I know how to calculate the resistor values for a comparator circuit for both the inverting and non-inverting case given known, fixed high and low switching points, and the supply voltage.

So far so good.

Now I have a situation where the reference for the comparator is derived from an error amplifier, so the voltage at which the comparator will switch is also variable (though bounded to the range [0V, 5V]).

The input voltage is from a ramp generator, and when the input ramp exceeds the reference voltage, the output used to arm a trigger, and is supposed to also discharge the capacitor to ground via a 74HC05. The problem is that the output pulse is very short without hysteresis, and the (1nF) capacitor doesn't discharge fully.

The problem is to either:

1) Configure the comparator hysteresis so that the output pulse lasts long enough (0.5uS should be enough) even though I don't know what the reference voltage is.

or

2) Stretch the output pulse to 0.5uS or so.

I'd welcome suggestions for either (note that I don't really want to add a monostable like a 74HC221 to the circuit for reasons of board space).

Thanks

 

Hello,

Do you have a schematic of the current situation?
How does the threshold of the comparator "follow" the signal?

Greetings,
Bertus

 

Well I could draw a schematic, but all you would see is a +7.5v supply feeding a 39k2 resistor and a 1nF capacitor and then to ground.

There is a comparator + input connected to the junction of the resistor and capacitor. The - input is connected to the threshold voltage. The output is connected to a pullup, and to the input of a 74HC05, and the output of the '05 is connected to the junction of the resistor and capacitor!

Arbitrarily set the - terminal to (say 3v) and observe the behaviour (either on a breadboard, or a simulator). Make sure you start with initial conditions set to zero and enable accurate emulation of digital parts. Typically the output of the comparator is only high for long enough to partially discharge the cap.

 

Try this: Add hysteresis by adding a 10k resistor in series with the (-) input of the comparator, then take an unused section of the HC05 and feed it back from comparator output to the (-) input. This will give you positive feedback, reducing the threshold to near zero while the cap is discharging. If the threshold is too low, you can reduce the value of the 10k resistor, or add a resistor in series with the HC05 output that you just added.

 

Sneaky - I will give that a try ...