The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> limiting amplifier gain
https://designers-guide.org/forum/YaBB.pl?num=1210062036

Message started by aaron_do on May 6^th, 2008, 1:20am

Title: limiting amplifier gain
Post by aaron_do on May 6^th, 2008, 1:20am

Hi all,

for constant envelope modulation schemes, we can use a limiting amplifier before demodulation to set the signal level. These limiting amplifiers are sometimes staged to get an RSSI at the same time. In order to prevent too much distortion, you need to filter the signal first using a channel filter.

Anyway i've noticed some schemes having limiting amplifiers with over 80 dB gain. The amplifiers are operated in an open loop so linearity is poor. Anyway their job is basically to compress the signal so their intermodulation is bound to be high especially when the signal is large. My question is basically "will this matter"? I understand the channel filter will ease the requirements of the limiter linearity, but with 80 dB gain, you would need an enormous out-of-band rejection in the channel filter....

thanks for any help,
Aaron

Title: Re: limiting amplifier gain
Post by pancho_hideboo on May 6^th, 2008, 1:58am

aaron_do wrote on May 6^th, 2008, 1:20am:

The amplifiers are operated in an open loop so linearity is poor. Anyway their job is basically to compress the signal so their intermodulation is bound to be high especially when the signal is large. My question is basically "will this matter"?

As far as limiter is well designed, limiter produces only odd order harmonics of carrier.
If an input envelope to limiter is already constant, a main lobe of spectrum around carrier will not spread even after limiter.

Although actually limiter produces even order harmonics, you can suppress this with careful design.
If an input envelope to limiter is not constant, a spectrum around carrier after limiter will spread.
But after band limiting by channel selection filter, it comes to have envelope fluctuation again.

Anyway if limiter is well designed, it never causes unrecoverable distortion as far as there is no large interference input to limiter.

aaron_do wrote on May 6^th, 2008, 1:20am:

I understand the channel filter will ease the requirements of the limiter linearity, but with 80 dB gain, you would need an enormous out-of-band rejection in the channel filter....

Large rejections are required at higher order harmonics frequency.

In attached figure, an input envelope to limiter is already constant, a main lobe of spectrum around carrier will not spread even after limiter.

Title: Re: limiting amplifier gain
Post by pancho_hideboo on May 6^th, 2008, 2:13am

AM/AM and AM/PM characteristics of Limiter.

BTW I couldn't apply Cadence's SpectreRF in evaluation of this. >:(
http://www.designers-guide.org/Forum/YaBB.pl?num=1162921214/0#0

It might be possible to evaluate this if current SpectreRF is used.
Maybe simulation speed must be very very very slow.
Anyway I will not use Cadence's SpectreRF. :P8-)

Title: Re: limiting amplifier gain
Post by Ken Kundert on May 6^th, 2008, 4:32pm

I have used SpectreRF to examine the AM/PM characteristics of limiter, and it has worked great, and I have had no trouble with the speed.

I have found that people often have trouble with SpectreRF because they tighten the tolerances when they don't need to. My experience is that the default tolerances are generally work really well, and I rarely have to use anything but the defaults.

Perhaps you are being too conservative in your use of tolerances.

-Ken

Title: Re: limiting amplifier gain
Post by pancho_hideboo on May 6^th, 2008, 5:20pm

Ken Kundert wrote on May 6^th, 2008, 4:32pm:

Perhaps you are being too conservative in your use of tolerances.
-Ken

No, my accuracy setting is not so tight. That is "errpreset=moderate".

SpectreRF can work for simple gentle circuits.

My log envelope detector circuit(RSSI) is composed of limiter, rectifier, bias etc.
Circuit size is fairly large. So it takes much time even for DC analysis.
SpectreRF is not robust at all for such large actual circuits on the contrary to Cadence Advertisement.

I have various simulators working on Cadence ADE.
So I don't have to insist on using SpectreRF at all.

Title: Re: limiting amplifier gain
Post by Ken Kundert on May 6^th, 2008, 10:44pm

That has not been my experience. I have found it to be quite robust.

-Ken

Title: Re: limiting amplifier gain
Post by aaron_do on May 8^th, 2008, 12:02am

Hi pancho_hideboo,

thanks for your reply. Could you tell me what simulation you ran to get the envelope before and after limiting. It seems to be a transient analysis, but how did you get the frequency spectrum? Usually i use dft, but i'm sure it can't get that many points...Also did you use transient noise? Any tips on how to set the parameters?

Last question...you showed the AM/AM and AM/PM characteristics for three harmonics. Isn't the 1st harmonic enough?

thanks,
Aaron