Hi everyone,

I have a little design challenge here. I made a ramp oscillator with a 555. Simple enough.
A ramp oscillator is simple if you stick to the 2/3VCC and 1/3VCC treshold. Since I use a 5V VCC, it would simply oscillate between 1.66v and 3.33V. I use a constant current source to charge the capacitor C1.

The interesting part is to adjust that lower limit to something closer to 0V. Ideally I would fix that lower limit somewhere around 0.2V.

In order to do so, I used a simple diode D1 to tie the OUTPUT to the CONTROL input. Doing so, when the output drops to 0V, the input CONTROL voltage drops to the Vf of the diode, in my case somewhere around 0,4V. Now that the 2/3VCC point of the 3 resistor bridge is forced to the diode Vf, the lower voltage comparison should be 0.2V instead of 1.66v.

I realized this circuit and there is a interesting behaviour that I cannot explain. Maybe someone will be able.
The diode makes the Ramp signal back to zero! I would have expected the lower comparator to Set the toggle when the voltage of the capacitor gets lower than the Vf/2 but it doesn't. I am missing something here, maybe someone can explain.

The two waveform are as follow:

Picture 1, Yellow is OUTPUT3, Pink is CONT5, and Blue is THRES+TRIG+DISCH (the actual ramp signal in the dicharge portion)

Picture 2 is without the feedback diode (open circuit), Yellow is OUTPUT3, Blue is THRES+TRIG+DISCH (ramp)

I would have expected the Ramp to stop discharging once the BLUE ramp crossed the PINK diode control voltage feedback. But unfortunately it doesn't.

Thanks in advance for your help, any input will help shine the light on this.
JF

àà

 

One method of reducing the voltage ramp on pin 6 of the 555 timer is to reduce the voltage on the control voltage at pin 5.

However my 555 timer cookbook states that the minimum voltage at pin 5 is 0.17 x Vcc in the astable mode. Bear in mind that this will cause a change in the timing of the circuit.

A simpler method to achieve what you want would be to feed the 1/3 – 2/3 ramp signal into a unity gain op-amp with added offset circuitry to allow adjustment of the signal in relation to 0V.

 

I used a current-mirror to generate the constant-current for the timing capacitor (below).
It is simple and has the advantage of the frequency being essentially independent of the supply voltage.

I would use an op amp with offset to get the zero voltage you want, as Hymie suggested.

 

Thank you gentlemen, I appreciate your feedback.

Maybe I should have started with the full picture. I make a 0-10 to PWM converter. The circuit in attachment works pretty well including the Diode on the 555 which make a perfect ramp from 0V to 3.5V. Frequency doesn't need to be precise. The problematic I face is in relation with the following points:

1- the incoming signal 0-10V rise a little with temperature (operation -30C to 80C) and what should be a 0V is in fact 0.1V per say. If it wasn't from that, I would be happy with the current circuit. But I need to rise this lower voltage of the ramp to 0.2V.

2- This brings me to the original post above. I am trying to understand the principle behind the diode between OUT3 & CONTROL5 that I have seen in some other documents, and I cannot explain the behavior. In reality the voltage brought to the comparators is Vf of the Diode. I don't understand why the comparator doesn't pickup, (probably because of the common mode of the comparator within the 555) The voltage could be too low to be picked up, hence the discharge continuing until the capacitor is completely discharged. I shall go have a look at the full diagram of the 555 not only of the function block.

The idea of using an op amp to offset the resulting ramp signal to 0.2V came to my mind a while ago but it made the circuit a bit more complicated with more parts. (trying to keep the cost low for repeatability). At first I thought of Negative voltage offset which was way to complicated with too may parts to implement. Now you make me think I could try capacitive coupling with an offset based on 2V Zener. But it does make the circuit more expensive. Will test it.

What bothers me the most, is not understanding why the tricks I did with the diode doesn't follow the laws of electricity. (I know it is most likely me not picking up on one detail)

 

the incoming signal 0-10V rise a little with temperature (operation -30C to 80C)

Silicon diodes have a temperature coefficient of about 2mV/°C so is that a problem in your circuit?

 

But I need to rise this lower voltage of the ramp to 0.2V.

You can add a small value resistor (47ohm) in series with C1 to raise the low end of the ramp.

 

I used a current-mirror to generate the constant-current for the timing capacitor (below).
It is simple and has the advantage of the frequency being essentially independent of the supply voltage.

I would use an op amp with offset to get the zero voltage you want, as Hymie suggested.

View attachment 230730

Besides the offset you could have a gain of 3 stage to get you back to a full 5V to GND swing. Make sure the part and the power supply will allow all of this.