# Build a nearly 50% duty cycle, adjustable oscillator using D-FF

Starting from an example circuit, I'd like to simulate and physically realize a D-FF circuit that generates a 50% duty-cycle clock.

Scope of this:

• Clock generator for a test circuit
• Out 0-5v

I don't have particular restrictions on the output clock:

• 45-55% duty cycle it's ok
• Freq. imprecision of 5-10% it's also ok

I'd like to regulate the output frequency between 100kHz and 400kHz and between 10-400Hz , maybe with a switch that adds an additional resistor. For now I've added just a potentiometer to regulate the frequency.

Here's an example circuit simulation on Falstad

This circuit doesn't act correctly. Seems to glitch and frequency regulation is not always correct.

My question is:

Starting from this circuit, is it possible to design a reliable functional oscillator (and how?) Or is it completely wrong?

• Do you have a question? – Andy aka Apr 1 at 8:12
• I've added a question. It's something, I'd like: "I've done this, and it's not working" -> "How to correct it?". In reality it's like "I'd like to do this, and I'm here" -> "How to realize this thing? I'm near or far away?" – Singed Apr 1 at 8:26
• The potmeter's swiper is not connected, so adjusting 'Resistance' will not do anything. – Huisman Apr 1 at 9:54

try this one: two-inverter oscillator based around RS flip-flop

The ~Q output reinforces the R input via the capacitor, but the Q output opposes it via the resistor.The capacitor wins in the short term, but is soon over-come by the resistor. Thich cause the outputs 'Q' '~Q' to change state and the reverse to happen.

Adjusting the resistor adjusts the time taken for the resistor to win and thus the frequency.

The mysterious 1meg resistor stops excessive current flowing into the R input. this is needed because the voltage on the resistor end of the capacitor exceeds the supply voltage and goes below ground.

By swithing the capacitor you can have two frequency ranges that differ by a factoir of about 10000.

The 10u resistor is just there to keep the simulator happy, in real life just use a wire.

I have made oscillators with a 74HC74 that have worked tolerably well because this IC has hysteresis on the clock input. It only makes pulses so the other FF would divide down to get 50-50. But for frequencies in the range you want you should use a proper oscillator circuit and stop playing around with novelty circuits like this. Also, the two frequencies ranges are so far apart that you will need to change capacitors rather than resistors. The CMOS 555 is a much better choice than the bipolar one.

The problem with using an ordinary flip-flop as an astable multivibrator is that its inputs do not have well-defined switching thresholds and no hysteresis.

You would be much better off using something like the ancient , which has both.