I managed to find 555 timer at TLC555 data sheet, product information and support | TI.com. And there is something wrong with the result. Can someone debug the .lib file?
TLC555_astable.7z (1.7 MB)
I do not use NGSpice. But I am familiar with the 555 and I might be able to help if you show what is the difference between your waveform and that in the datasheet.
BTW the bypass capacitor values shown were devised before modern MLCCs (see original date on the datasheet). Unless you are using all through hole components, I would probably use a 1.0 to 4.7 uF 0603 chip at C1 and C3. If you are powering from 5V, the voltage rating of these capacitors could be 10V or more.
OK…
-
The waveform from the TI datasheet is measured at C2. Where is your waveform observed?
-
Your waveform indicates current and voltage. Which is it? I am almost certain that you are showing us a current waveform. I see a current scale but no voltage scale. There should be no negative voltage in this circuit.
-
Is your nomenclature such as 0.15u;25V workable? I do not know if NG Spice can resolve that. I would think that capacitance and voltage rating need to be shown separately but I could be wrong. I would omit the voltage rating.
Voltage measurement at C2 is the same as output so I decided to measure current on C2. I changed 0.15u;25V to 0.15u;12V before simulation…
NO. C2 should have an asymmetrical sawtooth voltage on it as you showed from the datasheet.
Output should be an asymmetrical pulse with a high duty cycle of approximately 8/11 = 73%.
And I suspect that you are probing the component (in LTSpice that will show you current.) You should be probing the wire (the net) to see voltage.
Give us an output plot which indicates voltage at C2 and another showing voltage at OUT.
That is better but the waveform seems sort of noisy.
I do not know the significance: It looks like your capacitors are all electrolytic. If they include some ESR that could be a problem. It also does not make any sense in the real world. Capacitors up to at least a few uF should be ceramic, non polarized.
I would change the cap values according to my earlier post. Also set a maximum step size…maybe 1 uSec?
The output voltage waveform now looks…reasonable.
Capacitors up to at least a few uF should be ceramic, non polarized.
Thank you for pointing it out.
Also set a maximum step size…maybe 1 uSec?
It looks better now
That is better but the waveform seems sort of noisy.
What can be done to improve?
Just those items I mentioned. Does the capacitor voltage waveform still look (irregular; noisy)?
To make an oscillator, I would do something more like this:
But do not try to get more than a few mA of drive out of it. This comparator is < $0.40 at DigiKey.
Today I am done. I will take care of capacitors tomorrow. The wave forms I will put here. Thank you.
Sounds good. But one more thing…I like this oscillator schematic better. R1 and R2 just evenly split the input supply.
Good values for a slow oscillator using a CMOS comparator would be:
R1=R2=499K
R3=249K
R4=124K
C1 based on desired frequency.
You can relax with a relaxation oscillator. (I never understood that term…) 
That sawtooth looks much better with 1 uSec max step size. It all looks even (not noisy) as it should be. 
You must have made some significant change to cause a different result in the last two red pulse voltage waveforms?
The upper one is 5V amplitude and >50% duty cycle. The lower one is about 19V amplitude and <50% duty cycle. Something changed?
Once you get everything working, it is time for some fun experimentation (in simulation only!!):
Try increasing your input voltage slightly…like to maybe 5000V and see what happens??? I suspect that everything will still work.
That’s right. I had something different in my ind though. It was to connect the two circuits and …
I have on output
it is not what I expected to get. My calculations
Do you see a mistake in them?
-
Your plot says “voltage” but does not indicate which voltage where?
-
R7 should be 124K (not 124 ohms) = the parallel combination of R5,R6,R8. But…
-
Those 4 resistor values were given for a CMOS comparator with picoamp level input bias current. An LM358 is a bipolar op amp:
3a) Input bias current is in 10’s to 100’s of nanoamps instead of picoamps.
3b) Output slew is in microseconds instead nanoseconds.
3c) Output does not swing rail-rail as does the recommended comparator.
3d) Nothing in 3 above is a “show stopper” but all will degrade accuracy. You may find that the op amp output voltage swing does not go low enough or high enough.
3e) I would recommend scaling all those resistor values down by a few X. This will require a bigger C3 for the same frequency. But items in (3) above are probably not critical. -
Schematic is a bit too small to read. How are the two ICs interconnected?
-
You have many cross wire dot junctions. Schematics are a bit more difficult to read. Do it like this:
Can you give me the same but for bipolar op amp as LM358
with a sign, try to magnify the page with Ctrl - +. Ctrl - - is to undo.
There is a function generator in LM358 datasheet to produce square waves. And thank you for what you say.I will be in touch.
I cannot read the letters in the net names. Your .png screenshot is only 16KB and the resolution is just not there. (I downloaded it and viewed it with IRFANVIEW. It is what it is.)
It looks to me like U1 pin 2 is connected to “a”??
It looks to me like U2 pin 4 is connected to “8”??
Why not divide all of those resistors (except R7) by 5??
R5=R6=100K
R8=49.9K
R7=24.9K
These are all standard E96 1% values.
As for your present connection of the 555 and I would need to study the datasheet to see just what that is supposed to do. Why not just fix R7 first and make it 124K? That MIGHT make a dig bifference.
Do not bother yourself that much. I will take care of it.