I am working on radio project for school. We we given a whole circuit to test. My question is that I want to change the detector circuit (AM envelope detector) with a better one. Any suggestions will be much appreciated.

 

Post the circuit. Better in which sense?

 

Unfortunately your circuit is not yet visible. This may be due to a delay for moderation as you are a new member, but there may be some other issue. You might try to upload again: .png files are preferred for pictures.

 

An AM radio has very poor audio. it does not produce high audio frequencies and it picks up all kinds of interference. How can you make it better?

 

Can't do anything about interference, but the high end audio is a function of the limits imposed by the FCC. Both CB and commercial frequencies are limited to 10Khz, but this is an artificial limit designed to allow more channels on the band.

We really do need to see a schematic by the OP. You could use a balanced mixer with a phase locked carrier frequency, but I have a feeling you would have no idea as to what I just said.

One really simple suggestion is to replace the diode with a Shottky diode, which drops less voltage and has a higher frequency response.

 

An AM radio cuts audio frequencies above about 2000Hz so that the 10kHz beat frequency between adjacent stations is not heard as a squeal.

An FM radio cuts audio frequencies above 15kHz so that the 19kHz stereo pilot tone is not heard as a whistle.

Here is what the Pros say about a lousy AM radio:

 

Again, that is by design, not function. Telephones roll of at 3.4Khz analog, this used to be SSB on a carrier wave. TV's (old style) used AM for their video, it rolled off at 3.5Mhz. You are talking specifications and treating them like they are performance. As far as sound goes it is dictated by the FCC (and whatever the Canadian equivalent is).

But I will slap myself on the back of the head, because this is off topic.

For the OPs question the Schottky diode is one answer. A Schottky diode is very close to the old razor blade principle of AM crystal radios. A Schottky diode drops between 0.2 and 0.3VDC, so that helps eliminate on source of loss.

Nowdays, with modern frequency synthesis they use the balanced mixer principle I mentioned in AM radios. There is no loss with this method, but the synthesized frequency must be rock solid.

You can also use a germanium transistor BE junction to provide built in gain, but I'm not too fond of that idea.

 

here is the link to my circuit:
http://dl.dropbox.com/u/40549321/doc1.PNG

 

The Schottky diode is still valid.

What is your objections to this circuit. BTW, it could have been larger, I can't make out a lot of the text, such as the part numbers. You could have attached this file to your post, or you can eventually (after 10 posts) use the local hosting service AAC provides.

 

AM radio stations boost 2kHz to 5kHz at 18db/octave to help make speech more intelligible on an ordinary AM radio.

I worked with tele-conference systems for banks. They complained that there were no high audio frequencies.
I measured -12dB at 3kHz round trip between two city telephone lines. Bell said it was fine because their limit is -7.5dB one-way. So I made and sold an equalizer that boosted 3kHz by 12dB (then a sharp drop above 3.3kHz) and it made speech sound crisp and clear.

 

as to the objections, the picked signal is too weak. for that reason I need to work on the detector and hopefully I can to solve that issue.

 

Come AG, I know you know radio theory and AM at least as well as I do, and no doubt better! Any drop off is purely an artifact of circuit design, not theory. Side bands next to the carrier should be exactly the same amplitudes as their audio counter parts. To hear you talk NTSC TV could have never worked, since it goes to 3.5Mhz and stays flat (more or less) over the range.

Fact is, the 10Khz range was mandated (gee, I think I said this before) and the audio circuitry has to roll off around there to prevent cross talk, which in this case sounds a lot like whistling. Ditto for the 3.4Khz roll-off on telephones, they picked 3.4Khz because the guys at the early Bell Labs determined you could still understand human speech at that range. So they sliced up a carrier using USB (upper side band, a form of AM) into 3.4Khz segments, each its own voice channel.

The really interesting part is a lot of this was developed during WWI, as laying a pair of twisted wires on the battle front was costing lives, and they really wanted to use that pair of wires for more than one voice channel. War tends to advance technology, this is a prime example.

The first 20 years of my professional life was Collins Radio, it was also my core subject in college (Communications). If you want to discuss this on another thread I can arrange it, but I'd rather avoid confusing the OP, who is new to the subject, at this time.

To the OP, I will tell you a lot of what you are learning is oversimplified. It is OK, all of us went down the same path. It is easier to understand the circuit the way your being taught, and then later they will go deeper into theory and correct the holes. Keep working on what you are doing for now, it applies in a lot of other things.

Note the small capacitor to ground right after the diode. It shorts the high frequency carrier components to ground and (hopefully) leaves the audio. Unfortunately it also distorts the audio frequency response. It is why the base is stronger than the highs.

The Schottky diode will absorb less of the signal than a conventional diode. A conventional diode absorbs 0.6V to 0.7V just turning on, compared to the Schottky's 0.2V to 0.3V. Schottky's typically have a much higher frequency response, a plus. Many times a germanium diode is used instead for the same reason, lower dropping voltages.

Your circuit attempts to compensate for the diode by adding gain in front. This works well. So it goes back to what your complaint is, how much signal are you loosing here? Or is it a frequency response issue?

There are circuits that are much better, but they are also much more complex.

 

as to the objections, the picked signal is too weak. for that reason I need to work on the detector and hopefully I can to solve that issue.

Today you added that you have a mixer and oscillator, so maybe the circuit you posted is the IF amplifier and detector?
You said today that you are using a "square Loop" antenna instead of a ferrite bar antenna wound with Litz wire like most AM radios.

A real AM radio sometimes has an RF amplifier transistor feeding the mixer and has a few high gain tuned IF stages. Your IF amplifier is not tuned.
A real AM radio usually has so much gain that it uses an automatic gain control circuit to reduce and adjust the gain so that all stations (local and distant) have the same volume.

 

Last I heard a crystal radio was also a real AM radio. This circuit is somewhere between.

 

thank you all for the quick response. how would I tune the IF amplifier?

 

AM radio has a lot of interference from power lines, as well as noise caused by signal decay. You can try to add in better high and low pass filters to better filter out some of this "noise", but i can't guarantee that this will do much. i would advise using a good, long antenna, keep it away from any power lines or appliances if at all possible, as this will cause interference.

 

thank you all for the quick response. how would I tune the IF amplifier?

Disable the automatic-gain-control, tune in a weak distant station and peak the IF transformers.

Or you can feed a weak signal generator signal to the antenna so that the AGC is at max gain then peak the IF transformers.

Peak means max output level.

 

I'm looking for IF transformers. Any suggestions on what model I need to buy and where I can get them from?

 

Modern radios use crystal filters instead of IF transformers.
Digikey has some 10.7MHz crystal filters for the IF amplifier in FM radios but nothing for AM radios.

 

Try Mouser. This page is for customers in my country, you would need to look for your own.
http://uk.mouser.com/Passive-Compon...5gbg?P=1z0wmc5&Keyword=IF+transformer&FS=True