Improvised symbol for Schmitt trigger

[Gasboss775]

The Schematic capture app that I use on my tablet ( Quick Copper ) doesn't have a symbol for a Schmitt trigger inverter, but it does allow you to make custom symbols. What do you think of this idea? Would you know what it is if you seen it in a schematic?

 

[audioguru]

Your symbol for a Schmitt Trigger inverter does not look like a Schmitt Trigger.
I use Microsoft Paint to make schematics. Usually I simply copy and paste parts symbols from datasheets or from other schematics.
I fixed yours:

 

[Gasboss775]

Your symbol for a Schmitt Trigger inverter does not look like a Schmitt Trigger.
I use Microsoft Paint to make schematics. Usually I simply copy and paste parts symbols from datasheets or from other schematics.
I fixed yours:

Thanks.

Unfortunately I can't get Microsoft paint for my tablet, though undoubtably there are alternatives. At the moment I don't have working computer or laptop in the house. I have three broken laptops that I need to try to fix!

 

[Gasboss775]

Been playing about with the component editor again and have managed to create a better Schmitt symbol.

 

[cowboybob]

Very nice.

 

[Gasboss775]

Very nice.

Thanks.

Here's an example of it in use as an alarm sounder:

 

[fezder]

shcmitt trigger was my introduction to electronics, teacher showed how to make blinking led with it, I was amazed still use it doing PWM and such

 

[Gasboss775]

shcmitt trigger was my introduction to electronics, teacher showed how to make blinking led with it, I was amazed still use it doing PWM and such

Yeah, I love them too. You can also adjust the frequency of them by varying the supply voltage, higher voltage = lower frequency. I modified a friends TR606 analogue drum machine this way. Made the hex Schmitt inverters power supply variable so that the oscillators could be retuned.

 

[fezder]

You can also adjust the frequency of them by varying the supply voltage, higher voltage = lower frequency.

Neat! Didn't know that

 

[AnalogKid]

Just about everyone here knows what an emitter follower is and what a differential amplifier is. Ever wonder where they came from? ***SAME GUY***

https://en.wikipedia.org/wiki/Otto_Schmitt ::
Otto Herbert Schmitt (April 6, 1913 – January 6, 1998) was an American inventor, engineer, and biophysicist known for his scientific contributions to biophysics and for establishing the field of biomedical engineering. Schmitt also coined the term biomimetics and invented the Schmitt trigger, the cathode follower, the differential amplifier, and the **broken link removed**.

https://en.wikipedia.org/wiki/Schmitt_trigger - Squid nerves? Really!

ak

 

[Gasboss775]

Awesome guy!

 

[schmitt trigger]

There is a reason for all electronic component symbols. When one understands what is going on inside the component, one can appreciate that its symbol somehow indicates its construction or operation.
With the schmitt trigger, the symbol indicates the output vs input transfer function...namely that the transition from high to low follows a different path than that from low to high.

 

[Gasboss775]

I had this idea to make a Schmitt trigger from an XOR gate. Why? I have a load of 4070 quad XOR gate IC's and Schmitt triggers are very useful circuit elements. The XOR is basically being used as a non inverting buffer. Will this be stable or will it oscillate? Tomorrow I will try it out for real.

 

[Gasboss775]

It did work, but not with the resistors shown in the last post. With R2 ( the I/p resistor ) 130K and R1 ( the feedback resistor ) 150K it worked well. With a second XOR gate wired as an inverter, the output was very clean and I was able to build a Schmitt trigger oscillator with it.

Beware that with feedback resistors > OR = 180K you start to observe spurious oscillation on the negative going transitions at about 5 MHz. The higher the resistor value the more oscillations are observed. Note, I did have a 100n decoupling cap in place, so that's not to blame for the oscillations.

 

[MrAl]

There is a reason for all electronic component symbols. When one understands what is going on inside the component, one can appreciate that its symbol somehow indicates its construction or operation.
With the schmitt trigger, the symbol indicates the output vs input transfer function...namely that the transition from high to low follows a different path than that from low to high.

Hi there,

Well, if anyone would know it would be you
Interesting screen name, just glad you did not choose "Chopper Stabilized Amplifier" instead

 

[MrAl]

It did work, but not with the resistors shown in the last post. With R2 ( the I/p resistor ) 130K and R1 ( the feedback resistor ) 150K it worked well. With a second XOR gate wired as an inverter, the output was very clean and I was able to build a Schmitt trigger oscillator with it.

Beware that with feedback resistors > OR = 180K you start to observe spurious oscillation on the negative going transitions at about 5 MHz. The higher the resistor value the more oscillations are observed. Note, I did have a 100n decoupling cap in place, so that's not to blame for the oscillations.

View attachment 96935

Hi,

I like to see stuff like that. Creating higher functioning circuits from lower functioning circuits and components is what it is all about.

A Schmitt Trigger is really just a comparater with hysteresis. Your feedback resistor provides the hysteresis.

If you would like to measure the input signal state transition levels on the raw gate itself we can come up with equations that choose the resistors for optimum operation. That would mean applying an input voltage and watching the output as it changes from low to high and then again from high to low and noting what the inputs were when this happened. There is already some hysteresis in every gate so the two input levels should be different.
Alternately, we could just take a look at the equations for a theoretical model. It will be interesting.

 

[Gasboss775]

MrAl: I did measure the transitions on the raw gate but forgot to write them down.

I did however note the transitions with 130K & 150K f/back resistors; Vhi = 8.12 V, Vlo = 6.59 ( Vdd = 15 V )

I can go back and remeasure if you're interested, the IC is still on the breadboard.

__________________________

The username was a screw up during the registration process, it's actually my real name, but google auto-fill put it in the username box and I hit OK before realizing. Stuck with it now, I guess.

 

[Gasboss775]

Here are the measurement results:

Another experiment with this gate didn't go so well:

 

[MrAl]

Hi,

Thanks for the measurements.

What value did you try for the feedback resistor Rf with the amplifier test? That might have to be around 1Mohm, but you could try other values too.
If you connect just that one resistor, you should see the input bias to about 1/2 Vcc and the output about the same. That is with no other resistors or capacitors connected.
The internal gain is very high so the transfer function curve is very steep, which means the slightest change in input could cause the output to shoot very high (or low). Noise would look like oscillation too, so power supply bypassing would be necessary most likely, right at the chip power pins.

LATER:
Are you sure about those input levels? I ask because CMOS usually has more noise immunity than that would suggest. Is that for the circuit with no resistors at all?

Equation for R1:
R1=(-R2*VL+R2*VH-Vhys*R2)/(VL-VH+VoutL-VoutH)

where
VH is the high threshold,
VL is the low threshold,
VoutL is the output voltage when low,
VoutH is the output voltage when high,
R2 is the feedback resistor value, and
Vhys is the desired hysteresis voltage.
Vhys must be greater than the original hysteresis voltage.

 

[Gasboss775]

Hi,

Thanks for the measurements.

What value did you try for the feedback resistor Rf with the amplifier test? That might have to be around 1Mohm, but you could try other values too.
If you connect just that one resistor, you should see the input bias to about 1/2 Vcc and the output about the same. That is with no other resistors or capacitors connected.
The internal gain is very high so the transfer function curve is very steep, which means the slightest change in input could cause the output to shoot very high (or low). Noise would look like oscillation too, so power supply bypassing would be necessary most likely, right at the chip power pins.

I tried a few values for both R feedback and Rin, including 1.5 & 3.3M for feedback. Was definitely oscillations observed, almost sinusoidal at around 700 KHz. I even tried a small capacitor (39pF) in parallel with Rfb but still got oscillation. There was a 100n ( 0.1uF ) ceramic capacitor across the power rails.
I've got this circuit to work previously using a 4069UB.

LATER:
Are you sure about those input levels? I ask because CMOS usually has more noise immunity than that would suggest. Is that for the circuit with no resistors at all?

Yes, without the resistors. I expected there to be a bigger gap myself.