I managed to change comparator to LMC7221A
A) I do not know what voltage are you showing? You need to label this on a schematic diagram. Such as voltage at “U2 pin 3”
B) The LMC7221 has open drain output. Did you add a pullup resistor? Without a pullup resistor your oscillator will probably not run. With a pullup resistor, the pullup resistance will figure into the frequency and duty cycle calculation.
The comparator which I recommended has active pullup. TI also makes similar.
Another possibility is to use LMC7221 with pullup resistor, and drive a 5V CMOS gate to provide the output and feedback waveform.
I insert LMC7221 into that circuit…
OK. I am on it.
Beware of inversion. In other words if you use an inverter gate, then you need to swap the comparator input connections. If you post again please include a new schematic with somewhat better resolution. Better yet would be to also include the KiCad format schematic.
LMV7239 would be a better option because it has push pull output. There is a bewildering choice of devices from TI; different pinouts and there is probably a cheaper choice with push pull output; cheaper than LMV7239.
You know I am trapped. I try to figure out how to debug a .lib file and figure out how thing works at the same time. Probably I will be better off if I read datasheet carefully in the first place.
There are many “fine points” which differentiate the devices which we are discussing:
Rail-rail input, rail-rail output (or not). Many comparators such as LM339 work well down close to the negative rail, but the input becomes “blind” within 1-3 volts of the positive rail…
Open drain/collector, bipolar active pullup, CMOS active pullup
Input bias current
Speed (probably not so critical in your situation)
Output drive current capability: 2-3 mA up to maybe 50 mA.
Pinout
These all can affect the suitability of an IC for a design. Hopefully they are all adequately defined by the datasheet.
As for me: My best trick is power conversion circuit design. I have been pretty good at composing a design and having it just work on the bench. But when it comes to simulation, I am mostly just a dumb user. With a few simple/obvious exceptions, I would not know how to edit a .lib file.
A few years ago, another guy and I were both consulting for the same small company. This other guy was good at simulation modeling. He designed some circuitry which played well in steady state operation according to his simulation. But I could see that his design tended to self-destruct during power up and power down. And it did that…
The simulation I provide here is just to look at. I am inclined to do this circuitry on breadboard and see the things.
control.7z (27.1 KB)
I have fixed a couple of things…
control_BobZ1.kicad_sch (48.0 KB)
Your connection of R5 was worng.
BTW it is best to rename the file every time you post or transmit it. A version number or date+revision does the trick.
I think you need to think a bit more when you look at a circuit. Do you know about Norton and Thevenin equivalents?
I will keep it my mind, I hope I will follow it in the future.
Looks like it is more readable now. I never saw a circuitry that way before.
Heard some I may try the method later on on the circuit.
See this:
To explain the values and connection of R5,R6,R8.
This is very cool script, thank you. I will try to find some time tomorrow to read it.