24V output square wave

ronv

Joined Nov 12, 2008
3,770
It is start voltages a zero volts in the edit spice analysis window. Make sure to run it with a very low frequency input to see the effect of the AC coupling.
 

Thread Starter

Martino Chiro

Joined May 1, 2015
128
I am surprised at the persistant confusion caused by one capacitor.
At the moment, i am confused about the behaviour of my LTSpice simulation, not by the capacitor.
Also with an input square wave of 20Hz i can't reproduce the absence of the first two or three periods in the Vout.
 

ScottWang

Joined Aug 23, 2012
7,397
I am surprised at the persistant confusion caused by one capacitor.
If you want to compare the digital and analog signal, maybe you could do some test, you can using a square wave as 1Hz, 10Hz, 100Hz, 1Khz, 10Khz, 100Khz, 1Mhz to adding to the circuit you provided and adding to a 74HC04(2V~6V)) with 6 inverters, but you only need to use two and one in series with another, using the O'Scope to measure the output waveform, and to see what's their difference.
 

ronv

Joined Nov 12, 2008
3,770
At the moment, i am confused about the behaviour of my LTSpice simulation, not by the capacitor.
Also with an input square wave of 20Hz i can't reproduce the absence of the first two or three periods in the Vout.
Ahh, I see what is bothering you. The bottom transistor doesn't turn on for the first few pulses, but the load brings the output voltage back to ground. With no load or the load returned to +24 the pulses will magically disappear. As a matter of fact I suppose you could eliminate the bottom transistor and the cap and just drive it that way.
 

Thread Starter

Martino Chiro

Joined May 1, 2015
128
Ahh, I see what is bothering you. The bottom transistor doesn't turn on for the first few pulses, but the load brings the output voltage back to ground. With no load or the load returned to +24 the pulses will magically disappear. As a matter of fact I suppose you could eliminate the bottom transistor and the cap and just drive it that way.
Yes ronv, exactly as you explained, here the LTspice result. Now let me put this in a breadboard.
 

Attachments

ScottWang

Joined Aug 23, 2012
7,397
an analog amplifier will amplify a digital signal too. if you want a 50-50 duty cycle 1khz square wave, run a 2 khz square wave thorugh a cmos flip flop first, that will turn it into a 50-50 duty cycle 1 khz.
Although the analog amplifier can be amplifying the digital signal, but they still have some difference in different frequencies, specially is the waveform.

If you want to comparing the digital and analog signal, maybe you could do some test, you can using a square wave as 1Hz, 10Hz, 100Hz, 1Khz, 10Khz, 100Khz, 1Mhz to adding to the circuit as AnalogKid provided and adding to a 74HC04(2V~6V)) with 6 inverters, but you only need to use two and one in series with another, using the O'Scope to measure the output waveform, and to see what's their difference.
 

Thread Starter

Martino Chiro

Joined May 1, 2015
128
Yes ronv, exactly as you explained, here the LTspice result. Now let me put this in a breadboard.
I apologize for this delay.
I put the circuit in a breadboard (the only difference is that i used an BC337-25 instead of the 2N4401) and supply it at 24V and a 10Vpp input square wave. The output doesn't show any lost cycles at the beginning of the capture. The same at lower frequency then 1KHz. I see that "lost cycles" only when i lower the amplitude of the input signal, 8V and 7V as shown in the pictures.
Final results: the circuit proposed by Analog Kid is fine for 1KHz 10Vpp input square wave: it gives the desired output. It is just a little more difficult to assimilate, but it is stimulating.
Every consideration is welcome.
 

Attachments

Thread Starter

Martino Chiro

Joined May 1, 2015
128
I apologize for this delay.
I put the circuit in a breadboard (the only difference is that i used an BC337-25 instead of the 2N4401) and supply it at 24V and a 10Vpp input square wave. The output doesn't show any lost cycles at the beginning of the capture. The same at lower frequency then 1KHz. I see that "lost cycles" only when i lower the amplitude of the input signal, 8V and 7V as shown in the pictures.
Final results: the circuit proposed by Analog Kid is fine for 1KHz 10Vpp input square wave: it gives the desired output. It is just a little more difficult to assimilate, but it is stimulating.
Every consideration is welcome.
Sorry : these are the correct attachments.
 

Attachments

Thread Starter

Martino Chiro

Joined May 1, 2015
128
I know mine works because I simulated it and have built it in various incarnations, lots of times. I would have to analyze the other circuit...

Mine has a higher input impedance (loads the source less).

Mine is shown driving a 1000Ω (24mA load) while the other circuit shows a 47K load.
The circuit is a standard of voltage converter, the output current calculating as this:
Vce_Q1 = 0.2V
Vbe_Q2 = 0.7V
Vce_Q2 = 0.2V
I_R4 = (24V - Vbe_Q2 - Vce_Q1)/R4
= (24V - 0.7V -0.2V)/4.7K
= 23.1V/4.7K
= 4.91 mA
Ic_Q2 = I_R4 * 10 = 49.1 mA
So if you want to reduce the current then you can increase the value of R4, or you want to increase the current then you have to reduce the value of R4, you just calculate as above formula.
Vo = 24V - Vce_Q2 = 24V - 0.2V = 23.8V
R5 is a false load.

View attachment 86463
I put in a breadboard the circuit you proposed, and the output was exactly as espected.
 
Top