Very simple Audio compressor for FM transmitter

[Willen]

...falling in more confusion by these answers hehehe

If I cannot use KSK596 or BS170 (even with modify) then I am going to make a garden for gardening, I think it is VERY useful to me. Because I never find each n every components which I need! Electro-Sahara Desert

 

[MrAl]

Hello again,

Another idea for the CdS cell is to limit the dynamic range by design. For example, if we find a cell that has a 5ms time constant then that is going to be over a wide dynamic range. Limiting the dynamic range to 1/2 of that by design brings us down to 2.5ms, and that's not asking too much really.

But apparently the OP wants to use specific parts so that really means a redesign of the entire limiter section.

 

[Willen]

But apparently the OP wants to use specific parts so that really means a redesign of the entire limiter section.

Yeh, I wanted to get alternative option for 2N5459 for #2 post's limiter circuit. But thinking to leave the project. But will make if I find

 

[MrAl]

Hi,

What would it cost to mail one from USA to you?

 

[ronsimpson]

Willen, where are you. What country? Is there any one close?

MrAl,

Another idea for the CdS cell is to limit the dynamic range by design. For example, if we find a cell that has a 5ms time constant then that is going to be over a wide dynamic range. Limiting the dynamic range to 1/2 of that by design brings us down to 2.5ms, and that's not asking too much really.

Tried that. Won't work. A cad cell has a delay much like a RC. Cutting the drive voltage in half does not change the delay it is still one RC. I did a frequency sweep and found the cell had a frequency response of about 100hz. So in the drive amplifier for the LED I added a speed up capacitor that caused that amplifier to have a gain of one below 100hz and have a gain of 10 at 1khz. This really helped. That gave a nearly flat response to 1khz. The response was some what dependent on drive level.

What I did in the end: I used a MOSFET to do a small amount of limiting very fast. AND a CAD-CELL to do a large amount of limiting very slowly.

I no longer use MOSFETs or JFETs or CAD Cells but I still use multiple feed back loops. A fast response to handle that one too loud ward and a slow response to handle the over all trend.

 

[MrAl]

Hello again,

So you are saying that it should be modeled more like a light controlled current source charging a capacitor then? That would make some sense then.

But on the other hand these are used in compressors. So we'd have to find out what they are doing. Maybe they dont need them to be that fast.

What makes even more sense is to use a gain controlled amplifier. There was a part number we used to use a long time ago, but i would imagine these things are even better today. If i remember right it was all about controlling the bias on the differential amp stage, but it's been a really long time now.

 

[Willen]

Hi,

What would it cost to mail one from USA to you?

Hi,
Once a designer and admin from www.techlib.com/electronics Charlese Wenzel sent me all needed parts (mosfets, transistors, ICs, diodes, inductors) from his home Austin, Texas. He was saying that Even now in the world It cost only $4 to send mail to each part of the world.

He included 12 ICs, 10 big inductors, 20 diodes, A big special capacitor, 30 TO-92 components to make his all types of Lightning detectors. Then I made spherics detector whis is more sensitive. You see can the device made by me at the bottom of his page, he included mine too- http://www.techlib.com/electronics/spherics.htm (my formal name is- Rajkumar) was so happy!

It means even with little heavy weight with more components, mailing rate is not more high according to him.

Ronsimpson> Willen, where are you. What country?
Is there any one close?

I am from Nepal (between China and India). 'is there any one close' mean?

 

[MrAl]

Hi again,

Well how bad do you want to build this thing? Maybe one of us here could get the FET's and mail a few to you. Would that be enough to get the thing built?

 

[Willen]

Hi again,

Well how bad do you want to build this thing? Maybe one of us here could get the FET's and mail a few to you. Would that be enough to get the thing built?

Wow............ It would be extremly happy moment if I got these few!!!

Almost a year ago I made this (attached) basic FM Tx designed by audioguru. I made few other Tx too as an experiment but this one has VERY crystal clear sound like a professional radio. But sound was little small than professional radio. I didn't know about such limiter/compressor so I experimented lot more with audio amplifier stage of the Tx, simulated with different bias, applied differents voltages, used differents transistors, even I used a 1000 hFE transistor MPSA18; But got nothing more than distortion. Later audioguru said about compressing audio to trasnmit loud audio. Very later I found few circuits on internet, made me excited to do! Excited to see its comprecessed and bypassed audio spectrum on computer screen by making it myself, wow! It would be cool thing to me. In past I asked 100ths of questions to audioguru and learnt lot more, sometime he was irritated too by my childish questions hehehe.

I was much more excited to make "stereo encoder" but after searching 3 years I made my hands up (same components problem).

Anyway, In real- I didn't find the jFET anywhere in past days too. But when I sent an email to a local distributor of Kathmandu (capital city), they said "we have only TL072 ICs, not others", then I stopped once in past.

Kathmandu is far and no home delivery service is stablished here in such rural places of mine. There is a friend who can buy few TL072 but he will not return soon, only after 7/8 months later he can buy them (he is out of Kathmandu now so). Currently, I have two pieces of LM358 opamp but most of experts said that it is not good opamp for audio purpose which can create hiss.

So in conclusion- to find jFET (2N5459 or any) is 'impossible' and to find TL072 is 'hard'. If it is not bad to request few things then along with jFETs I would request these things too-

- few pieces of TL072
- few pieces of high value resistors like: 1M, 2M2, 3M9, 4M7... I need 1M and 3M9 resistors here on this Limiter too. I didn't find such Meg resistors in past so searched hardly on old TVs or radios' circuit.

But, if it is hard to manage and send these additional things, I will manage myself by taking even long time as I said before- more than 8 months, maximum a year. Almost I have an addiction of searching components, so searching unavailable components from 4 years. That is why each semiconductor has very high important on my life. I have bought more than 100 pieces of general purposes BJTs which I can find here. Sorry for long text.

Thank you lot!!!

 

[Willen]

Hi MrAl,
Are there difficulties? You just stopped...

 

[MrAl]

Hi again,

Oh ok well i'll see if i can find some FET's then and try to find out what it would cost to ship to your area. Where are you located (maybe you should PM me with info and i can call the post office on Monday).
I might be able to get a few op amps too but not sure yet about the other parts.
Are you sure you will build this though if i go through the trouble of gathering up all the parts? I would hope that you do <smile>

P.S.
I am getting all "obsolete" messages from my distributors, where the new FET part would be a surface mount part. This would be harder to deal with.
I'll have to look around some more i guess.

P.S. P.S.
This FET is looking like it is hard to find. The replacement part is not the same. We might have to turn to a total new design. That will take a little experimentation on your part.

 

[Willen]

OK, as you said I am turning the process into PM...

 

[audioguru]

Instead of trying to find an old Jfet that is now obsolete, why not use a modern VCA (voltage controlled amplifier) IC instead?
But the modern ICs I have found are only made in a tiny surface-mount package.

 

[ronsimpson]

Instead of trying to find an old Jfet that is now obsolete, why not use a modern VCA (voltage controlled amplifier) IC instead?
But the modern ICs I have found are only made in a tiny surface-mount package.

The 4301 comes in a DIP. But if you can't find simple parts then that another problem. THAT corp has many good audio processors. Only one comes in DIP I think. Read their application notes. Good reading. If there are questions on these parts I can help.

 

[MrAl]

Hello again,

Well i suggested a voltage controlled amplifier a long time ago, but i also realized that he will have a hard time getting one because he has a hard time obtaining parts in his bad location for obtaining parts.

So with that in mind, did anyone here ever try using a low power MOSFET for the limiting device rather than a JFET? Of course the biasing would be different but that is to be expected. The main thing is if you did then did you find any major pitfalls before i begin working on this? He has a couple already so if we can do a limiter with a MOSFET he wont need any parts right away.

 

[ronsimpson]

MOSFETs work well.

I think I brought up "US 5808516 A" back many posts.

 

[ronsimpson]

 

[MrAl]

MOSFETs work well.

I think I brought up "US 5808516 A" back many posts.

Hi,

Oh ok great. Then do you have a reference circuit we can take a look at?

 

[ronsimpson]

I think there are examples in old national notes, and there are several in old HAM books. This one is not the best on distortion. The MOSFET need two resistors on the gate to make it more linear. I think I used 2meg resistors.

 

[audioguru]

When the FET has the two resistors on the gate with a series capacitor then the distortion is reduced a lot but the attack time is very slow. The circuit from Rod Elliot that I posted (post #2) reduces distortion and is fast.