Improvised symbol for Schmitt trigger

[MrAl]

Hello again,

Well, you had posted 7.52v as input high and 7.41v as low, so that's only 0.11v of noise immunity voltage. There's no logic family that has it that low so i suspect something must be wrong somewhere.
For CMOS, the spec is usually around 30 percent for low and 70 percent for high, so for a Vcc of 10 volts that would be 3v and 7v.
Perhaps you can check another gate as there may be something wrong with this one. Maybe check a different kind of gate too such as an inverter.

 

[Gasboss775]

Hello again,

Well, you had posted 7.52v as input high and 7.41v as low, so that's only 0.11v of noise immunity voltage. There's no logic family that has it that low so i suspect something must be wrong somewhere.
For CMOS, the spec is usually around 30 percent for low and 70 percent for high, so for a Vcc of 10 volts that would be 3v and 7v.
Perhaps you can check another gate as there may be something wrong with this one. Maybe check a different kind of gate too such as an inverter.

Yes, I think I'll try some other measurements as this doesn't seem right. Just to clarify the method that I used was to increase the input voltage to the point where there is significant shift in the output, note this input voltage as the high threshold, then reduce the input voltage to the point where there is a significant shift in the output in the opposite sense then to note this input voltage as being the lower threshold.

 

[Gasboss775]

I have augmented the post #14 circuit to make a VCO ( it works, I built it ! )

By adding a monostable in place of the toggle flipflop a variable PWM circuit could be made though it would also have variable frequency, which may or may not be a problem depending on what you use it for.

 

[Gasboss775]

Neat! Didn't know that

 

[Gasboss775]

Hello again,

Well, you had posted 7.52v as input high and 7.41v as low, so that's only 0.11v of noise immunity voltage. There's no logic family that has it that low so i suspect something must be wrong somewhere.
For CMOS, the spec is usually around 30 percent for low and 70 percent for high, so for a Vcc of 10 volts that would be 3v and 7v.
Perhaps you can check another gate as there may be something wrong with this one. Maybe check a different kind of gate too such as an inverter.

Evidently I am going about this the wrong way. I have noted the input and output voltages for a 4069 inverter for input from 0 to 15 volts. I will compile the results and post them later, perhaps you could explain to me how to identify the high and low threshold points.

 

[MrAl]

Hello again,

Well, a variable voltage input with low pass filter like 10k and 0.1uf might do it. That will filter any momentary spikes when the pot for adjustment is turned.

It would be good if you could test a Schmitt Trigger device too, such as the 74HC14 (one section), or similar.

 

[MrAl]

Hello again,

It seems that the threshold for the 'normal' CMOS gate could be 1/2 Vcc (which would be 7.5v at 15v Vcc) it's just the high and low definitions that are different. This must be so or else the linearized gate would not work. That is, when we place a resistor from output to input with an inverting gate of some kind, the gate is self biased into linear operation. That means there can be no hysteresis in a 'normal' gate, or at least one that is capable of linear operation.

The Schmitt Trigger on the other hand must have two distinct levels, so that makes it very different than the typical gate.

So in the formula for the resistors we can just make VL=VH and set that to about 1/2 Vcc and that should give us the right results. For the Schmitt Trigger however, we'd still have to set these two to two different values.

This actually makes it a little easier for normal gates.

 

[Gasboss775]

Hello again,

It seems that the threshold for the 'normal' CMOS gate could be 1/2 Vcc (which would be 7.5v at 15v Vcc) it's just the high and low definitions that are different. This must be so or else the linearized gate would not work. That is, when we place a resistor from output to input with an inverting gate of some kind, the gate is self biased into linear operation. That means there can be no hysteresis in a 'normal' gate, or at least one that is capable of linear operation.

The Schmitt Trigger on the other hand must have two distinct levels, so that makes it very different than the typical gate.

So in the formula for the resistors we can just make VL=VH and set that to about 1/2 Vcc and that should give us the right results. For the Schmitt Trigger however, we'd still have to set these two to two different values.

This actually makes it a little easier for normal gates.

On the 4069 data sheet ( National Semiconductor, yes they're old ones!! ) the definition of the low threshold is that which puts the output at 13.5 volts ( for VDD = 15 V ) and the upper threshold is that which places the output at 1.5 V ( again with VDD = 15 V )

 

[MrAl]

On the 4069 data sheet ( National Semiconductor, yes they're old ones!! ) the definition of the low threshold is that which puts the output at 13.5 volts ( for VDD = 15 V ) and the upper threshold is that which places the output at 1.5 V ( again with VDD = 15 V )

Hello again,

Well in any case, the device would not work in a linear mode if there was built in hysteresis because it would have to oscillate or stay at one fixed level.
So the transfer function causes the output to shift quickly from one state to the other, which makes the gate look like one big amplifier with a huge gain.

Here is the formula for a 'normal' gate then:
R1=Vhys*R2/(VoutH-VoutL)

The switch levels will be centered around 1/2 Vcc, so if Vhys=2 and Vcc=10 then the switch levels will be 4v and 6v for example. You could easily test this too if you like.

As an approximation, we could also write:
R1=Vhys*R2/Vcc

so with Vcc=10v for example we have:
R1=Vhys*R2/10

and if Vhys=2 and R2=100k then we have:
R1=2*100k/10=20k

For another example, with Vcc=15 and R2=150k we have:
R1=2*150k/15=20k

 

[Gasboss775]

Another improvised symbol, this time for a Schmitt input Nand gate.

Does anyone know why there were never any other logic gates with Schmitt inputs other than the NAND gate and the NOT / inverter gate?

 

[jpanhalt]

Does anyone know why there were never any other logic gates with Schmitt inputs other than the NAND gate and the NOT / inverter gate?

There is a simple buffer (non inverter) and per se "and":
https://www.ti.com/lit/ds/symlink/sn74lvc1g17-q1.pdf
https://www.ti.com/product/SN74HC7001

I didn't check for others, but assume they may exist too.

John

 

[Gasboss775]

There is a simple buffer (non inverter) and per se "and":
https://www.ti.com/lit/ds/symlink/sn74lvc1g17-q1.pdf
https://www.ti.com/product/SN74HC7001

I didn't check for others, but assume they may exist too.

John

Thanks John, When it comes to digital I'm about 20 years out of date!

 

[MrAl]

74HC002 (not the 74HC02)

 

[Gasboss775]

74HC002 (not the 74HC02)

Can't find any sign of it. What is it?

 

[Gasboss775]

Another Schmitt trigger PWM circuit. Works well up to around 20Khz with TL082. Could go faster with a comparator.

 

[MrAl]

Can't find any sign of it. What is it?

Hi,

Well that was a NOR gate with Schmitt Trigger inputs.

Also i realized that we might have a Schmitt Trigger input if we use two resistors like we had been. One input resistor, one feedback resistor. You can try that if you like to see how well it matches up to the formula. It should be close. I cant try this just yet myself.

 

[jpanhalt]

Hi,

Well that was a NOR gate with Schmitt Trigger inputs.

Also i realized that we might have a Schmitt Trigger input if we use two resistors like we had been. One input resistor, one feedback resistor. You can try that if you like to see how well it matches up to the formula. It should be close. I cant try this just yet myself.

"Was" is the operative term. That chip is so old, no one lists it.

I'll give you $0.10 for one with free S&H.

John

 

[Gasboss775]

Idea for Schmitt input NOR gate:

 

[gophert]

Idea for Schmitt input NOR gate:

View attachment 97111

If I set one input to 2.5V and sweep the other between 0 and 5V, I do not get any change in the output.

What are the rules for a two-input schmitt trigger? Is this an adjusted test?

On the other hand, I set one input to triangle wave at 10 Hz and a second triangle wave at 9 Hz. No output signal change.

I cannot imagine a situation where such a device would be needed, other than for completeness of a set for those of us who have OCD.

 

[MrAl]

"Was" is the operative term. That chip is so old, no one lists it.

I'll give you $0.10 for one with free S&H.

John

Hi,

Referring again to the 74HC002 (not the 74HC02)...
I could not find one either. I never purchased one myself either at any time in the past, just read about them.