Thanks for the video. One comment -- to be honest, I take a bit of issue with your description of SNR at ~13:00. What you show there on the plot is not SNR but rather a dBc dynamic range measurement. For SNR, it's the integration of the noise. Let's say I have a 10kHz channel. Within that channel I have a +5 dBm tone and -80 dBm/Hz "noise" across the channel. The SNR is therefore 5 - (-80 + 10LOG10^3) = 55 dB. (The dBc measurement would be 85 dB.)
The noise voltage average over time tends to reach to zero but it can be nonzero at any instant. In fact, there is no guarantee that the average will be nonzero unless we measure it for an infinite time.
Dear Dr. Remillard, I am sure you possess invaluable experience in Cryogenic MW measurement. I would be thankful if you consider posting some information in this regard. For instance I understand calibrating a VNA for measuring Q of a cavity sitting in cryogenic fridge is not an easy task. Any comment or tutorial about cryogenic Microwave will be greatly appreciated. Also any tutorial about making cryogenic switches and low noise amplifiers would be great. Thank you for all of your great videos,
(To prevent confusion among other readers, I'll clarify that my comment here has to do with Q and not noise figure.) VNA calibration can be avoided for Q measurement if you use the full-two-port method Q measurement that I describe in the other video. When the coupling is weak, it is still helpful to flatten the background by subtracting out the standing waves in the reflection. Because all you are trying to do is to flatten the background, you can either remove as much of the reflection as possible, up to the feedthroughs, or remove all of it at the resonator connections before sealing the cryostat. Since you are only measuring differences in S11 over a miniscule frequency range, the change in cable attenuation upon cooling doesn't matter. I don't have background with designing cryogenic switches or LNAs. The LNA design was done by other team members. I was the filter designer. Regards
@@stephenremillard1 Thank you very much, it will saves me from lots of hassle. I thought I have to design a cryogenic switch similar to what is used in microchip industry, so I can calibrate the VNA cold. The cavity will be a superconductor one, for calculating the tangent loss of different materials placed inside it at sub-kelvin temperature. I will email you the design of it. It is not mine, but look interesting.
Thanks for the video. One comment -- to be honest, I take a bit of issue with your description of SNR at ~13:00. What you show there on the plot is not SNR but rather a dBc dynamic range measurement. For SNR, it's the integration of the noise. Let's say I have a 10kHz channel. Within that channel I have a +5 dBm tone and -80 dBm/Hz "noise" across the channel. The SNR is therefore 5 - (-80 + 10LOG10^3) = 55 dB. (The dBc measurement would be 85 dB.)
shouldnt it be 10LOG(10^4) = 40 dB since it is a 10 kHz channel ?
@@Berk-lf6ge Yep. Sorry, error on my part.
The noise voltage average over time tends to reach to zero but it can be nonzero at any instant. In fact, there is no guarantee that the average will be nonzero unless we measure it for an infinite time.
Correct. Thanks.
Dear Dr. Remillard, I am sure you possess invaluable experience in Cryogenic MW measurement. I would be thankful if you consider posting some information in this regard. For instance I understand calibrating a VNA for measuring Q of a cavity sitting in cryogenic fridge is not an easy task. Any comment or tutorial about cryogenic Microwave will be greatly appreciated. Also any tutorial about making cryogenic switches and low noise amplifiers would be great.
Thank you for all of your great videos,
(To prevent confusion among other readers, I'll clarify that my comment here has to do with Q and not noise figure.) VNA calibration can be avoided for Q measurement if you use the full-two-port method Q measurement that I describe in the other video. When the coupling is weak, it is still helpful to flatten the background by subtracting out the standing waves in the reflection. Because all you are trying to do is to flatten the background, you can either remove as much of the reflection as possible, up to the feedthroughs, or remove all of it at the resonator connections before sealing the cryostat. Since you are only measuring differences in S11 over a miniscule frequency range, the change in cable attenuation upon cooling doesn't matter.
I don't have background with designing cryogenic switches or LNAs. The LNA design was done by other team members. I was the filter designer.
Regards
@@stephenremillard1 Thank you very much, it will saves me from lots of hassle. I thought I have to design a cryogenic switch similar to what is used in microchip industry, so I can calibrate the VNA cold. The cavity will be a superconductor one, for calculating the tangent loss of different materials placed inside it at sub-kelvin temperature. I will email you the design of it. It is not mine, but look interesting.
🙏🌹