Home |
Search |
Today's Posts |
#11
|
|||
|
|||
effect of cascading LNAs
|
#12
|
|||
|
|||
effect of cascading LNAs
wrote in message ups.com... I'm dealing with a DSSS current signal at 70MHz. Before I start designing anything, I am calculating the signal amplitude. It's going to be 5nArms. If I have a 50ohm resistor, it's -153dBm. I'd be happy to know any solid proof that it's not doable because it's my job now. In theory DSSS signal can work under noise level, but can it be so much? Well, I have no idea of what a DSSS signal is, so I can't be of much help. Whatever it is, it better be in a very narrow bandwidth. At that signal level, it will take a 50 Hz bandwidth to achieve minimum discernable signal (MDS) with a front end NF of 2.4 dB. If you're talking about using autocorrelation to bring it up out of the noise, all bets are off depending on how much time you can sacrifice. If you're planning on bringing this signal to -10 dBm, better prepare for a lot of shielding and decoupling, not to mention filtering and frequency stability at 70 MHz center frequency. W4ZCB |
#13
|
|||
|
|||
effect of cascading LNAs
wrote in message ps.com... I googled and saw GPS signal is -150dBm. So it's detectable, but it's much higher frequency. What if it's 70Mhz? Well, the GPS signal is detected all right, but they sure don't put all their gain at the signal frequency. W4ZCB |
#15
|
|||
|
|||
effect of cascading LNAs
Harold E. Johnson wrote: wrote in message ups.com... I'm dealing with a DSSS current signal at 70MHz. Before I start designing anything, I am calculating the signal amplitude. It's going to be 5nArms. If I have a 50ohm resistor, it's -153dBm. I'd be happy to know any solid proof that it's not doable because it's my job now. In theory DSSS signal can work under noise level, but can it be so much? Well, I have no idea of what a DSSS signal is, so I can't be of much help. Whatever it is, it better be in a very narrow bandwidth. At that signal level, it will take a 50 Hz bandwidth to achieve minimum discernable signal (MDS) with a front end NF of 2.4 dB. If you're talking about using autocorrelation to bring it up out of the noise, all bets are off depending on how much time you can sacrifice. If you're planning on bringing this signal to -10 dBm, better prepare for a lot of shielding and decoupling, not to mention filtering and frequency stability at 70 MHz center frequency. W4ZCB Some comments: DSSS is direct sequence spread spectrum. It is wideband on the front end. It is probably not a valid comparison to compare DSSS detection levels with GPS. GPS encodes much redundancy into the signal to enhance S/N, and unless the DSSS encoder does something similar we can't use the same detection level numbers. To say DSSS works under the noise level assumes some gain due to encoding, and probably also assumes an information bandwidth much less than the spread spectrum bandwidth. Here are the main reasons you can't cascade a lot of gain at the operating frequency. Others have mentioned this--so this is just a summary with clarification: 1. Stability. A lot of gain at a single frequency is difficult to do without causing stability problems. Even careful shielding between stages doesn't`always work. At 70 Mhz, about 40 dB is the best I can do reliably and even then I would use two separate metal enclosures, each with 20 dB of gain. Above, say, 60 dB all kinds of things go wrong--enclosures don't shield well enough, power supplies don't decouple enough, interconnects talk to each other, etc. With single-chip MMICs and LNAs it's really easy to build something and see for yourself. 2. Intermodulation and blocking. Any unwanted signal at the antenna would be amplified--assume you are trying to detect a -150 dBm signal and someone in the house has a wireless headphone, which transmits at about 72 MHz. With 140 dB of gain your amplifiers would saturate completely and pass none of the wanted signal. Radio design uses a careful balance between gain, mixing, AGC and filtering to minimize stability and overload problems. For narrowband signals you want to apply selectivity as soon in the signal train as you can. For spread spectrum you do the same: a filter as narrow as you can get it and still pass the spectrum, then your despreader at the lowest level you can make it, then a narrow band filter that just passes the information, and finally more gain and the detector. 3. Components Some components such as crystal filters and mixers work better at lower signal levels before a lot of gain is applied. This is really just an extension of reason 2. Hope this helps. Glenn Dixon AC7ZN |
#16
|
|||
|
|||
effect of cascading LNAs
Well, I may have to go back to the old question, because it's not
exactly a DSSS depending on cross correlation. What are the possible bad effects when I keep connecting amplifiers? Shorter bandwidth? Smaller IP3? There should be something I don't know. Mike Andrews wrote: On 11 Oct 2006 13:41:07 -0700, wrote in om: wrote: I'm dealing with a DSSS current signal at 70MHz. Before I start designing anything, I am calculating the signal amplitude. It's going to be 5nArms. If I have a 50ohm resistor, it's -153dBm. I'd be happy to know any solid proof that it's not doable because it's my job now. In theory DSSS signal can work under noise level, but can it be so much? Harold E. Johnson wrote: wrote in message ps.com... It seems that I know the Friis equation. I am asking because it seems difficult, and maybe impossible. I've never heard of this high gain before, but I dont' know why I cann't stack amplifiers and I am looking for somebody who can tell me what exactly the problem is: you see the NF is nice, and I can get the SNR if the band is narrow enough. It also seems that I "don't know what I'm doing". So please explain a little if you want me to understand. Thanks. Well, how about a little additional information? What is it that you're trying to accomplish? A gain of 140 dB would amplify a signal of strength 8 x 10^-10 Watts (About the equivalent of a received signal of 10 dB over "S"-9) to a kiloWatt. I don't think you're trying to do that, so what convinces you that you NEED 140 dB of gain? I googled and saw GPS signal is -150dBm. So it's detectable, but it's much higher frequency. What if it's 70Mhz? It probably isn't going to make a lot of difference whether it's 25 cm. or 40 cm; the principles are the same for DSSS decoding. GPS signals make heavy use of PN sequences so that they can be detected using autocorrelation techniques. They're down below the terrestrial noise threshold, and require autocorrelation techniques for synchronization and detection. But you probably know that already. This probably will be true of your solution as well, and especially so at those power levels. If you're going to be receiving and detecting signals at those power levels, you'll need to use something other than *just* amplification to pull the modulation out. One presumes you'll have a known short spreading sequence that you can use to try to sync up on the signal, or (if it's a longer sequence) that you'll have a rough idea of the code epoch, so that you can try to sneak up on it from behind, as it were. The point here is that a _lot_ of the process gain comes from the autocorrelation and despreading processes. You may not need all that analog gain. Getting it up to ... something like -30 dBm or -15 dBm may be enough -- or even overkill. -- Mike Andrews, W5EGO Tired old sysadmin |
#17
|
|||
|
|||
effect of cascading LNAs
Can I know how did you get that number of 50Hz? Thanks.
Harold E. Johnson wrote: wrote in message ups.com... I'm dealing with a DSSS current signal at 70MHz. Before I start designing anything, I am calculating the signal amplitude. It's going to be 5nArms. If I have a 50ohm resistor, it's -153dBm. I'd be happy to know any solid proof that it's not doable because it's my job now. In theory DSSS signal can work under noise level, but can it be so much? Well, I have no idea of what a DSSS signal is, so I can't be of much help. Whatever it is, it better be in a very narrow bandwidth. At that signal level, it will take a 50 Hz bandwidth to achieve minimum discernable signal (MDS) with a front end NF of 2.4 dB. If you're talking about using autocorrelation to bring it up out of the noise, all bets are off depending on how much time you can sacrifice. If you're planning on bringing this signal to -10 dBm, better prepare for a lot of shielding and decoupling, not to mention filtering and frequency stability at 70 MHz center frequency. W4ZCB |
#18
|
|||
|
|||
effect of cascading LNAs
wrote in message oups.com... Can I know how did you get that number of 50Hz? Thanks. Boltzmans constant times temperature KT for any resistor at room temperature =-174 dBm/Hz. That's as good as you get without cryogenic cooling. Add 17 dBm to increase bandwidth from 1 Hz to 50 Hz = -157 dBM/Hz Add 3 more to get to MDS and it's -154 dBm, your signal level. W4ZCB |
#19
|
|||
|
|||
effect of cascading LNAs
|
#20
|
|||
|
|||
effect of cascading LNAs
|
Reply |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
FCC WRC-03 Amateur Rule Changes Now in Effect | Policy | |||
Can Touchlamps damage or effect speakers? | Homebrew | |||
How does rain effect antennas | Antenna | |||
Ku band Dish - effect of distortion | Antenna | |||
Bricks effect in dipole resonance? Help! | Antenna |