Automatic variable potentiometer circuit needed

[neazoi]

Hello,
I need a bit odd circuit for a construction of a soft CW keyer.

One way I think of is to automatically vary the power gain of an RF amplifier when pressing a switch. I think this can be done by varying its RF input using a single potentiometer.
Do you know any other way for a soft CW keyer?

I give you the scenario in my case:

1. close the switch

2. the potentiometer value changes progressively and automatically from 0K to 100K say. (the amount of change I need it to be variable)

3. open the switch

4. the potentiometer value changes progressively and automatically from 100K to 0K. (the amount of change I need it to be variable)


I have seen these digital potentiometers but they require that you hold the switch on until they go for full resistance and then you have to keep it hold until they go for low resistance. I just need this to be automatically done with a single switch press/release.

I coulkd do that with a photo or a voltage variable resistor I guess but then I have to perform the same press/release switch operation to vary the voltage. Have you got of any ideas or what to look for?

 

[IanP]

You need a small microcontroller and a digitally controlled potentiometer ..
As microcontroller I suggest the PICAXE-08M ..
As digital potentiometer: DS1803-100 ..
**broken link removed**

And here are hints on how to connect both together via the I2C bus ..
http://www.rev-ed.co.uk/docs/axe110_i2c.pdf

Rgds,
IanP

 

[neazoi]

The microcontroller/digital potentiometer was what I first thought.

The problem is that these digital potentiometers are good for audio but not for RF. Since RF will be passed through the potentiometer (Shortwave frequencies) I was looking for a more "classic" way like for example light variable resistors. But then there must be a way of increasing/decreasing a voltage to control the brightness of the lamp. I bet there must be an easier microcontoller-less way to do this, but I do not know where to search.
Any ideas?

 

[IanP]

How about using a FET as volttage controller resistor?
See fig. 5.7:
**broken link removed**

:?:

 

[neazoi]

Do you think the FET will not degrade the RF signal quality in terms of phase noise and intermodulation distortion?

hm... I have found what I was looking if photo-resistors are to be used:
http://www.analog-synth.de/synths/mod2/trigdiv/attrel.gif
A simple attack/release envelope generator contoled by a single switch, yet simple enough and variable.

So the thing still remains, the decision of the variable resistor element and how it behaves in Radio frequencies. Any clues?

Added after 21 minutes:

This seems interesting. using discrete components...

http://media.photobucket.com/image/...nsistors schematic/Robertsunrise/pict2105.jpg

 

[FvM]

What makes you think that LDR are better suited for RF applications than FET variable resistors? You most likely didn't consult
the respective data sheets. The LDR circuits shown in the links are dedicated to AF apllications.

Actually FET variable resistor and LDR both have limited bandwidth, they can be used up to several 10 MHz, but then the resistance
dynamic isn't very high. I assume that FETs can perform slightly better, but I'm not sure about. For high dynamic RF variable
attenuators, variable gain amplifiers are the better alternative.

 

[neazoi]

Yes that is what I try to do in my application, a variable gain amplifier having the gain controlled by an attack-release like circuit.
At the beginning of the post I thought I could have a potentiometer at the RF input, but you gave me an idea with the variable gain amplifier.

The class-A broadband amplifier that I am using is attached. The gain ,I think, Can be adjusted by the 3K3 anf the 750 ohm resistors.

An LDR could be added there so that the gain can be adjusted using an attack-release circuit. So the potentiometer will not be at the RF path and also the input/output impedance remains the same.

how do you think of that?

I have to say, I am anaware if by changing the value of the bias resistor the circuit falls into class-AB...