Prev: interfacing three adau1701
Next: interesting DSP application and note about engineers
From: Tim Wescott on 17 Feb 2010 02:49
On Tue, 16 Feb 2010 20:10:56 -0800, rgh wrote:

> Hello all
>
> looking for a device to add two signals x(t) and y(t) x(t) - is
> a signal having a bandwidth of upto 5GHz y(t) - more or less
> noise with uniform ditribution
>
> Would appreciate some pointers/links in this direction.
>
> Regards
> rhg

Electrical signals? Optical signals? Source impedances? Load impedance?

If it's electrical on a coax or in a waveguide then the pro-equipment/lab-
equipment solution is to use a gizmo called a "hybrid combiner" suitable
for your desired bandwidth. I _think_ the same term is used for
optical. If you want to combine the signals without power loss then pay
attention -- some hybrid combiners incorporate some attenuation, and any
device at all has an upper power limit that it can handle.

--
www.wescottdesign.com
From: Jerry Avins on 17 Feb 2010 10:50
Eric Jacobsen wrote:
> On 2/16/2010 9:36 PM, Vladimir Vassilevsky wrote:
>>
>>
>> rgh wrote:
>>
>>> Hello all
>>>
>>> looking for a device to add two signals x(t) and y(t)
>>> x(t) - is a signal having a bandwidth of upto 5GHz
>>> y(t) - more or less noise with uniform ditribution
>>>
>>> Would appreciate some pointers/links in this direction.
>>
>>
>> x(t) >---[R]----
>> |----> x(t) + y(t)
>> y(t) >---[R]----
>>
>>
>> VLV
>
> That's two devices.

A packaged op-amp is one device, but a pair of resistors it two? OK.
Combine the resistors into a single package. Call it a center-tapped
resistor.

> He must be looking for a splitter/combiner,
> assuming the signals are electrical and the center frequency is low enough.

It's a nothing problem. If he doesn't know how to combine the signals,
his next question will be about how to use the combination. Someone
handed him a task he doesn't understand. He has my sympathy.

> Hard to say.

Hard to ask. This is an analog question put to a digital newsgroup by
someone who doesn't even understand that noise doesn't influence the
solution.

Jerry
--
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������
From: Darol Klawetter on 17 Feb 2010 11:28
On Feb 16, 10:10 pm, rgh <mailme....(a)rediffmail.com> wrote:
> Hello all
>
>         looking for a device to add two signals x(t) and y(t)
>         x(t) - is a signal having a bandwidth of upto 5GHz
>         y(t) - more or less noise with uniform ditribution
>
>         Would appreciate some pointers/links in this direction.
>
> Regards
> rhg

Many microwave companies sell power combiners/dividers. Below is a
link to a 2 way combiner/divider that works from DC to 6 GHz. What's
the minimum freq you need to support? Do you have isolation and
insertion loss requirements?

http://www.ifengineering.com/html/powerdiv.htm

You can also try minicircuits.com and miteq.com.

Darol Klawetter
From: Jerry Avins on 17 Feb 2010 11:35
Darol Klawetter wrote:
> On Feb 16, 10:10 pm, rgh <mailme....(a)rediffmail.com> wrote:
>> Hello all
>>
>> looking for a device to add two signals x(t) and y(t)
>> x(t) - is a signal having a bandwidth of upto 5GHz
>> y(t) - more or less noise with uniform ditribution
>>
>> Would appreciate some pointers/links in this direction.
>>
>> Regards
>> rhg
>
> Many microwave companies sell power combiners/dividers. Below is a
> link to a 2 way combiner/divider that works from DC to 6 GHz. What's
> the minimum freq you need to support? Do you have isolation and
> insertion loss requirements?
>
> http://www.ifengineering.com/html/powerdiv.htm
>
> You can also try minicircuits.com and miteq.com.

rgh says that he needs a 6 GHz bandwidth. We don't know if that is some
low frequency to 6 GHz, or, say, from 100 t0 106 GHz.

Jerry
--
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������
From: Tim Wescott on 17 Feb 2010 12:35
On Wed, 17 Feb 2010 11:35:41 -0500, Jerry Avins wrote:

> Darol Klawetter wrote:
>> On Feb 16, 10:10 pm, rgh <mailme....(a)rediffmail.com> wrote:
>>> Hello all
>>>
>>> looking for a device to add two signals x(t) and y(t) x(t) -
>>> is a signal having a bandwidth of upto 5GHz y(t) - more or
>>> less noise with uniform ditribution
>>>
>>> Would appreciate some pointers/links in this direction.
>>>
>>> Regards
>>> rhg
>>
>> Many microwave companies sell power combiners/dividers. Below is a
>> link to a 2 way combiner/divider that works from DC to 6 GHz. What's
>> the minimum freq you need to support? Do you have isolation and
>> insertion loss requirements?
>>
>> http://www.ifengineering.com/html/powerdiv.htm
>>
>> You can also try minicircuits.com and miteq.com.
>
> rgh says that he needs a 6 GHz bandwidth. We don't know if that is some
> low frequency to 6 GHz, or, say, from 100 t0 106 GHz.
>
> Jerry

rgh left out a lot, which is why he's getting the "Rorschach test"
responses. Sure would be nice if he'd expand on his question.

--
www.wescottdesign.com
First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: interfacing three adau1701
Next: interesting DSP application and note about engineers