How can this circuit work as a toggle on-off switch?

Thread Starter

Jony130

Joined Feb 17, 2009
5,487
I found this circuit

1.png
And this circuit was advertised by the seller as a toggle on-off switch but to be honest I do not see how it can work this way.
What do you think about this? Can it work?
 

AnalogKid

Joined Aug 1, 2013
10,986
No. The switch pulls both ff inputs below the CMOS transition level. Vout is forced high while the switch is closed. After the switch is released the output is undetermined.

Here is the classic toggle ff circuit using gates. Any inverting gates will work.

ak

NOTE: the lower 100K resistor should be smaller, like 10K or less.

upload_2019-3-11_12-59-23.png
 
Last edited:

sghioto

Joined Dec 31, 2017
5,376
The circuit posted above works but when pressing the button a second time to set the output low you have to release the switch within about .2 seconds or the output will return to a high state. Changing the 1uf to 4.7uf will give you a full second to release the button making it more practical if using manually.
SG
 

crutschow

Joined Mar 14, 2008
34,281
You can also use a D flip-flop for a toggle circuit as shown below:
R1C1 provides a debounce function if the input is from a mechanical switch.

upload_2019-3-11_12-1-8.png
 
Last edited:

crutschow

Joined Mar 14, 2008
34,281
And here's a variation of the post #2 circuit that requires only one capacitor, and is insensitive to how long the button is held.
It is also ignores contact bounce.

upload_2019-3-11_15-41-31.png
 
Last edited:

AnalogKid

Joined Aug 1, 2013
10,986
Using a capacitor to control ramp rate / delay into a non schmidt triggered conditioned CMOS gate input generally a no-no.
Fortunately for us, the circuit posted has positive feedback around two inverters in sreies, an excellent example of a Schmitt trigger from 1961.

ak
 
Last edited:

danadak

Joined Mar 10, 2018
4,057
Posted circuit does not sim well. Note I had to use NOR could not find
NAND in simulator. I also threw in a small amount of stray C.

I used 10 mV of noise added to ramp.

upload_2019-3-11_19-19-3.png


Regards, Dana.
 

danadak

Joined Mar 10, 2018
4,057
Post 2, except I use a ramp rather than a cap. And of course the gate difference.
I have done the same sim in the past with inverters, same result. It makes sense
as well.

Basically all I and TI ap note trying to explain is when transiting thru threshold region
with some ever present noise if the ramp rate, Trise, too slow lots of transitions occur,
undesired in a logic element. Even with + fdbk. And this does not look at ground bounce
generated noise referred back to gate input effects, more undesired effects.

Over the years I have seen many CMOS gate circuits to implement delay, pulses,
etc.. and always thought long times spend at Vth asking for problems.

Regards, Dana.
 

crutschow

Joined Mar 14, 2008
34,281
always thought long times spend at Vth asking for problems.
True. I have no argument with that.
But the circuits we posted uses a switch to rapidly switch through Vth, not slowly as the voltage source you used.
The zero source resistance eliminates the positive feedback that provides the hysteresis to prevent the oscillations you show.
Add some source resistance and the oscillations will stop.

Look at the adage at the bottom of my posts.
 

danadak

Joined Mar 10, 2018
4,057
crutschow, I must be blind as a bat, I see no source R in either post # 2 or # 8 or # 9,
other than cap esr ? Because the caps are acting as the source......or so I think.

What adage did you mention ? You mean this "--The important thing about having knowledge is knowing where it correctly applies.--"

Great adage. I don't have one yet for my posts, I will have to think about that.
Like "Is the earth flat at relativistic speeds ?"


Regards, Dana.
 
Last edited:

crutschow

Joined Mar 14, 2008
34,281
Because the caps are acting as the source......or so I think
Okay, I see your point, and its low impedance would kill the positive feedback voltage.
But the switch uses that source to rapidly switch the circuit between states, not slowly as in your sim
And 10mV is a lot of noise for such a circuit configuration.

If hysteresis were needed (and I don't think it is for these circuits) you can add a small resistor (e.g 100Ω) in series with the cap in my post #9 circuit.
 

danadak

Joined Mar 10, 2018
4,057
Yes I should sim switch bounce as well. Then noise becomes the switch bounce
effects piled on top of linear element noise.

Picked the 10 mV just thinking in a CMOS 3 or 5V circuit, switching noise would
generate that. But then it would be instructive to characterize G of gate which I
have not done.

Hysteresis is good. I defer to TI ap note.

Regards, Dana.
 

AnalogKid

Joined Aug 1, 2013
10,986
The circuit posted above works but when pressing the button a second time to set the output low you have to release the switch within about .2 seconds or the output will return to a high state. Changing the 1uf to 4.7uf will give you a full second to release the button making it more practical if using manually.
That is solved, as in Wally's circuit, by reducing the lower 100K resistor to something smaller than the other one, like 10% or less (his is 4.7K). It then swamps out the impedance of the R-C timing network, holding the latch in state no matter the voltage of the cap. I updated post #2.

Note - this has *nothing* to do with a slow voltage ramp causing a problem. There is no slow voltage ramp on any CMOS input in this circuit.

ak
 

danadak

Joined Mar 10, 2018
4,057
In post # 2 circuit, if there is no slow Tr of an input, how does 1 uF get charged rapidly,
like within a prop delay of the gate(s) ?

Regards, Dana.
 
Top