The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Instrumentation Amp Noise
https://designers-guide.org/forum/YaBB.pl?num=1256105820

Message started by aaron_do on Oct 20^th, 2009, 11:16pm

Title: Instrumentation Amp Noise
Post by aaron_do on Oct 20^th, 2009, 11:16pm

Hi all,

I designed a simple 3-opamp instrumentation amp (ins-amp) in order to buffer the output of a low-IF receiver front-end, but when measuring the output noise of the ins-amp, there was a lot of noise between the frequency of 1 MHz and 5 MHz (See the attachment below). This noise did not appear in simulation, and the stability of the ins-amp seemed fine in simulation also. Does anybody have any clues as to where this noise originated, and how I can take care of it in future?

thanks,
Aaron

Title: Re: Instrumentation Amp Noise
Post by thechopper on Oct 21^st, 2009, 7:39pm

Hi Aaron,

It's quite difficult to tell without knowing more details from your design. Is your IA a chopped one?
It looks you have significant flicker noise so I guess it must be a N(P)MOS input diff pair.
Did you check your noise measurement set up to rule out any external noise component at those high frequencies where the noise peak shows up?

If you can provide more details may be an answer could be found.

Best
Tosei

Title: Re: Instrumentation Amp Noise
Post by aaron_do on Oct 22^nd, 2009, 6:42am

Hi Tosei,

thanks for the reply. It is not a chopped design. It was actually quite a quick and dirty design of just three two-stage miller op-amps. The ins-amps were designed to contribute negligible noise to the actual design so I did not fabricate them separately. However, they were powered by a separate supply from the main DUT, and when I powered down the main DUT so that only the ins-amp was on, I got the noise PSD I showed above. The only blocks which were powered were the op-amps in the ins-amps. Since the ins-amp was designed to drive 50-ohm, there was nothing of significance in between the ins-amp and the spectrum analyzer (a DC-block with 1 kHz corner and a cable...).

The first thing that came to mind was some kind of instability, but I figured that if it was really unstable I would see a single sharp and high peak.

thanks,
Aaron