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.
hi
there are various reasons: the one I know is that with two stages (or more) the gain is spread over the two stages so the whole system is less sensitive to oscilations. A chain with a great gain tends to oscillate but the idea is to divide the gain in different frequencies (RF, IF1 and IF2)
regards
another advantage of Superheterodyne receiver arises if you have to select a narrowband channel (Δf of some kHz) in a frequency band B of several MHz: with Superheterodyne you can split the high Q requirement you would need for the channel filter (Q = B/Δf) betweeen RF and IF filters. You can use a RF filter with a low Q to select the band B (or a tunable filter to select only a part inside the full band) and after IF conversion you can implement a fixed frequency filter (i.e. SAW filter) with very good Q with less cost and better performances than if you do direct selection of the channel inside the band B with a tunable filter with same Q.
Another advantage is that the IF frequencies often are standardised and so there are a lot of components available on the market.
As dch316 says high RF gain without oscilation is easier with a superhetrodyne receiver.
The other main reason is image reception.
Consider a simple direct conversion receiver (not I/Q) which receives an FSK signal using 450.005 to represent a zero and 450.010 to represent one.
You can use a local oscillator of 450.000MHz to convert that to baseband giving 5Khz for a 0 and 10Khz for a one.
Suppose that there is another transmission in the next channel down occupying 449.98-449.99
That is too close to filter out in the bandpass filtering before/after the LNA.
The unwanted signal ends up in your baseband output.
another advantage of Superheterodyne receiver arises if you have to select a narrowband channel (Δf of some kHz) in a frequency band B of several MHz: with Superheterodyne you can split the high Q requirement you would need for the channel filter (Q = B/Δf) betweeen RF and IF filters. You can use a RF filter with a low Q to select the band B (or a tunable filter to select only a part inside the full band) and after IF conversion you can implement a fixed frequency filter (i.e. SAW filter) with very good Q with less cost and better performances than if you do direct selection of the channel inside the band B with a tunable filter with same Q.
Another advantage is that the IF frequencies often are standardised and so there are a lot of components available on the market.
1. Image rejection filter is no different, either down convert to baseband or IF band. We just use a bandpass filter to reject those image channels.
2. An high Q filter is much easier for low pass instead of bandpass. Directly down convert to baseband can make this channel selection easier.
3. as to the components, I think it is the result instead of the cause.
Added after 4 minutes:
throwaway18 said:
As dch316 says high RF gain without oscilation is easier with a superhetrodyne receiver.
The other main reason is image reception.
Consider a simple direct conversion receiver (not I/Q) which receives an FSK signal using 450.005 to represent a zero and 450.010 to represent one.
You can use a local oscillator of 450.000MHz to convert that to baeband giving 5Khz for a 0 and 10Khz for a one.
Suppose that there is another transmission in the next channel down occupying 449.98-449.99
That is too close to filter out in the bandpass filtering before/after the LNA.
The unwanted signal ends up in your baseband output.
Part of mowgli's point is that fixed frequency filters are more convinent in receivers that
can be tuned to different frequencys
>Image rejection filter is no different, either down convert to baseband or IF band.
It is much easier to make a filter that is 12.5KHZ wide at 10.7MHz than a filter
that is to make a filter that is 12.5KHz wide at typical communication frequencys such as 110-174MHz and 400-500MHz.
>We just use a bandpass filter to reject those image channels.
No, you can't.
A cavity filter for 144MHz that is about 30dB down at 100KHz from the centre
is the size of a propane cylinder and requires precise mechanical tuning.
You can get crystal filters for VHF but they are big and expensive and still not
narrow enough to give good rejection of adjacent channels 12.5KHz away.
Look at this page
**broken link removed**
This filter is bigger than some walkie-talkies.
What if you want to make a walkie talkie that can be programmed to cover
any of thousands of channels? Are you going to have thousands of crystal
filters?
>An high Q filter is much easier for low pass instead of bandpass.
>Directly down convert to baseband can make this channel selection easier.
A low pass filter does not help when your wanted baseband signal is at 10KHz and your unwanted signal is also at 10KHz.
>Is there any difference if we use superhetrodyne?
In a direct conversion receiver the image frequencys are adjacent to the wanted frequencys.
In a superhet they are further away. A good superhet receiver has 50dB or more rejection of image freqenucys. A non-IQ direct convertion receiver has zero rejection.
I don't know how much rejection you can get with I/Q but I suspect it is 30-40dB at most.
If it is a true direct conversion receiver, then there is no image problem! If you remember, image is always located at 2*IF away from the desired signal. If IF=0Hz, where is the image?
But we do have DC offset problem to solve, for example, LO leaking to the antenna, get reflected and remix in the mixer.
Some architecture is talking about low IF to resolve this DC issue.
If it is a true direct conversion receiver, then there is no image problem! If you remember, image is always located at 2*IF away from the desired signal. If IF=0Hz, where is the image?
You usually don't convert your wanted signal to 0Hz.
You can make a direct conversion AM or DSBSC receiver where the LO is tuned to the carrier frequency but AM is not used much these days.
Most people listen to broadcast radio on 88-108MHz FM instead of 550-1330KHz AM.
Analog mobile phones are FM and have been entirely replaced by digital in most countrys.
Nobody ever uses DSBSC.
Converting an FM signal down to 0Hz dosn't work.
For SSB the LO is offset from the wanted signal so that the output is not 0Hz but 300-300Hz voice frequencys. There is an image problem.
If it is a true direct conversion receiver, then there is no image problem! If you remember, image is always located at 2*IF away from the desired signal. If IF=0Hz, where is the image?
But we do have DC offset problem to solve, for example, LO leaking to the antenna, get reflected and remix in the mixer.
Some architecture is talking about low IF to resolve this DC issue.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
