So I need to see all the voltages while it's ON with the button pressed, trying to turn it off.

They are included im my previous post. Images #2,#4,#6. Or you need something else?

 

They are included im my previous post. Images #2,#4,#6. Or you need something else?

It's not clear what trace is what voltage.
With the output on (with a 1kΩ load), then press and hold the push-button.
With the button pressed, show and label the voltages at nodes out, G1, G2, and 2.

 

It's not clear what trace is what voltage.
With the output on (with a 1kΩ load), then press and hold the push-button.
With the button pressed, show and label the voltages at nodes out, G1, G2, and 2.

On all the images, yellow is Vg2 and blue are Vg1, Out, V2. Load is mentioned relay with HF41F relay. Do you want me to put a 1kOhm resistor and make an oscilloscope images again?

 

Do you want me to put a 1kOhm resistor and make an oscilloscope images again?

Did you not have an output load previously?
I won't work properly without a load.

I would like the voltages as stated in my post #62.
After 1 second, release the button while recording the voltages.

 

I was always having earlier mentioned relay attached to the circuit. Resistance of the relay circuitry is arround 600Ohm (measured with voltage/current consumption by the relay).
However, now I connected 1kOhm resistor at the output. Also connected like 10m of wire from the circuit to the momentary switch. The results (yellow is always Vg2):
- Output (blue):





- V2 (blue)





-Vg1 (blue):

 

Okay.
It appears that it is trying to turn off (the Vg2 negative going spike), but C1 is not large enough to hold G2 low, long enough.
Try increasing the value of C1 to a few microfarad and see if that helps.

 

Okay.
It appears that it is trying to turn off (the Vg2 negative going spike), but C1 is not large enough to hold G2 low, long enough.
Try increasing the value of C1 to a few microfarad and see if that helps.

Added 10uF and now it works With 10m of cable between momentary switch and circuit. It can be turned on immediately but to switch it off, you have to wait about 6 seconds before pressing the switch. So, the sequence is:
1. power on the circuit
2. switch output on by pressing the momentary switch
3. wait 6 seconds
4. press momentary switch to switch the output off
5. switch on by pressing momentary switch
6. repeat from step 3.

This delay between switching ON and then be able to switch OFF depends on C1? Does it also depend on the line length between momentary switch and circuit?

 

If the cable length seems to be having an effect then it seems like noise of some type, picked up through the cable, is having an effect. That was my thinking when I suggested using a shielded cable.
The circuit shown last is not noise resistant on the push button connections. Those are moderately high impedance points..

 

If you have this setup and working, can you try it with different wire lengths and tell us?

 

This delay between switching ON and then be able to switch OFF depends on C1? Does it also depend on the line length between momentary switch and circuit?

The wire length should have zero effect on any of the circuit's time constants.

ak

 

This delay between switching ON and then be able to switch OFF depends on C1?

Yes. the delay before you can either turn it ON or OFF Is determined by the R2C1 time-constant.
If you look at the purple trace V(2) in post #53 you will see the rise/fall time for that.
Since it is now working, you can likely reduce C1 some, to reduce that time while still having the circuit work, if you want that time to be shorter..

The circuit shown last is not noise resistant on the push button connections. Those are moderately high impedance points..

Those connections have a high DC impedance, but the impedance to AC noise is determined by the values of C1 and C3.
Using twisted-pair wire to the switch should reduce the sensitivity to noise, with shielded twisted-pair wires (tie shield to circuit common) even better.
(Note, you can readily twist two wires by putting one end of the pair in a vice, and the other pair end to a drill chuck. Holding the wires taut will then allow the drill to rapidly twist the wires).

 

I have tested the circuit with shorter wire to the momentary switch. All works like described when 10m of cable.
crutschow thank you a very very much! And thanks for sharing your knowledge that is "...knowing where it correctly applies"

 

Yes. the delay before you can either turn it ON or OFF Is determined by the R2C1 time-constant.
If you look at the purple trace V(2) in post #53 you will see the rise/fall time for that.
Since it is now working, you can likely reduce C1 some, to reduce that time while still having the circuit work, if you want that time to be shorter..
Those connections have a high DC impedance, but the impedance to AC noise is determined by the values of C1 and C3.
Using twisted-pair wire to the switch should reduce the sensitivity to noise, with shielded twisted-pair wires (tie shield to circuit common) even better.
(Note, you can readily twist two wires by putting one end of the pair in a vice, and the other pair end to a drill chuck. Holding the wires taut will then allow the drill to rapidly twist the wires).

Making twisted pair as described produces a poor version because each wire is also twisted. AND the twisting is never uniform along the length.for installing that system in a house or anything except a very small room, buy the cheap alarm cable which is twisted pair uniformly without the wound up conductors.

 

Making twisted pair as described produces a poor version because each wire is also twisted.

What's a "twisted" wire and why is it a problem?

buy the cheap alarm cable which is twisted pair uniformly without the wound up conductors.

Good idea if you are running more that a few feet of twisted wire.

 

Twisted wire is what you get making a twisted pair by spinning wires around with a drill motor. Not only do the wires wrap around each other, but also each wire twits about it's own central axis.that is undesireable in a cable of any length beyond a few inches.

 

each wire twits about it's own central axis.that is undesireable in a cable of any length beyond a few inches.

Again, why?
Do you know for a fact that they do the wire twist commercially with twisting the wire?

 

Again, why?
Do you know for a fact that they do the wire twist commercially with twisting the wire?

.
Yes, I know for a fact .........
You need to visit a Wire manufacturing plant.
It will be immediately obvious how "un-twisting" each individual Wire,
when making a multi-conductor Cable,
is automatically avoided in the process.

90-Second Video ..........
.

 

OK, LQC has shown a more current picture from a manufacturing system, which operates in the same manner. So the fact is that we both know what we are talking about.
The drill method may be OK for short lengths of under a foot or so, and if the ends of the wires are allowed to rotate rather than be clamped in a vise. It gets repeated by those who have not used it, is what I see.

 

90-Second Video ..........

Nice music.

ak

 

And YES, I know for a fact that assuring no additional twisting of individual strands is a standard part of manufacturing cables.