Pulse Width Modulation DC Motor Control

[audioguru2]

Hello,
I have been researching circuits for PWM speed control of DC motors and found our project here:
http://www.electronics-lab.com/projects/motor_light/033/index.html
Has anybody tried it?
What does it do?
It appears to be a simple adjustable-frequency "Classic" 2-inverter CMOS oscillator driving a Mosfet.
Comments are welcome. Its schematic is here:

View attachment 35782

 

[ante1]

Audioguru,

I tried it in ISIS, no go! Dead in the water! Q1 is open all the time and the motor runs full speed. No pulse just a strait line on the scope.

 

[audioguru2]

Hi Ante,
I don't know what's wrong with your simulator, but I have seen many oscillators like this, and Pin 4 of the CD4011 should be swinging from 0V to +6V with about a 50% duty-cycle. Since the gate threshold of the IRF511 is typically 3V, it should supply pulses of about 2A. Don't turn the pot to minimum or the oscillator will stop!

Do you see PWM occuring with this circuit? I don't.

 

[ante1]

It

 

[audioguru2]

Hi Ante,
Does your simulator have a 74C00? Or 4001, 74C02, 4069, 74C04. They will all work the same way in this circuit.

 

[ante1]

This is what I have:

74ALS00, 74AS00, 74HC00, 74HCT00, 74LS00, 74S00 and 74STD00.
74ALS02, 74AS02, 74HC02, 74HCT02, 74LS02, 74S02 and 74STD02.
74ALS04, 74AS04, 74HC04, 74HCT04, 74LS04, 74S04 and 74STD04.

The 4001 and 4069 is the same as the 4011, it has hidden vcc and vss pins?!

Maybe I have to read the manual ;D, there must be some way to feed these IC:s.

 

[audioguru2]

Hi Ante,
Never mind your darn simulator. Just quickly build this simple circuit.
My point is that it does not produce Pulse Width Modulation because it is just a variable frequency oscillator with a constant duty-cycle.

The author says, "This (is a) simple pulse width modulation DC motor control (circuit)". I don't believe him. Do you?
The author also says, "R2 adjusts the speed of the oscillator and therefore the speed of M1." Again I don't believe him. Do you?

 

[GPG1]

Try this

View attachment 35807

 

[audioguru2]

Thanks, GPG,
Of course your circuit will produce PWM and control the motor's speed by adjusting the pot. The diodes allow the oscillator's output duty-cycle to be varied over a wide range while keeping its frequency constant.
The paralleled inverters at its output charge the Mosfet's gate capacitance quickly, reducing its heating during switching transitions.
Your circuit does motor speed control very well that our project does not do at all!

Did you know that when CMOS gates are used as inverters, only one input should be used, with the extra input(s) turned-on by connecting them to a supply (positive for NAND gates, ground for NOR gates)? This way, the input threshold voltage is much closer to half supply, allowing a 50-50 duty-cycle with your pot at half-way, and when used in an ordinary oscillator.

 

[audioguru2]

Thanks, Sasi,
Not only does your circuit control the motor's speed very well with PWM, it reverses its direction, too! ;D
What would happen if you rev'd the motor full speed forward, then quickly slammed it into reverse? Over-temperature shutdown of the tiny IC?

 

[Kevin Weddle]

I have seen that there is more than one way to build this oscillator. All the oscillators posted seem to pan out. The one that I have has a capacitor between the two gates and a pullup on the output of the first gate. It does not show the feedback resistor but I know now that it must be there. Surely we can use any of the posted oscillators with the same degree of accuracy.

 

[ante1]

Audioguru,

Yes you are right, just frequency adjustment and not even a nice looking square on the scope! 8)

 

[ante1]

It changes the sound of the motor not the rpm. I agree, dump it!

 

[audioguru2]

The project is still there, unchanged.
Where's my fix for it? ;D

 

[audioguru2]

Instead of dumping it, at least the project should have a link to this thread.

 

[ante1]

Audioguru,

The circuit generates a bad square wave, which is not very effective for PWM purposes. So there are circuits as simple but with better performance.

 

[audioguru2]

Hi Ante,
I'm glad that you agree with me that this project doesn't work.

It is really a new topic that you find its oscillator producing a lousy square wave. Kevin also mentioned that his "Classic CMOS 2-inverter Oscillator" makes lousy square waves. But the ones that I have made do not! Why?
1) I use mine with a 12V supply, where the CMOS has a much lower output impedance so that there is less loss at its output that is driving the load of the frequency-adjustment resistor through the timing capacitor. Using a supply of only 5V or 6V forces the resulting high-impedance output to struggle at trying to drive the oscillator's internal load.
2) I use a minimum value of 33K ohms for the frequency-adjustment resistor, so the output is hardly loaded by it. Maybe you and Kevin are using a low-value resistor that loads the output too much.
3) I always try to use B-series (buffered) gates instead of A-series (original version), nor 4069UB (unbuffered) inverters.
4) If there are any spare inverters or gates in a package, I always parallel them to obtain a lower output impedance.

Try these values in this oscillator and see how perfect its square wave is:

View attachment 35827