I*********y
发帖数: 185 | 1 【 以下文字转载自 EE 讨论区 】
【 原文由 IloveWindyy 所发表 】
向牛人请教。
在WDM光网络中，假如某时刻在fiber link上没有有效信号( 但
有noise)，网管能否shutdown其上的光放大器，使得噪声不会继
续传播？
我搞不清楚EDFA究竟算有源器件，还是无源。如果有源应该可以
由网管控制它的开合，可假如象coupler, fiber是passive components,
那就不能控制了吧？
很confused。 | s*****n
发帖数: 536 | 2 Since it is an amplifier, according to energy constant it must
be active---right? :)
Theoretically it can be shut down, worth or not, don't know.
And practically it also can be shut down---but most of concern
of research is real-time gain shape tuning.
Another thing is, when you shut down and some time turn it on,
how long will the EDFA becomes stable. I remember the edfa's response
time is around 1 ms. Don't know if this is the only parameter.
【在 I*********y 的大作中提到】
: 【 以下文字转载自 EE 讨论区 】
: 【 原文由 IloveWindyy 所发表 】
: 向牛人请教。
: 在WDM光网络中，假如某时刻在fiber link上没有有效信号( 但
: 有noise)，网管能否shutdown其上的光放大器，使得噪声不会继
: 续传播？
: 我搞不清楚EDFA究竟算有源器件，还是无源。如果有源应该可以
: 由网管控制它的开合，可假如象coupler, fiber是passive components,
: 那就不能控制了吧？
: 很confused。
| I*********y
发帖数: 185 | 3 Thanks soliton.The reason why I raised this stupid question is
that I read a paper discussing the amplifier placement problem
in PON(Passive Optical network). If the amplifier is an active
component, how can the network still be called PON? Well the
authors are networking people who may not know much about
components, but guess it is still quite a fundamental
mistake, huh?
I also wonder if it's realistic or worthy to shutdown the amplifiers
not in use. For example in a ring topology, if we don't
【在 s*****n 的大作中提到】
: Since it is an amplifier, according to energy constant it must
: be active---right? :)
: Theoretically it can be shut down, worth or not, don't know.
: And practically it also can be shut down---but most of concern
: of research is real-time gain shape tuning.
: Another thing is, when you shut down and some time turn it on,
: how long will the EDFA becomes stable. I remember the edfa's response
: time is around 1 ms. Don't know if this is the only parameter.
| s*****n
发帖数: 536 | 4 Sorry, today I don't have time on PON, which needs google.
You are afraid of lasing in a ring, or quasi-lasing?
Actually I feel EDFA's noise is not a big issue at all--
after some thinking.
You see, we need amplifier due to loss. This in general
prevents quasi-lasing in a ring with a certain amount of los
More important, the edfa is a saturating
device, the output power can not bigger than P_maximum,
seems large in total, however, the power of noise with a
range (the bandwidth of the receiver f
【在 I*********y 的大作中提到】
: Thanks soliton.The reason why I raised this stupid question is
: that I read a paper discussing the amplifier placement problem
: in PON(Passive Optical network). If the amplifier is an active
: component, how can the network still be called PON? Well the
: authors are networking people who may not know much about
: components, but guess it is still quite a fundamental
: mistake, huh?
: I also wonder if it's realistic or worthy to shutdown the amplifiers
: not in use. For example in a ring topology, if we don't
| I*********y
发帖数: 185 | 5 Thanks for detailing your response. You are exactly right that I
tried to prevent the quasi-lasing of the ASE noise in the ring, or
a mesh with loop. And my concern is exactly the possibly degraded
BER performance. It's just my personal thinking that, if the unused
amplifier can be temporarily shutdown until it has active light
signals again, the system may perform better in terms of steady
state BER than no such shutdown is allowed. But it's really just
my idea, even myself feel it a little fis
【在 s*****n 的大作中提到】
: Sorry, today I don't have time on PON, which needs google.
: You are afraid of lasing in a ring, or quasi-lasing?
: Actually I feel EDFA's noise is not a big issue at all--
: after some thinking.
: You see, we need amplifier due to loss. This in general
: prevents quasi-lasing in a ring with a certain amount of los
: More important, the edfa is a saturating
: device, the output power can not bigger than P_maximum,
: seems large in total, however, the power of noise with a
: range (the bandwidth of the receiver f
| I*********y
发帖数: 185 | 6 Oh, regarding that the output power of an EDFA can never exceed a certain
Pmax. Why is that? I checked the commercial specs of many non-AGC EDFA products,
they do have a parameter called "output saturated power", but its definition is
the output power when the amplifier gain drops to a half of the small signal
gain.It by no means implies that it's the upper bound of the amplifier output
power, according to my understanding.
As a matter of fact, Pout increases when Pin increases, even though gain
【在 s*****n 的大作中提到】
: Sorry, today I don't have time on PON, which needs google.
: You are afraid of lasing in a ring, or quasi-lasing?
: Actually I feel EDFA's noise is not a big issue at all--
: after some thinking.
: You see, we need amplifier due to loss. This in general
: prevents quasi-lasing in a ring with a certain amount of los
: More important, the edfa is a saturating
: device, the output power can not bigger than P_maximum,
: seems large in total, however, the power of noise with a
: range (the bandwidth of the receiver f
| s*****n
发帖数: 536 | 7 OK. still energy is constant. The power of signal comes from the power
of pump laser, right? Then how much power a pump laser can provide
actually set the highest power a signal can get, right?
Saturation power just define a region of saturation, within which
the gain is not constant over input power, the fundamental of this
saturation comes from what?----two things, all of the Erbium ion are
excited, or, all of your pump power are used. So there must a upper
limit for the output power, otherwis
【在 I*********y 的大作中提到】
: Oh, regarding that the output power of an EDFA can never exceed a certain
: Pmax. Why is that? I checked the commercial specs of many non-AGC EDFA products,
: they do have a parameter called "output saturated power", but its definition is
: the output power when the amplifier gain drops to a half of the small signal
: gain.It by no means implies that it's the upper bound of the amplifier output
: power, according to my understanding.
: As a matter of fact, Pout increases when Pin increases, even though gain
| s*****n
发帖数: 536 | 8 Personally feel: don't waste time on that EDFA shut down, lasing's first
condition is phase add-up, not possible for noise. Second, if you
want BER, then there must be signal. Another word, if only noise,
nobody measure BER and nobody care about lightpath---as long as
it won't be damaged, and this is also impossible unless you got a
lucky fishing boat.
The beauty of EDFA is transparency, don't try to damage this by
adding control for information flow.
Iknow there is AGC circuit, but Iam not quit
【在 I*********y 的大作中提到】
: Thanks for detailing your response. You are exactly right that I
: tried to prevent the quasi-lasing of the ASE noise in the ring, or
: a mesh with loop. And my concern is exactly the possibly degraded
: BER performance. It's just my personal thinking that, if the unused
: amplifier can be temporarily shutdown until it has active light
: signals again, the system may perform better in terms of steady
: state BER than no such shutdown is allowed. But it's really just
: my idea, even myself feel it a little fis
|
|
