it is evil to have a constellation point on the origin?
> Or is it?
>
Besides issues with amplifier linearity, I think one good reason is that
the optimal spacing for any 2^N constellation can't have a point in the
middle. You want to maximize the Euclidean distance between the points.
I can't think of a configuration in which this would include having
a point in the middle. Now, maybe if you were making a 5-point
constellation, the ideal configuration would look like QPSK but with a
point in the middle. But nobody uses 5-point constellations.
-Kevin
|
|
| | From: | Jerry Avins | | Subject: | Re: A small doubt about signal constellation | | Date: | Fri, 21 Jan 2005 12:51:43 -0500 |
|
|
| Kevin Neilson wrote:
> dspguru_wannabe wrote:
>
>> Hi all,
>>
>> Why is it that it is evil to have a constellation point on the origin?
>> Or is it?
>>
> Besides issues with amplifier linearity, I think one good reason is that
> the optimal spacing for any 2^N constellation can't have a point in the
> middle. You want to maximize the Euclidean distance between the points.
> I can't think of a configuration in which this would include having a
> point in the middle. Now, maybe if you were making a 5-point
> constellation, the ideal configuration would look like QPSK but with a
> point in the middle. But nobody uses 5-point constellations.
> -Kevin
For seven points, it's nearly perfect. :-)
Jerry
--
Engineering is the art of making what you want from things you can get.
ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
|
|
| | From: | Mike Yarwood | | Subject: | Re: A small doubt about signal constellation | | Date: | Fri, 21 Jan 2005 09:52:13 +0000 (UTC) |
|
|
|
"dspguru_wannabe" wrote in message
news:1106293929.409408.61510@z14g2000cwz.googlegroups.com...
> Hi all,
>
> Why is it that it is evil to have a constellation point on the origin?
> Or is it?
>
Depends on what you had to do to get there but often this will be associated
with a null in the RF energy transmitted and a large dynamic range on the
signal power - when you push this through a power amplifier where you want
maximum DC to RF efficiency with tolerable interference you will have
difficulty keeping spectral regrowth and ISI low.
Best of Luck - Mike
|
|
| | From: | Steve Underwood | | Subject: | Re: A small doubt about signal constellation | | Date: | Fri, 21 Jan 2005 19:04:50 +0800 |
|
|
| This is a multi-part message in MIME format.
--------------080900010703000605030703
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Mike Yarwood wrote:
>"dspguru_wannabe" wrote in message
>news:1106293929.409408.61510@z14g2000cwz.googlegroups.com...
>
>
>>Hi all,
>>
>>Why is it that it is evil to have a constellation point on the origin?
>>Or is it?
>>
>>
>>
>Depends on what you had to do to get there but often this will be associated
>with a null in the RF energy transmitted and a large dynamic range on the
>signal power - when you push this through a power amplifier where you want
>maximum DC to RF efficiency with tolerable interference you will have
>difficulty keeping spectral regrowth and ISI low.
>
>Best of Luck - Mike
>
>
This would be true for things like PSK. For high order QAM there isn't a
whole lot of difference in the dynamics if the symbols close to the
origin actually move onto the origin.
Regards,
Steve
--------------080900010703000605030703
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Mike Yarwood wrote:
"dspguru_wannabe" <gogichen@gmail.com> wrote in message
news:1106293929.409408.61510@z14g2000cwz.googlegroups.com...
Hi all,
Why is it that it is evil to have a constellation point on the origin?
Or is it?
Depends on what you had to do to get there but often this will be associated
with a null in the RF energy transmitted and a large dynamic range on the
signal power - when you push this through a power amplifier where you want
maximum DC to RF efficiency with tolerable interference you will have
difficulty keeping spectral regrowth and ISI low.
Best of Luck - Mike
This would be true for things like PSK. For high order QAM there isn't
a whole lot of difference in the dynamics if the symbols close to the
origin actually move onto the origin.
Regards,
Steve
--------------080900010703000605030703--
|
|
| | From: | Steve Underwood | | Subject: | Re: A small doubt about signal constellation | | Date: | Fri, 21 Jan 2005 17:33:15 +0800 |
|
|
| dspguru_wannabe wrote:
>Hi all,
>
>Why is it that it is evil to have a constellation point on the origin?
>Or is it?
>
>
It isn't evil. Its how most fibre communication occurs. On-off keying is
just one constellation point that is at the origin, and one that isn't. :-)
Regards,
Steve
|
|
