Soft latch on-off circuit glitches

[Krotow]

In search for soft on-off solution found Latch and Toggle Power Circuits with as it look - simple and useful soft latch circuits. Before taking one of them decided to test them on breadboard. Also because device where I want to use them works from 9V. Due to a bit limited component supply @home at this afternoon tested them with different MOSFETs. Turned out that

1) Circuits with 2 logic inverter gates (Press ON - Press OFF soft latching circuits) when made as it is in article, with CD4011BE NAND gates (have no other "inverters" here now), switch on and off erratically in all voltage range up to desired 9V. As turned out, 1 uF capacitor charge too fast and gates become flipped before button is released. Capacitor must have higher value: at least 2.2 uF, but better 4.7-10 uF. That seems already usable. Extended version with P-channel MOSFET (tested with SPP18P06P) on VCC works too. There is side effect: when button is pressed for 3+ seconds, gates turn into multivibrator and output voltage fluctuate till button is released. Nothing of that was mentioned in description. Would like to know how to avoid multivibrating effect on long button pressing for switching off. Put a diode somewhere? Also I'm curious which logic IC author used - no mentioning about that too. It certainly was none of CD4xxx.

2) Circuit with two complementary MOSFETs (IRF7319) instantly turn on when powered and doesn't turn off on button press. I used SPP18P06P (P-channel) + STP20NE06L (N-channel) MOSFETs for that. Wrong MOSFETs?

[Krotow]

As for second circuit with dual MOSFET... At today got FDS8958A (IRF7319 substitute) that worked as expected. A bit later later in office electronics dumpster box noticed decommisioned laptop board with few Si4542DY (another IRF7319 substitute) on it. Took them out and tested in switch circuit. They also works. Seems the main problem in this circuit indeed was wrong MOSFETs. Noticed erratic switch behavior above 6V (I need 9V). Seems 1 uF capacitor value is too low again. Found 10 uF making switch very stable on 3V and above. Another erratic switch behavior on inductive load (12V 1A motor tested in 3V-12V range prevented turning the circuit off) was solved with 10 kOm resistor connected in parallel to load together with diode connected in series between output and load as described in article chapter Driving capacitive loads. Both Schottky (1N5817) and regular (1N4001) diodes works just fine here. Final circuit version is below.

[jfiresto]

Don Lancaster gives a reliable alternate-action switch on page 317 of his CMOS Cookbook. You can download the book with this link. He uses the circuit to explain clocked logic to teach his readers how to design their own circuits.

[Krotow]

Thanks about the book. Indeed it is useful reading. Plus have pretty detail 4xxx CMOS logic list that I will require for my next DIY project.