Using a lock-in amp for weak signals like satcom and radio astronomy?

[rwgast_lowlevellogicdesin]

I wasnt sure if I would get more answers here or in the test equipment forum but im hoping some one here knows a bit about this topic. I found an article about a diy lock in amp and realized I have all the chips to build it, Ive read up alot on it but I haven't seen these used in RF at all and cant figure out why. My idea is to downconvert l band an 70c'ish frequencies down and feed them in to the lock in amp, and use and RF signal gen to feed the carrier input, then feed the data in to an SDR or ADC.

From what I can tell a lock in amp will dig signals out of the noise floor be comparing there phase with the carrier reference, what I dont really understand if I downconvert things like the natural hydrogen line spike or complex signals like qam for a sat, will the lock in amp reconstruct them properly by just feeding it the correct carrier and if so why am I not seeing all kinds of people using these for space comms or HF weak signal, etc. I have a feeling im missing the limits of what this tool can do, i.e I don't want to waste time putting one together if all I can do is dig weak signals out of the noise for simple things like photo transistors and other analog sensors with simple output.

[KE5FX]

I wasnt sure if I would get more answers here or in the test equipment forum but im hoping some one here knows a bit about this topic. I found an article about a diy lock in amp and realized I have all the chips to build it, Ive read up alot on it but I haven't seen these used in RF at all and cant figure out why. My idea is to downconvert l band an 70c'ish frequencies down and feed them in to the lock in amp, and use and RF signal gen to feed the carrier input, then feed the data in to an SDR or ADC.

From what I can tell a lock in amp will dig signals out of the noise floor be comparing there phase with the carrier reference, what I dont really understand if I downconvert things like the natural hydrogen line spike or complex signals like qam for a sat, will the lock in amp reconstruct them properly by just feeding it the correct carrier and if so why am I not seeing all kinds of people using these for space comms or HF weak signal, etc. I have a feeling im missing the limits of what this tool can do, i.e I don't want to waste time putting one together if all I can do is dig weak signals out of the noise for simple things like photo transistors and other analog sensors with simple output.

In the RF world, you'll see the term "synchronous detection" or "coherent detection" used to describe a similar process, and it's indeed very common.  The idea is that it's easier to recover a noisy signal if you can generate a local copy of the carrier to compare it with.  The local carrier can be used to demodulate the signal at a zero IF where the bandwidth can be easily constrained to the minimum necessary. 

In a classical lock-in amplifier, the carrier tends to be easier to recover because you generated it in the first place to provide the stimulus signal to the test apparatus.  The job of the amplifier is to recover the difference signal for phase and/or amplitude comparison.  In a synchronous detector, the receiver has to maintain its local copy of the carrier in phase with the noisy remote carrier -- which ideally isn't even transmitted in the first place -- to aid recovery of the intelligence in the sidebands.  An example of the former would be an atomic clock, while high-grade short wave receivers often include the latter capability. 

For hydrogen-line astronomy you're looking for narrow signals near 1420 MHz that are subject to Doppler broadening, so that might be a good application for coherent detection.  In fact, if you read about how a hydrogen maser works, it will probably sound familiar.  A very weak signal at a relatively high RF frequency needs to be recovered in a very narrow bandwidth.  A passive H-maser is a globally-closed lock-in application, while the astronomer would use a locally-closed loop to study a system that's running open-loop in the larger sense.

[kridri]

Do you have a link to this diy lock in amplifier?

[cj]

I don’t think a lock-in amplifier is the right instrument to receiver neutral hydrogen because it can only pick-out one narrow band channel from below the noise floor.
If you want to measure the neutral Hydrogen spectrum with a lock-in amplifier you'll have to do a lot of measurements because the neutral Hydrogen is several hundred kHz in width.
An SDR with appropriate software (FFT + averaging) can pick-out the neutral Hydrogen spectrum from below noise floor.

For more information and software see project section on http://parac.eu/links.htm .
 
CJ

[rwgast_lowlevellogicdesin]

Cool thank you all for the info, I didn't realize a synchronous detector is basically the same thing.

here is a link to the instructables, if you google AD630 lock-in amplifier youll get links to some more detailed pdf files.
http://www.instructables.com/id/Lock-in-Amplifier/

[LaserSteve]

The other name for this is "Autodyne" converter. 

Here are two examples at HF, but are narrowband, so they probably are of no use at the hydrogen line

https://sites.google.com/site/kennnick/autodynereceiver

http://www.vk2zay.net/article/154

A search for "Autodyne Microwave Receiver" yielded some interesting things for radar, near field communications, rangefinders,  and lab measurement gear, but not much for astronomy. There was some interesting things in homodyne receivers as well, but only for millimeter wave astronomy.

Steve

[Alex Nikitin]

Cool thank you all for the info, I didn't realize a synchronous detector is basically the same thing.

It is not quite the same thing  . A lock-in amplification requires first a modulation and than a synchronous demodulation. The main reason to use it is to remove 1/F noise and low frequency drifts and offsets for low level signals. At least when I was working on a radiotelescope (about 30 years ago) a lock-in technique was used in receivers, by modulating the RF signal at the front-end and demodulating it after the detector and LF amplification/filtering.

Cheers

Alex

[rwgast_lowlevellogicdesin]

Good info guys thanks, even though this is all narrowband stuff I was kind of thinking one could sweep the amp from 1420 to 1421 and then put the signals together.