The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Gain step control
https://designers-guide.org/forum/YaBB.pl?num=1241710123

Message started by aaron_do on May 7^th, 2009, 8:28am

Title: Gain step control
Post by aaron_do on May 7^th, 2009, 8:28am

Hi all,

I want to design a variable-gain common-source amplifier with a resistive load. The gain step should be of discrete values (6 dB for example), and the gain should be controllable by the bias voltage of the amplifier. Rather than inputting a reference signal into a matched amplifier and recording the gain step, is there a way to set the gain?

thanks,
Aaron

Title: Re: Gain step control
Post by pancho_hideboo on May 7^th, 2009, 8:35am

aaron_do wrote on May 7^th, 2009, 8:28am:

Rather than inputting a reference signal into a matched amplifier and recording the gain step

What do you mean ?

Your VGA is single-ended or differential ?
What value of dynamic range for VGA do you require ?
What value of "dB_per_step" do you require for VGA ?
What accracy do you require for DR and "dB_per_step" ?

For example, DR(Dynamic Range)=15dB, 15steps, "dB_per_step"=1dB/step

Title: Re: Gain step control
Post by aaron_do on May 7^th, 2009, 9:33am

Hi pancho_hideboo,

at the moment I'm just looking for conceptual ideas. Suppose for arguments sake however, that i would need a dynamic range of 24 dB with 6 dB steps. The VGA would be differential. It actually occurred to me that I can use a peak detector to detect the gain change, but i'm still kind of playing the idea around in my head. The actual situation is somewhat more complicated than what I described.

Most of the AGCs I've seen change a control voltage until the output level is within certain limits. My goal however is to change the control voltage until the gain drops by a desired amount (for example6dB).

I'm merely curious about whether it can be done without the use of a peak detector.

Quote:

What do you mean ?

The first idea that came to me was to design a matched amplifier and inject a reference signal into the amplifier. Then compare the input and output to get the gain you want. However, the absolute value of the gain is not that important to me, so the method I described above (peak detector) is probably better...

cheers,
Aaron

Title: Re: Gain step control
Post by pancho_hideboo on May 7^th, 2009, 9:42am

aaron_do wrote on May 7^th, 2009, 9:33am:

a dynamic range of 24 dB with 6 dB steps.

DR=24dB, Steps=4, Resolution=6dB/step ?

aaron_do wrote on May 7^th, 2009, 9:33am:

The first idea that came to me was to design a matched amplifier and inject a reference signal into the amplifier.
Then compare the input and output to get the gain you want.

You mean a replica amplifier for reference of gain cotrolling which could be small in sizes and currents by scaling ?

Xamp architecture of ADI(Analog Devices Inc) might be helpful.

Title: Re: Gain step control
Post by aaron_do on May 7^th, 2009, 6:39pm

Quote:

You mean a replica amplifier for reference of gain cotrolling which could be small in sizes and currents by scaling ?

Yes

Quote:

Xamp architecture of ADI(Analog Devices Inc) might be helpful.

thanks. I'll have a look.