Question about LNA and out-of-band rejection filter
I need some clarification on this quoted specification
"For the purposes of rejecting undesired signals, the active portion of the GPS antenna shall have the following out-of-band rejection filter characteristic requirements as measured at the antenna assembly...
>25 dB at +/- 50 MHz of 1575.42 MHz"
My question is this, let's assume the LNA has a gain of 25 dB and I was measuring the performance of the LNA. Assuming calibrated VNA would I expect signals outside of +/- 50 MHz to be 0dB (25 dB down from the gain) or -25 dB (50 dB down from the gain) in order to meet spec?
I interpreted my requirements such that a -25 dB filter would be sufficient, is this the correct way to think?
thanks in advance
Out-of-band rejection is referenced to the passband. This means 25 dB down from the gain = 0 dB gain in your case.
To give this kind of spec at dB form is not sufficient.
I think the best way of express is give it at dBc form. Ex:
"25 dBc @ 50 MHz of 1575.42 MHz"
This means at +/- 50 MHz of 1575.42 the signal must be 25 dB lower than the desired signal.
So VSWR is right.
I Think this spec is for the filter . if filter has an IL of 0.5 dB , at an offset of 50 mHz it should have a rejection of 25.5 dB.
Nothing to do with your LNA , your LNA will be having a bandwith greater than 100 Mhz and it will amplify both in band 1575.42 MHz and the signal at an offset of 50 MHz almost to the same level.
Test your filters response using a NA for the same
regards
GV
Yes, I also think he is right.
However,the expression maybe better as "-25 dBc @ 50 MHz of 1575.42 MHz"
I think that such spec means that antenna itself must have gain reduction of 25 dB @+/-50MHz without any additional filters. In this case dBc should not be used because this term means dB below carrier. This case hardly related to the VNA measurements because usually VNA do not use for antenna measurement, at least for the case when 25 dB attenuation @ 50 MHz is involved. It is hard to make GPS LNA with good noise and linearity performance with such a narrow frequency response. Also considering that sensitivity of the latest GPS receivers is actually approaches kTB there probably no preselection filter, therefore this spec is related only to antenna. All of mentioned above means that VNA may be used only for LNA characterization and should show actual LNA parameters without antenna and there will be no effect from antenna?s 25 dB attenuation @ +/-50MHz.
Hi,
Generally, this filter design as I think could be a solution with an LC filter or ceramic filter.
If only LC components, it could acheive 25dBc for rejection @ +/- 50MHz. If you want LC component network give 50dBc on this frequency, you will lose much inband noise performance.
Ceramic filter could be expensive but can give good ( such as 50dBc) rejection on outband without losing much inband gain. That could be use for the mirror frequency of mixer or other tough blocking requirement , when you have a system consideration.
To jupitorcuu,
For 1.5GHz ceramic filter is the best and least expensive solution, but if you read initial question it is not a concern for the author.