going crazy with... AGC IF amplifier 120kHz

[caga]

Hi,

I am getting almost crazy with the project I am dealing with days. I need any idea or advice.

Here is the project:

Target: communicate over power cables tx & rx
Concern: price and space is limited, standart components as many as we can use

Problem: When the cable length changes, the resistance, better impedance is changing and the rx should trim itself and normalize the received signal and give it to the uP. The system is using PCM with 120kHz carrier. I am planning to use;

Power Lines--Coupler--High/Band pass filter--Amplifier w AGC--Envelope detector/tone decoder-- uP

I already make some tests being succesfull in specific lengths.

I tried many circuits. The main problem with the amplifier is bandwith. To be standart LM358 seems to be good, with 120kHZ if I am not wrong we can have 5-10 gain. With the two op-amps this should be enough but there is a lot of noise at the output! Classic transistor amplifier design is beyond me!??

The input signal is changing from 100mV-pp to 20-40V-pp. (Mostly on down side, 40v is reached when I used an inductor at the tx side!)

After the amplifier I planned to use a LM567 tone decoder, which select the signals very carefully and outs enveleope. Just a plan, could not amplifie the signal yet.

So if you have some ideas about these please post;

1- AGc amplifier for 120kHZ bursts
2- any standart IC doing one of the jobs above
etc


Many thanks to you all.

 

[Nigel Goodwin]

Please post the circuit you've tried so far, so we can make suggestions.

 

[caga]

IF AGC amp

I tried standart op-amp amplifiers inverting and non-inverting with capacitor coupling, the signal amplified but the noise is too much and the gain is not enough, even with the two stages of lm358.
Actually there is no such one schematic, but let me draw few and post it here.

Before that, can I ask ;

How do you amplifie a 120kHz signal? any IC, op-amp? which one? or a transistor circuit? And as the target is to get TTL, do I have to rectifie it and amplifie? or does anyone know an IC for this?

Thanks,

 

[Optikon]

Re: IF AGC amp

I tried standart op-amp amplifiers inverting and non-inverting with capacitor coupling, the signal amplified but the noise is too much and the gain is not enough, even with the two stages of lm358.
Actually there is no such one schematic, but let me draw few and post it here.

Before that, can I ask ;

How do you amplifie a 120kHz signal? any IC, op-amp? which one? or a transistor circuit? And as the target is to get TTL, do I have to rectifie it and amplifie? or does anyone know an IC for this?

Thanks,

This sounds like the design wasn't properly thought out for transmitting in the intended environment. Please post a schematic so you can get some real help with your troubles. Otherwise, we're going to be stuck answering questions like how do you amplify a 120kHz signal? You do it with an amplifier. Will it work though? show us your schematic.

 

[caga]

schematics

Hope, last posting twice is not my mistake!

Ok, I am trying to reduce the complexity but you are right without seeing it is hard to give an idea 8)

Here is what I tried;

1- first I try this microchip app-note deisgn;

https://www.electro-tech-online.com/custompdfs/2005/09/00236a.pdf

please see pic1

using ordinary NOT gates as an amplifier is a good idea, but the gain is not enough when I try a 100m cable. With little modifications it worked with 1- 30meters, freq is 120kHz.

2- Then I tried standart op-amp amplifier circuits a few just to see and get familiar with the op-amps, and especially concentrated on one circuit, please look at

**broken link removed**

**broken link removed**

I understand this circuit pretty well, the problem is this is designed for audio which is far beyond my PCM freq 120kHz. If this circuit can work for 120kHz I think I will be able to tune its AGC. As far as I can see LM358 is not suitable for such a gain.

So I tried to use the two stages of LM358, make the first stage an amplifier only and feed the second stage with this to amplifie again and use the AGC circuit. But could not sucessfull. The signal is not amplified enough in neither of the circuits. And there is noise at the output, even when I cut its relation with the coupler. I tried to decouple of the power supply, the noise attenuated a little but not dissaperaed.

3- I searched web for amplifier ICs, like MC1350 but its price is rather much for such a circuit, so did not tested yet.

Maybe I am missing some key points to ask, so I ask you "how do you amplifie a 120kHz signal", I try to write some details in my first posting such as voltage levels, but maybe these are not enough also?

Thanks,

 

[Nigel Goodwin]

Re: schematics

Hope, last posting twice is not my mistake!

Yes it was :lol: but I've deleted it for you (Moderator).

Ok, I am trying to reduce the complexity but you are right without seeing it is hard to give an idea 8)

Here is what I tried;

1- first I try this microchip app-note deisgn;

https://www.electro-tech-online.com/custompdfs/2005/09/00236a-1.pdf

please see pic1

using ordinary NOT gates as an amplifier is a good idea, but the gain is not enough when I try a 100m cable. With little modifications it worked with 1- 30meters, freq is 120kHz.

These types of systems aren't intended to work over those sorts of distances - not much use if you turn your light OFF, and the entire streets lights go OFF 8)

2- Then I tried standart op-amp amplifier circuits a few just to see and get familiar with the op-amps, and especially concentrated on one circuit, please look at

**broken link removed**

**broken link removed**

I understand this circuit pretty well, the problem is this is designed for audio which is far beyond my PCM freq 120kHz. If this circuit can work for 120kHz I think I will be able to tune its AGC. As far as I can see LM358 is not suitable for such a gain.

So I tried to use the two stages of LM358, make the first stage an amplifier only and feed the second stage with this to amplifie again and use the AGC circuit. But could not sucessfull. The signal is not amplified enough in neither of the circuits. And there is noise at the output, even when I cut its relation with the coupler. I tried to decouple of the power supply, the noise attenuated a little but not dissaperaed.

An LM358 doesn't look a very good choice for 120KHz?, I would suggest you use a much more modern opamp than that! - you probably couldn't find one much worse!. The TL072 series are usually pretty decent, and cost very little - no doubt Audioguru will dazzle us with specifications?, but it should provide a lot better gain than an LM358.

Maybe I am missing some key points to ask, so I ask you "how do you amplifie a 120kHz signal", I try to write some details in my first posting such as voltage levels, but maybe these are not enough also?

This is all old technology from back in the 1970's, it was never very successful, and X10 seems to be about the only system these days?.

The common way back then was to use a tuned transformer on the input - just like an IF transformer - this accepts the 120KHz (and amplifies it by transformer action), while rejecting the noise. You can then amplify it using decent spec opamps, or even discrete transistor stages. As you're only wanting to amplify a specfic frequency you could have tuned stages, or simply low value coupling capacitors to reduce low frequencies. I don't recall ever seeing one with AGC, it's just not needed, if the signal is too strong it will clip, and that's fine, just as with an FM radio.

 

[caga]

tuned inductor

Thank you Nigel for yr advices.
Actually I did wrote a new protocol for two way communication which have some error correction etc and want it to be used in large gardens, also I will design a trap filter not to let the signals go out to public, so I am trying to make it work in a long distance also.

The AGC idea is great I think, the enveleope detector or the tone decoder will catch the signal even it is clipped, then the only problem is to amplify the signal enough for it to catch. I inspired from power carrier network appliances which uses AGC, but they are of cource more complex designs.

I also tried the transformer version, as a coupler but not a tuned one. As a digital guy it is hard to make the coils. By the way does anybody know a good source for transformers/coils, I means a tutorial to calculate my tuned transformer/coil? I want to try some other examples.

Thanks.

 

[Nigel Goodwin]

Re: tuned inductor

I also tried the transformer version, as a coupler but not a tuned one. As a digital guy it is hard to make the coils. By the way does anybody know a good source for transformers/coils, I means a tutorial to calculate my tuned transformer/coil? I want to try some other examples.

An AM IF transformer is around 470KHz, you may be able to tune it down low enough by adding external capacitance across it?.