The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: email@example.com (Thorsten Brandt)
Subject: Re: Variable Gain Amplifier
Date: 19 Oct 2002 18:10:42 -0700
References: <firstname.lastname@example.org> <email@example.com>
NNTP-Posting-Date: 20 Oct 2002 01:10:42 GMT
Jim Thompson wrote in message news:<firstname.lastname@example.org>...
> On 18 Oct 2002 15:56:22 -0700,
> email@example.com (Thorsten Brandt),
> In Newsgroup: sci.electronics.design,
> Article: <firstname.lastname@example.org>,
> Entitled: "Variable Gain Amplifier",
> Wrote the following:
> |i'm looking for help and information in order to design a variable
> |gain amplifier with the following spec :
> |bandwidth : 1 Ghz
> |continously variable gain from 1 to 50
> Gain-bandwidth product of at least 50GHz ?:-) I don't think so.
well,first i didn't say that it must be a one stage approach, so your
gbw is not applicable in the way you do it, correct ;-) - second, if
you look for examples take a look at the already mentioned AD8367.
by the way - ADI has also a nice differential opamp with a fixed 20db
gain and a BW of 1 GHz - you can also design similar ones for yourself
if you like if you have a foundry with an adequate bicmos-process - so
far concernig GBW ...
> |input imedance : 10kohm/2pf
> |output resistance : 50-100 ohm
> |noise density : < 5nV/Sqrt(Hz)
> |best regards
> |Thorsten Brandt
> You can get < 5nV/rt-Hz at the highest gain, but not at unity gain...
> it will be *much* larger.
ok, that's right - but "*much*" means for a well designed broadband
VGA like the CLC522 from National for example that the input referred
noise density doubles approximately. in this case you will have about
13nV/sqrt(Hz) for unity gain and < 7 nV/sqrt(Hz) for gains from 7 to
100. one might think about an intelligent attenuation and
amplification scheme if this scenario would be the one which would fit
the other requirements ...
from tektronix input amplifiers i heard that the avg. input referred
noise density for the input stage for digitizing oscilloscopes is
around 6 nV/sqrt(Hz). as they have a similar vga in the input stage of
some of their designs AFAIK - i think this constraint shouldn't be out
of range. but of course I'm open to reasonable discussion about this
constraint - if we say 5nV/sqrt(Hz) is unreachable but 10nV/sqrt(Hz)
is - please go ahead and tell me how you would do that with respect to
the other constraints.
> ...Jim Thompson
thanks anyway for your comments - if you have any good design idea -
or can guide me what structure of multilpier, gilbert cell or other
VGA-structure might fit best to my needs with of the shelf components
on a multilayer-board - let me know - i would be very glad ;-)
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup