The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> switch noise when it is switched off
https://designers-guide.org/forum/YaBB.pl?num=1269373309

Message started by lunren on Mar 23^rd, 2010, 12:41pm

Title: switch noise when it is switched off
Post by lunren on Mar 23^rd, 2010, 12:41pm

Hi All,

I think when a switch is off, it does not generate thermal noise, right?

However, in my simulation, I found that when a switch is off, its thermal noise still dominates the overall noise performance of my circuit. If a switch which is off should not generate noise, how can I get correct noise results? I am thinking to replace the switch with equivalent parasitic cap (like junction cap). In this way, it seems the results are normal, other wise, the noise is like 50mV!!! :o

Any idea about this are welcome.

Title: Re: switch noise when it is switched off
Post by Ken Kundert on Mar 23^rd, 2010, 9:50pm

Is the switch use to control the band of a VCO? If so you might want to look at the Oscillator Puzzle on the Theory page.

-Ken

Title: Re: switch noise when it is switched off
Post by vivkr on Mar 24^th, 2010, 8:41am

Just a thought. What does OFF mean for your switch? How much leakage do you have and what frequency range are you interested in? If for instance, you are working in the lower frequency range (< 1 kHz), then you might be seeing this.

Of course, there might be other reasons why you are seeing this large noise contribution. Perhaps a bit more information might help.

Vivek

Title: Re: switch noise when it is switched off
Post by lunren on Mar 24^th, 2010, 10:25am

Ken Kundert wrote on Mar 23^rd, 2010, 9:50pm:

Is the switched use to control the band of a VCO? If so you might want to look at the Oscillator Puzzle on the Theory page.

-Ken

It is not a VCO, it is a swiched capacitor gain amplifier.

Title: Re: switch noise when it is switched off
Post by lunren on Mar 24^th, 2010, 11:24am

vivkr wrote on Mar 24^th, 2010, 8:41am:

My circuit is a switched capacitor gain amplifier. There are many switches connected to the input of the amplifier, only one of the switches are turned on, others are turned off. Attached please find the schematic and the integrated noise results. With the turned off switches connected to the nodes inp/inn, the results is like 268mV and the noise summary reported that the two turned off switches contribute 97% noise. While with 500fF cap to replace the switches, the noise become 500uV.

I have also tried to sim the AC noise of a simple NMOS connected to a cap (the other side connected to a fix bias voltage). It turns out that even the NMOS is turned off, it still generate a lot noise and the noise depends heavily on the bias level at the other side of the NMOS.

The last picture shows one output noise of another circuit. It seems that there is a step around 10MHz (discontinuity) and I saw this behavior with several foundry's model. I am wondering if it is because of the noise equation discontinuity in the model? How do you deal with this?

Thanks,

Lunren

Title: Re: switch noise when it is switched off
Post by lunren on Mar 24^th, 2010, 11:27am

With switches connected (sorry I don't know how to attach several pictures at a time)

Title: Re: switch noise when it is switched off
Post by lunren on Mar 24^th, 2010, 11:28am

With equivalent caps connected!

Title: Re: switch noise when it is switched off
Post by lunren on Mar 24^th, 2010, 11:29am

Noise curve with discontinuity!

Title: Re: switch noise when it is switched off
Post by loose-electron on Mar 24^th, 2010, 12:37pm

a little detail - a resistance in series with a capacitance can still AC couple noise

Think about that with respect to an opens MOS switch, and its parasitics.

Title: Re: switch noise when it is switched off
Post by lunren on Mar 24^th, 2010, 1:10pm

loose-electron wrote on Mar 24^th, 2010, 12:37pm:

a little detail - a resistance in series with a capacitance can still AC couple noise

Think about that with respect to an opens MOS switch, and its parasitics.

Hi Jerry, what do you mean by "AC couple noise"? Can you explain more about your idea?

Even though there is some similarity. The open switches should not increase the noise from 500uV to 268mV. In another simulations, the noise is still dominated by one open switch, the OTA itself just generates less noise compared to the open switch (only one).

Thanks,

Title: Re: switch noise when it is switched off
Post by Frank Wiedmann on Mar 25^th, 2010, 3:17am

Ken Kundert wrote on Mar 23^rd, 2010, 9:50pm:

Is the switch use to control the band of a VCO? If so you might want to look at the Oscillator Puzzle on the Theory page.

-Ken

Regarding the circuit presented in http://www.designers-guide.org/Theory/puzzle.pdf : We recently had a presentation from Berkeley Design Automation (http://berkeley-da.com/). They have examined this circuit both with their AFS simulator and with another RF simulator (guess which one...). Their AFS simulator did not show any noise contribution from the switched-off transistors. With loosened tolerance settings for reltol and vntol, the other RF simulator also gave results that were quite similar to those from AFS and did not show the additional noise anymore. In contrast, the AFS results did not change when the accuracy settings were tightened. Their customer considered AFS to have the correct results.

Title: Re: switch noise when it is switched off
Post by vivkr on Mar 25^th, 2010, 3:34am

lunren wrote on Mar 24^th, 2010, 11:24am:

While with 500fF cap to replace the switches, the noise become 500uV.

You seem to have very large switches if the equivalent load is 500 fF!

Quote:

I have also tried to sim the AC noise of a simple NMOS connected to a cap (the other side connected to a fix bias voltage). It turns out that even the NMOS is turned off, it still generate a lot noise and the noise depends heavily on the bias level at the other side of the NMOS.

If simulating just a simple switch + cap combination with standard noise analysis gives you a lot of noise, then it is an indication that your switches are not sufficiently OFF in the OFF state. So although the off-resistance of the switches is high, it is not high enough to prevent the noise from entering the frequency range of interest. Incidentally, if you consider a large enough frequency range (from low-f to high-f), then you should see a fixed amount of total integrated noise in this simple setup, which should only depend on the cap and not on the switch anymore.

Look at the noise plot in this simple case across frequency.

You may also want to see how to get the switches to really turn OFF, maybe using other transistors which leak less, or considering pumping the gate to a more negative voltage to turn your switches OFF. Try a simple experiment, where you sweep the gate voltage negative and check the point at which the noise spectrum of the switch disappears from the frequency range you are interested in.

Vivek

Title: Re: switch noise when it is switched off
Post by subgold on Mar 25^th, 2010, 9:35am

of course it is a must to check if the off resistance is high enough.

but i wonder if it is really a design issue or simulation issue. what if you tighten the gmin setting?

Title: Re: switch noise when it is switched off
Post by lunren on Mar 25^th, 2010, 5:37pm

vivkr wrote on Mar 25^th, 2010, 3:34am:

lunren wrote on Mar 24^th, 2010, 11:24am:

While with 500fF cap to replace the switches, the noise become 500uV.

You seem to have very large switches if the equivalent load is 500 fF!
Yes, it is a group of switches, but it should not generate that much high noise.

Quote:

Actually if the gate was connected to much negative voltage, the integrated noise is only 7.8uV (much less than sqrt(KT/C)=20uV) in the simple experiment setup. If you replace the NMOS with ideal resistor from analogLib, you will find that if you set the resistance very high like 10G (the other side just connect to ideal voltage source, which means there is no other noise source except the resistor), the integrated noise is only 0.7uV (where C=10pF). If the resistor is 1M, we could get 21uV which matches KT/C. Any one who can give this some comment?

Title: Re: switch noise when it is switched off
Post by Ken Kundert on Mar 26^th, 2010, 2:03am

Frank,
Concerning your comment on the Oscillator Puzzle and BDA. It seems like you are saying that they are suggesting that effect documented in the oscillator puzzle is an error and that their simulator is not subject to this error. Both claims are dubious. They do not have the circuit I simulated, and just because the circuit they did simulate might have been similar in structure to the one I described, it is not the same circuit. As I pointed out in the paper, the effect is strongly dependent on seemingly insignificant circuit parameters.

I am not saying that the parametric amplification effect I describe could not be an artifact of the simulator. That is always a possibility. But I am not a naive simulator user, and that was the first thing I considered. I ruled it out as an explanation after a lot of digging.

-Ken

Title: Re: switch noise when it is switched off
Post by Frank Wiedmann on Mar 26^th, 2010, 6:04am

They do indeed suggest that the other RF simulator is in error. For the circuit they simulated (whatever it was exactly), the other RF simulator showed a strong increase in noise for tightened tolerance settings whereas their AFS simulator did not.

Title: Re: switch noise when it is switched off
Post by Ken Kundert on Mar 27^th, 2010, 5:35pm

It does not sound like the parametric amplification problem. That effect was independent of the accuracy settings.

-Ken

Title: Re: switch noise when it is switched off
Post by vivkr on Mar 29^th, 2010, 3:07am

ahem! Please forgive me for saying this, but isn't this discussion going a bit off track? I think we want to discuss the problem with the switch noise. I think the side discussion could easily go on in another thread.

Vivek

Title: Re: switch noise when it is switched off
Post by sheldon on Mar 30^th, 2010, 6:02am

Lunren,

I have built a switched capacitor VGA for DSC AFE. It contains a
number of on and off switches. The I ran pss with time domain noise
and calculated the noise at the several time points. Looking at the
noise summary, the dominant noise sources are the input pair,
the current sources for the folded cascodes, and the input common
mode switches.

Questions:
1) The common-mode feedback circuit seems to contribute a lot
of noise. The common-mode feedback were the largest noise
contributors until I turned them off. This makes sense sine the
the common mode has correlation, that is, equally effects each
of the differential outputs.

2) Have not seen the details of your results, could it be that the
switch noise is stored on the capacitor during one phase and
show up in the noise summary during another phase? That is,
the switch noise is stored on the capacitor during sampling
and show up at the output during hold.

In the next two appends, I will attach the noise summary and
the PSS waveform.

Best Regards,

Sheldon

Title: Re: switch noise when it is switched off
Post by sheldon on Mar 30^th, 2010, 6:05am

Noise Summary

Title: Re: switch noise when it is switched off
Post by sheldon on Mar 30^th, 2010, 6:07am

PSS waveform

Title: Re: switch noise when it is switched off
Post by sheldon on Mar 30^th, 2010, 6:31am

Sorry the schematic did not make it.

Title: Re: switch noise when it is switched off
Post by sheldon on Mar 30^th, 2010, 6:32am

Another controlled maneuver into the terrain!

Title: Re: switch noise when it is switched off
Post by lunren on Mar 31^st, 2010, 12:10pm

sheldon wrote on Mar 30^th, 2010, 6:02am:

I should mention that the previous simulation results were gotten with general AC noise sims. However, in Pnoise (option sources or timedomain) sims, it seems that the group of off switches didn't contribute much noise and the dominant noise sources are the input pair and the current mirror (load) of the first stage the amp, the sim results with Pnoise seem to be rational. But in my case, the common mode feedback does not dominate the noise.
I think in switched capacitor circuit, noise indeed will be sampled to cap during one phase and shows up in another phase.