Image signals.

[lord loh.]

I would like to know what exactly are image signals (in the context of superhetrodyne receivers) and what is their origin?

My textbook says that it is to be eliminated in the RF amplifier because if it reaches the mixer, it can never be eliminated.

The same textbook says that, the tuned stages of the IF amplifier eliminates the image signals as the lower local oscillator makes the image frequency to deviate further more from the desired IF.

These somewhat contradictory explanations have left me confused...Please help...

I would appreciate any explanations, etext, weblinks or anything that helps me clerify this idea.

Thank you.

 

[Nigel Goodwin]

I would like to know what exactly are image signals (in the context of superhetrodyne receivers) and what is their origin?

My textbook says that it is to be eliminated in the RF amplifier because if it reaches the mixer, it can never be eliminated.

The same textbook says that, the tuned stages of the IF amplifier eliminates the image signals as the lower local oscillator makes the image frequency to deviate further more from the desired IF.

These somewhat contradictory explanations have left me confused...Please help...

I would appreciate any explanations, etext, weblinks or anything that helps me clerify this idea.

Image frequency is very easy to understand.

A double-balanced mixer gives two outputs, the sum and the difference, so for a specific local oscillator frequency the input to the mixer can be either higher than the local oscillator, or lower than it, by an amount equal to the IF frequency.

For an example:

The local oscillator is running at 100MHz, and the IF is 10.7MHz. This gives reception of two frequencies 100+10.7=110.7MHz (sum) and 100-10.7=89.3MHz (difference). In this case only 89.3MHz is in the FM broadcast band, so the front end tuning should be tuned to 89.3MHz as well - attenuating the unwanted 110.7MHz signal (if there was one).

You will notice that the wanted signal and image frequency are 21.4MHz apart, this is double the IF frequency - for the image rejection to be effective, the two frequencies need to be as far apart as possible, which is why 10.7MHz is used for VHF, and 455KHz for much lower frequencies.

Very high end receivers often use much higher IF's, often higher than the tuning range of the receiver! - in this way the image is shifted too far away to give any problems, and a simple low-pass filter is all that's required on the input.

And YES, you MUST get rid of the image before the mixer, it's the front end tuning which does this!. The IF frequency simply sets where the image is, at twice the IF frequency away from the desired frequency.

 

[lord loh.]

Okay, I think I am begining to understand...but some bits and pieces are still remaining...

I am thoroughly embarrassed asking a question which might be obvious...

...A double-balanced mixer gives two outputs, the sum and the difference, so for a specific local oscillator frequency the input to the mixer can be either higher than the local oscillator, or lower than it, by an amount equal to the IF frequency...

This sounds as if there are two mixers...A double-balanced mixer which feeds the sum and difference to another mixer which may have an input above or below the local oscillator frequency.(My textbook had two mixers only in double superhetrodyne.)

If the mixer gives two outputs, then does it not mean that the image frequency is being generated by the first mixer ?

...For an example:

The local oscillator is running at 100MHz, and the IF is 10.7MHz. This gives reception of two frequencies 100+10.7=110.7MHz (sum) and 100-10.7=89.3MHz (difference). In this case only 89.3MHz is in the FM broadcast band, so the front end tuning should be tuned to 89.3MHz as well - attenuating the unwanted 110.7MHz signal (if there was one)...

How is the IF set to 10.7 MHz?

 

[Nigel Goodwin]

Okay, I think I am begining to understand...but some bits and pieces are still remaining...

I am thoroughly embarrassed asking a question which might be obvious...

...A double-balanced mixer gives two outputs, the sum and the difference, so for a specific local oscillator frequency the input to the mixer can be either higher than the local oscillator, or lower than it, by an amount equal to the IF frequency...

This sounds as if there are two mixers...A double-balanced mixer which feeds the sum and difference to another mixer which may have an input above or below the local oscillator frequency.(My textbook had two mixers only in double superhetrodyne.)

Yes, you only have two mixers in a double-conversion receiver. This is commonly done in VHF communications receivers, a first IF of 10.7MHz (to give good image rejection) and a second IF of 455KHz to give a good percentage of deviation.

The only reason I mentioned 'double balanced' was to make it simpler

A double-balanced mixer will only output sum and difference, a 'normal' (non-balanced) mixer will output the oscillator and input frequencies as well. I thought it was best to stick to just two outputs?.

If the mixer gives two outputs, then does it not mean that the image frequency is being generated by the first mixer ?

No, the image is an RF signal entering via the aerial - if there is no signal on that frequency then there will be no image! (and no image problem).

...For an example:

The local oscillator is running at 100MHz, and the IF is 10.7MHz. This gives reception of two frequencies 100+10.7=110.7MHz (sum) and 100-10.7=89.3MHz (difference). In this case only 89.3MHz is in the FM broadcast band, so the front end tuning should be tuned to 89.3MHz as well - attenuating the unwanted 110.7MHz signal (if there was one)...

How is the IF set to 10.7 MHz?

By it's tuning, originally with LC tuned circuits (IF transformers) but these days usually with ceramic filters - which give a better performance, and require no alignment.

 

[lord loh.]

A few more things that may help me understand....

Which of these quantities are fixed and which are varied ?

1. The RF is varied...as we tune into different ststions.
2. Is the local oscillator frequency varied ? No. I think.
3. The Intermediate frequency is the sum and differnece terms of the RF and LO so if RF varies, IF shall vary.

4. Should IF vary ? If it should not, then I am wrong about point 2. The local oscillator frequency should vary...

Please tell me if my assumtions are right...I seem to need a lot of thinking and reading befor I get an insight..

Thanks for all the help so far...

 

[Nigel Goodwin]

A few more things that may help me understand....

Which of these quantities are fixed and which are varied ?

1. The RF is varied...as we tune into different ststions.
2. Is the local oscillator frequency varied ? No. I think.

Yes it varies.

3. The Intermediate frequency is the sum and differnece terms of the RF and LO so if RF varies, IF shall vary.

No, the IF is fixed - notice I mentioned it often uses ceramic filters!.

4. Should IF vary ? If it should not, then I am wrong about point 2. The local oscillator frequency should vary...

Please tell me if my assumtions are right...I seem to need a lot of thinking and reading befor I get an insight..

The whole point of a superhet receiver is to use a fixed IF, it's basically a TRF receiver at the IF frequency. But a TRF receiver is difficult to tune, because you have so many tuned circuits which require simultaneous adjustment, with accurate tracking.

So a superhet operates at a single fixed frequency (the IF), which can easily be aligned and doesn't need tracking - it then adds a converter on the front, the mixer, oscillator, and front end. By adjusting the local oscillator, and front end tuning, you can alter the input frequency of the converter, while maintaining it's output at the IF frequency. The tracking requirement is simplified to the local oscillator, and either one or two front end tuned circuits.

 

[lord loh.]

Thanks Nigel Goodwin,

I read my text book all over again with the information that I got from you in my mind and seemed to understand it a bit further...

My text book had initially explained about a suphetrodyne receiver(1 mixer), then it progressed to a double superhetrodyne receiver(2 mixer).

So, now I understand that the image signal enters the 1st mixer and it's output contains components of RF, RF+LO and RF-LO.
The first IF amplifier tunes into the RF-LO (Desired frequency) rejecting the RF components and RF-LO(image frequency)...

The 2nd mixer recieves only the desired IF and hetrodynes it for better selectivity...

Is there still a flaw in my understanding ?

This means that the single superhetrodyne shall perform misrably as far as image signal rejection is concerned...And it appears that single superhetrodyne was designed for tracking and not image signal rejection...

Am I right here ?

Thanks a lot for all the trouble you have taken to answer...I am really greatful for it...

 

[Nigel Goodwin]

This means that the single superhetrodyne shall perform misrably as far as image signal rejection is concerned...And it appears that single superhetrodyne was designed for tracking and not image signal rejection...

Am I right here ?

No, there are two mutually exclusive requirements:

1) Image rejection - requires a high IF frequency.

2) Selectivity - requires a low IF frequency.

A double conversion receiver satisfies both requirements by having two IF's, the first high, and the second low.

For a single conversion receiver the IF is chosen to be best suitable for the frequency in use, so for low frequencies somewhere around 455KHz is used, which gives good selectivity, and acceptable image rejection due to the fairly low frequencies received.

At higher frequencies (VHF etc.) a 455KHz IF wouldn't provide any image rejection worth mentioning, so a 10.7MHz IF is used - this reduces selectivity - but isn't a problem as wideband FM is used for broadcast radio, which gives it far higher quality.

But for amateur radio, at 145MHz, only narrow band FM is used, so dual-conveersion is commonly used, with a 1st IF of 10.7MHz, and a 2nd of 455KHz.