Ionizationx: a clean environment is a human right!

Electronics => Best WORKING circuits for WFC => Topic started by: Steve on May 11, 2008, 23:16:59 pm

Title: Frequency generator perfect waves till 1 mhz, 1 till 10 FET channels
Post by: Login to see usernames on May 11, 2008, 23:16:59 pm
Title: Re: Frequency generator perfect waves till 1 mhz, 1 till 10 FET channels
Post by: Login to see usernames on June 17, 2009, 16:56:26 pm

The circuit you posted above  is not properly implementing the VCO section of the 4046.   By definition the oscillator is Voltage Controlled,  yet the control signal is missing and the VCOIN pin 9  is left floating.   Also,  R2 (pin 12)  is not being used.   This means the user could adjust the frequency  down to  0 hz,  causing full current DC across the load.  This will likely smoke a transformer.

When the 4046 VCO is implement properly,  R1 and C1 set the range and center frequency.  R2 sets the offset.    Then a voltage at VCO IN  (Pin 9)  controls the frequency in the ICs linear range.  This voltage will be from about . 5 Vdc  to about 4.2  Vdc assuming a 5 V Vdd supply .  

The output will then be a 50 percent duty cycle square wave from Fmin to Fmax  (as set by C1,  R1,  and R2)
 where ... .5v at VCO IN = Fmin Out  and 4.2V at VCO IN = Fmax Out .     R1 should not be used to control frequency,  but rather to set the center frequency / range.  Again I am assuming a VDD of 5V.

I would attach a schematic of an  actual working  circuit of a properly implemented 4046,  but I cant see how to do that here.   ( Thanks for the tip - file is attached now)

The NTE4046B is an overpriced replacement version of the CD4046.  The 74HC4046 is a somewhat  updated version with a better a PLL section and higher frequency capability.   Both the 74HC4046 and the CD4046 are still available through Mouser for less than $1 each.  The NTE4046 lists for over $3.50 each.   NTE parts many times do not perform to spec.  NTE does not even publish a Data Sheet for their 4046.  Personally,  I would avoid the NTE version.   Texas Instruments still makes this somewhat obsolete CD4046  IC.   These will probably be discontinued soon like many of the  older logic series  IC's.  

Meyer could not have  used an NTE4046  since NTE did not make a 4046 in 1985.  Meyer probably used a National Semiconductor  or ST Micro version of the 4046.    

Stevie,  you may want to double check your stated "400ma"  output specification.  These are low power devices designed to clock CMOS logic.  2- 4 ma would be about max.   At 12 Volts out and 400ma that would be about 2.4 watts at 50 % duty cycle.  Not likely to happen without some smoke since this chip can only dissipate 500mw max.    

Another interesting IC is the older 74LS629.  This is a dual VCO  Chip ( NO PLL)  good to about  20 mhz.  
Title: Re: Frequency generator perfect waves till 1 mhz, 1 till 10 FET channels
Post by: Login to see usernames on June 17, 2009, 23:23:23 pm

Thanks for your input. Looking good and is helpfull!
If you can add more schematics, please do so! ;) ;)


Title: Re: Frequency generator perfect waves till 1 mhz, 1 till 10 FET channels
Post by: Login to see usernames on June 18, 2009, 21:46:11 pm
The once heralded  Lawton Circuit ruled the boards for a while.  But as anyone who ever built or bought
one knows,  it leaves a lot to be desired.  It is unstable and barely controllable. The adjustments are too
sensitive.  Frequency changes with duty cycle and vise versa.  The gating is out of sync causing partial
pulses.  About all you can say for it,  is that it makes pulses and that with considerable tuning you might
can get the signal you want.

There have been many improved circuits based upon Lawton's use of 555 timers.  But 555 timers, as
handy as they are,  have certain limitations. Some folks have recognized these limitations and have
designed an built circuits that overcome these limitations. One is to use the 555 timer as sawtooth
generator/ oscillator, and then send this to a high speed comparator to make a classic old style PWM
generator.  This can actually work very nicely with well selected components and thoughtful design and
layout.   With the addition of some logic gates and a good FET driver,  a stable working circuit can be
designed that will work very nicely for experimenting with the Meyer VIC.

One of  the problems with the 555 timer in the standard configurations is that duty cycle and frequency
affect each other.   What we need is a sawtooth/triangle generator that is stable over the range of
frequency that we want to use. For Meyer that would  be between 100 Hz and 20K Hz.  But let's extend
it and go up to 50 kHz .  This is well within the capability of a 555.  However there are other options.
One is to use voltage controlled oscillators.  One for the main frequency and one for the gating signal. 

There are quite a few out there to chose from.  Meyer used a CD4046.  These are still available and
work very well up to about 1 mhz. The 74HC4046 is good to 10 mhz.  However this requires a 5 V
regulator.  So if we stay with 4000 series logic we can use 12V and only have 1 regulator. We need 2
VCO's.  We will also need 2 high speed comparators.  The LM311 will work nicely.  The logic gates will clean up
the signal and give about 30 ns rise and fall times to the FET Driver IC.  Pick a FET  driver IC that
operates with CMOS logic up to 16 volts.  Fairchild, National Semi, Microchip, etc all have good
Low Side FET Drivers.

The VCO  puts out a square wave and we need a triangle or sawtooth to do the comparator type
PWM.  This is not a problem.  We can take the triangle signal at the timing cap of the VCO,  buffer it
through an Op Amp and then send it to the comparator.   Adjust the gain and offset of the op amp so
that the triangle signal into the comparator goes from 3 volts to 9 volts.  Adjust the values of the
voltage divider resistors so that the control signal to the comparator goes from a little above 3 volts to a
little below 9 volts. 

Below is one circuit  variation that works very well.     Resistor and Capacitor values need to be
calculated based upon desired frequency range.  See the CD4406 Data Sheet from Texas Instruments.
The values shown will get you in the ball park. 

As you can  probably see  any 2 PWM's  can be fed into the Flip Flop/Mixer section.  They could be
made with 555's or anything else.  The flip flop syncs up the edges of the 2 PWMs assuring that there
will be no incomplete or partial pulses at either the beginning or end of the gated pulse stream output. 
The FET Driver IC assures proper gating of the MOSFET and reduced heat and stress on the FET.
A resistor  can be put in series with the gate should you want to slow down the rise and fall times.
Optionally,  an 18 v Zener diode can be put from the gate to ground to clamp the gate to source voltage
so that it never exceeds the +20V gate to source limit of most MOSFETS.

Bypass and decoupling capacitors are not shown,  but should always be used.  Each IC should have a
100nf  decoupling capacitor as close to it a possible. 

So what we have here is a stable circuit that can operate up to well over 50K Hz and down to 0 Hz  (not
recommended).  Frequency and duty cycle are independent of each other.  Gating is digitally  implemented,
eliminating the sync problems ( partial pulses) with Lawton and other circuits.   A dedicated
FET driver assures proper gating of the MOSFET.

Have fun.
Title: Re: Frequency generator perfect waves till 1 mhz, 1 till 10 FET channels
Post by: Login to see usernames on June 19, 2009, 11:48:30 am

Thank you for your input!
My planning was to scrap the the two red blocks and replace it with a micro controller block and use two PWM signals.
I have a few problems right now mixing the two PWM signals in the micro controller as one output for pulsing and gating. Getting synchronized signals with different duty and frequencies is difficult (for me to program at this moment  ;D).

It has all the signals I need for now for my alternator setup v2. I can use a few IO ports to control the gating and the output signal.
Be careful with micro controller signals, they can not be trusted. You need to have a manual shutdown that switch the system OFF. Or use a pulser signal (software) from the micro controller to generate a interrupt signal to a external chip. When this signal is lost the micro controller has a error and switch the system OFF. I want to control my frequency inverter to regulate the RPMs of the motor with this micro controller.

It think we must use AM signals, the alternator setup is the carrier wave and is mixing with the pulse gate signals (the digital or analog way).
I think we are trying to find the right mixing frequencies in the resonant cavity. My understanding is that no WFC (resonance cavity) is the same (at lost of parameters in here) thus different mixed signals must be used.

 This AM signal is finally generated / used by a VIC of course, but I want to study this the old fashion alternator way first. The alternator is just a (audio) transformer used in a different way till we can replicate a VIC unit.

I hope we can replicate a VIC unit  (or what you want to call this thing)  in parallel here. Later we must find balance in different system setups, to regulate it automatic in resonance I think.

Title: Re: Frequency generator perfect waves till 1 mhz, 1 till 10 FET channels
Post by: Login to see usernames on June 19, 2009, 18:13:14 pm

The mixer and driver  sections of the circuit will  work fine with your 2 microprocessor PWM inputs.
If your micro is  running at 5 volts you can use either CD4000  or 74HC series logic.  

Below is a circuit that will work for you using CD4000 Logic at 5 volts.  You will need an additional
12 regulated supply  for the FET Driver.  

NOTE:   Diagram was updated to add  Output Enable/Disable

Title: Frequency generator perfect waves with TL494
Post by: Login to see usernames on September 27, 2010, 17:02:07 pm

Here another squarewave generator for driving coils!