Note that a FET is not the only solution for your task. You also can use an OTA (integrated circuit) or a light-dependent resistor (LDR) instead. It is always good to know the alternatives...

I would like to use the FET and I tried to use a resistor but how do I need to calculate the resistance, like for a Voltage divider?

 

I would like to use the FET and I tried to use a resistor but how do I need to calculate the resistance, like for a Voltage divider?

I would suggest that you build a jig for a JFET with a variable gate voltage and a way to measure Vds and Id. You can easily do this in a simulator like LTSpice. What this will allow you to do is characterize a given collection of JFETS so you know what their non-linear resistance characteristic looks like.

 

The question how to use a FET as a resistor is also treated here:
https://electronics.stackexchange.com/questions/471498/using-an-n-channel-jfet-as-a-vcr

 

The question how to use a FET as a resistor is also treated here:
https://electronics.stackexchange.com/questions/471498/using-an-n-channel-jfet-as-a-vcr

I managed to understand(kind of) how the JFET works as a resistor and I saw that it should have a minimum resistance at Vgs = 0, but in my simulation, it goes to Mohms. I have no idea why the voltage looks like a AC not DC to the voltmeter but at the DMM it shows 1,6V.

 

You really should try to start systematically. At first, investigate the FET alone and become familiar with the limits of resistor application. Select a suitable value (and the corresponding control voltage) .....and - as a second step - start to determine the parts values for the GIC resonator, based on the selected FET resistance. And remeber to use the FET in series with another suitable resistor because of the amplitude limitations.

 

You really should try to start systematically. At first, investigate the FET alone and become familiar with the limits of resistor application. Select a suitable value (and the corresponding control voltage) .....and - as a second step - start to determine the parts values for the GIC resonator, based on the selected FET resistance. And remeber to use the FET in series with another suitable resistor because of the amplitude limitations.

Okay, but how do I determine this resistor value? And how can I find what the amplitude limitations are?

 

Stop worrying← , do something . . . anything - then ask questions about (even if it's totally off the track←)
. . .
- as infact the more you learn the more you become into realization of how little and non-precisely you actually know about the field you study ◄ this is what normally occurs - so you can't worry in advance of not being multiple doctor on that subject

 

Okay, but how do I determine this resistor value? And how can I find what the amplitude limitations are?

When you under which conditions the FET can be used as a resitor (part of the output chracteristic set of curves) it should not be a problem to measure or to simulate such a resistor. Look at the ID=f(VGS) characetristic curves and verify the corresponding resistances VDS/ID.

 

You can use MOSFETs too, and since you are working at such a low frequency you have a wide range from which to select.

Just a thought: Are you sure you are connecting your JFET correctly -in other words are you sure of which lead is which?

 

I managed to make a good simulation of the circuit but I changed the resistor from JFET to a photoresistor with a led. The whole system is working perfectly in the simulation. But when I made it on a breadboard I have a no changing frequency of 1.2 MHz and I want 10 - 15 kHz.

 

you can consider to use ICL8038 or equivient circuit. you can check the data sheet. If you want to buy PCBA, pls take the below parts as reference.
The frequency (or repetition rate) can be selected externally from 0.001Hz to more than 300kHz using either resistors or capacitors,

 

The first things I would examine are the differences between the simulation and the physical circuit.

 

Ahh, another newbie necro post.

 

There used to be an old app note from RCA semiconductor, which showed how to make a voltage tuneable sinewave oscillator with a transconductance opamp. The CA3080.

Unfortunately RCA semi has been kaput for decades, and the different companies which have successively acquired the assets, haven't kept the old app notes online.
Additionally, the CA3080 has been obsolete for at least 15 years, although you still find some on e-bay.

You can find two CA3080Es inside an LM13700.

 

Solid State Micro Technology (SSM), now a part of Analog Devices, Inc. (ADI) also has some.

SSM2010
SSM2013
SSM2018
SSM2120
SSM2122

ak

 

Hi,

Well this thread was started over 3 years ago so i doubt the original poster is that interested anymore, but hey there could be others that read this thread and like it enough to try building one themselves. That's what hobby electronics is all about.

One of the posts I ran into in this thread takes me way back, much further back then this thread when I used to play in chess tournaments. The post mentioned RCA, which is no longer with us. The reminder was a chess match i played in the old RCA building in New Jersey. Not sure if it is there anymore and even back then it was vacant. Can't remember what town it was in either.

My main reason for the reply though is that for the original problem statement, there are a number of ways to approach the problem using 'regular' parts like resistors, caps, inductors, and op amps. There is also a unique solution using either diodes or bipolar transistors where either one is used to shape a triangle wave. The triangle wave is shaped by having multiple diodes or transistors cut part of the triangle over the half cycle and that turns it into a pseudo sine wave.
This is actually a pretty interesting solution I was into a lot back in the 1980's and a little after that too. That is also how the aforementioned 8038 and similar chips work although they have all the bipolars inside the chip. It's not a perfect output but it's not too bad for some things. The only caveat is that when sine waves are really needed they usually need to be ultra pure. For example for testing audio equipment for distortion. That may be why these diode/transistor chips were mostly discontinued. They made a 20MHz version too, but that was also discontinued. That was more interesting and useful than the 1MHz garden variety.
I guess there were benefits though, such as simple control of frequency and amplitude.

If the goal is only one frequency output at a set output amplitude you can also use a square wave and decent bandpass filter. The bandpass filter takes out all the odd harmonics that appear in the square wave leaving you with just a sine wave at the fundamental frequency.
If you only need four such frequencies and amplitudes, you could just build four of those circuits and you'd get a pretty darn clean output sine wave.

Yes, the most modern method now is to use either a DAC or PWM and some filtering. With PWM and filtering you can get up in frequency pretty high, into the 10's of megahertz maybe higher. There are chips and even little PC boards with those chips on them and not too expensive, and they go up to around 40MHz.

 

Then there Is there is Direct Digital Synthesis (DDS) which uses a DAC to get anything more than pulses, an give high resolution but has inherent jitter which might be too high for some applications (and varies according to DDS design).

https://www.allaboutcircuits.com/te...thesis is used,and to enable frequency sweeps.