Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

LNA at low temperature

Status
Not open for further replies.

Darvin1555

Newbie
Joined
Jul 5, 2017
Messages
4
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Location
Russia
Activity points
70
Hello everyone.
Recently my chief has given me task to think about receiving part of our prospective system and additionally told that i should search for LNA with the lowest possible noise. Unfortunatelly im not RF circuits engineer, but nonetheless i have read necessary chapters from popular books like Microwave engineering by Pozar, and nowadays im slightly aware about principles of RF circuits design. But i decided to not design LNA, because i dont have much time and proper experience and i have been looking for appropriate LNA throughout all manufacturers. But i still have some questions. Receiver in the system will work with set of narrowband frequencies (15-30MHz for example.), but this set will be below 100MHz. And as far as im concerned at this frequencies there are a lot of terrestrial noise sources, judging by Internationall Telecommunication Unit recomendations, and this dependece works regardless of location of system (our will be located at bare field far from cities). And questions:
  • Could anyone point article or datasheet where i can find dependence between temperature and NF for LNA for temperature between -200C and -40C? I have seen some articles concerning cryogenic LNAs, but almost all charts with dependencies were measured at -200C, neglecting intermediate temperatures.
  • Is there any point to design cooling sytem to decrease temperature of LNA bellow lowest operating temperature from datasheet and has anyone had any experience concerning reliability of LNA at low temperatures?
Thanks in advance!
 

at this frequencies there are a lot of terrestrial noise sources

Yes exactly, at the short wave frequencies your receiver noise figure is NOT the limiting factor, and you can't improve signal/noise ratio by a better LNA.

almost all charts with dependencies were measured at -200C, neglecting intermediate temperatures.

That is because liquid nitrogen cooling at 77K (-196°C) is rather simple and cheap.
 
I understand that "intermediate" temperatures come into play if you are e.g. using Peltier coolers instead of LN2. I don't know about specific literature, but for a first estimation, you can consider that total noise is comprised of a thermal (~ T) and a temperature independent component. Having an ambient and a LN2 temperature noise specification gives you information about the intermediate range.
 
For frequencies below 100MHz, if you look to the noise levels chart (atmospheric, man made, etc.), you will find that any attempt to decrease the noise figure of the LNA, is pretty much useless.
 

Attachments

  • Noise_below_100MHz.jpg
    Noise_below_100MHz.jpg
    74.6 KB · Views: 101
I'm wondering how solid state device will operate at -200C ?? There must be some special materials for that ..
 

For frequencies below 100MHz, if you look to the noise levels chart (atmospheric, man made, etc.), you will find that any attempt to decrease the noise figure of the LNA, is pretty much useless.
Thanks for reply. Could you tell me where did you get that chart, because few days ago i was looking for similar charts, but i found information about noise power distribution only in one article and in one book?

I understand that "intermediate" temperatures come into play if you are e.g. using Peltier coolers instead of LN2.
You're right in your assumption, my colleague proposed to use Peltier coolers, but i asked about temperature dependence because im not completely sure until what temeperature level i should cool LNA and my colleague told me, that if i want to cool my device with power dissipation just 2-3 watts to temperature lower -50C at +50C ambient, i have to provide almost 1500 watts of electrical power, and bulky metall plate to dissipate heat, and im wondering is it worthy to design cooler just to reduce NF for 0.5dB. About your advice to evaluate NF using given data and assumption that noise of device has linear dependence from temperature. I saw only in one book, that NF has dependence roughly 0.01dB/C, but i decided to ask someone, because i realy affraid such 'simply' relations in nonlinear devices :grin:

Yes exactly, at the short wave frequencies your receiver noise figure is NOT the limiting factor, and you can't improve signal/noise ratio by a better LNA.
Thanks for reply. I made same conclusion and, i hope, my chief will agree with me and i wont have to design too complex system just for fancy-looking nubmers.
 

CMOS works pretty well at 77K. By 40K you will be
needing to do process modifications (PMOS carrier
freeze-out, needs excessive P+ source doping). A
cryo-specific process might make other changes for
the application, like a "buried channel PMOS" that
has low noise for imaging applications. This device
will have inferior drive but at low currents, have a
low noise due to the channel not actually abutting
the gate oxide interface.

It would be a rare application where you found the
reduction in noise worth the added complexity of
such a chiller setup. But if your plan already involves
liquid nitrogen, taking advantage of it might be a
sensible thing.
 

Interesting comment, but, thanks god, i dont have to design my own IC. My task more basic.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top