Discussion in 'Electronic Basics' started by Ruben Undheim, Jun 6, 2004.

1. ### Ruben UndheimGuest

I am trying to understand the principles of the superheterodyne receiver.
But I've came up with one question.

If I will receive wide-band FM radio with 200KHz bandwidth, and use two IF's
(10,7MHz and 455KHz), is it then possible to use a 455KHz ceramic filter
with a bandwidth of only 10+- KHz? Won't that give a bandwidth of only 20KHz
with much of the signal lost, or does it work? Can anyone explain this for
me?

Thank you! ;-)

Ruben

Ruben wrote:

3. ### DboweyGuest

ruben really did write:

<< I am trying to understand the principles of the superheterodyne receiver.
But I've came up with one question.

If I will receive wide-band FM radio with 200KHz bandwidth, and use two
IF's(10,7MHz and 455KHz), is it then possible to use a 455KHz ceramic filter
with a bandwidth of only 10+- KHz? Won't that give a bandwidth of only 20KHz
with much of the signal lost, or does it work? Can anyone explain this for me?
--

If you built the FM receiver as a dual conversion receiver with IFs of 10.7 MHz
and 455 kHz as described, the 455 kHz ceramic filter would destroy the signal
content. Beyond that, however, such a design would would fail for lack of a
wide-band discriminator at 455 kHz.

"FM band" receivers do not usually use dual conversion. Their design maintains
the wide-band response through to the wide-band discriminator.

Don

4. ### John LarkinGuest

Nope. It will trash the FM sidebands and create huge distortion.

Double conversion isn't usually necessary for broadcast-type FM
receivers; they usually have a single 10.7 MHz IF. Double conversion
is used to increase image rejection, not a big problem here, and that
would only work if the first IF frequency is increased beyond 10.7.

John

Worse, it will create horrendous distortion.

John

5. ### Ruben UndheimGuest

Thank you!

That's exactly what I had started to suspect during this hour after posting
my question. It's nice to have it cleared out in my head!!
Thank you again.

6. ### Ruben UndheimGuest

Thanks to you too!
;-)

7. ### John FieldsGuest

---
It doesn't work.

The reason is that even though heterodyning changes the location of
the carrier, the deviation of the sidebands will stay the same.

For example, assume you have an incoming carrier with a center
frequency of 100MHz modulated +/- 100kHz and you heterodyne that down
to 10.7 MHz with an 89.3MHz local oscillator.

Since you'll be getting the carrier,(F1) the LO,(F2) the upper
sideband (F1+f2) and the lower sideband (F1-F2) out of the mixer and
what you want is only the 10.7MHz, what you'll do will be to filter
out everything else in order to let only the 10.7MHz come through.

Now, since you've got an 89.3MHz LO, when your carrier is at exactly
100MHz you'll get 10.7MHz out of the filter. Then, when your carrier
goes to 100.1MHz, F1-F2 will be 100.1 - 89.3 = 10.8MHz and when the
carrier goes to 99.9MHz the output of the filter will be
99.9 -89.3 = 10.6MHz, so you can see that even though the center
frequency of the carrier was moved from 100MHz to 10.7MHz, the
sidebands remained displaced from the center frequency by +/- 100kHz.

The same thing will happen when the 10.7MHz center frequency of the
first IF is beat down to 455kHz with the second oscillator, and if the
2nd IF stage only has a bandwidth of +/- 10kHz, then anything put into
it by the second mixer below 445kHz or above 465kHx will be lost.

8. ### Michael BlackGuest

But what's the question?

You start off by saying you are trying to understand superheterodyne
It seems like you are really asking something else, but it's not been phrased.

It almost seems like you are trying to figure out the purpose of double
conversion, and before you get to the explanation you are puzzled over
why it's needed, at least in the example you use. If it is the issue
of double conversion, you need to back up and understand it (or ask
about it) before going into specific applications.

Ther are good reasons for double conversion.

Michael

9. ### Ruben UndheimGuest

Hello!
The reason for that I asked was that I have studied some schematics of the
superheterodyne receiver, and they have all been with either two IF's (10,7
and 455), or only one 455KHz IF. Therefore I couldn't understand how it was
possible to receive 200KHz bandwidth radio, when I couldn't find a 455KHz
filter with more than 35KHz bandwidth. My problem was that I thought that it
was not possible to get sound out of a 10,7MHz IF without the 455KHz stage,
since I had not seen any examples for it. But now I have got the answer I
searched for!
I hope you understand!! Curious problems appear when I am studying on my
own. ;-)

Thank you!

Ruben

10. ### Rich GriseGuest

It sounds to me like he doesn't realize that the 455KHz IF strip is
for the "AM" mode.

Cheers!
Rich

11. ### L. FiarGuest

Hi.

There is no problem with retrieving the sound from an FM signal with such a
design, dual conversion is used for other reasons.

A low IF is good for selectivity and demodulation reasons, but it leaves the
image frequency closer to the wanted one. For example, with a 455KHz IF, the
receiver will also pick up a signal 910KHz (twice the IF) away from the
wanted signal - just as if it were on the same frequency.
A dual conversion uses a high first IF, allowing the front end circuit to
attenuate the image at 21.4MHz away (for a 10.7MHz IF), and a low second IF
allowing selectivity. The first IF filter will have a bandwidth far below
910KHz - attenuating the 2nd IF image frequency before it gets to that
stage.

For broadcast stations, using wide band FM, an IF of 10.7MHz is reasonable.
A lower IF may not allow for the 200KHz bandwidth. As it already places the
image 21.4MHz away, another conversion stage before it is not required.
For communications radios, using speech, narrow band FM is used. It is very
possible to get such signals within a 10KHz channel spacing, with receivers
having 60dB or greater attenuation on that adjacent channel. This requires a
low IF systems.

I know I've said it before, but this is a great book for radio topics: