DIY soft-latch single power button - need help finishing it

[Vanarian]

Soooo guys I'm trying to make a compact soft-latch power button(single momentary push button type).

Required features are :

- Circuit should always start in OFF position (auto-reset or power on reset) ;
- Should work as quick press to power ON, hold 2 seconds to shut down ;
- Must be compatible up to 60V ;
- Must be able carry the battery's nominal voltage to the load it switches ON and OFF ;
- Must be as compact and affordable as possible without compromising reliability.

To be precise about the high voltage switch side, I'm going to use it to switch a Ben Vedder's / Fetcher anti-spark ON and OFF, it comes with a 12v zener diode on the excitation pins and needs at least 20V to go HIGH (else it stays LOW).

I posted on stackexchange about it ; the site is a bit unforgiving when you lack accuracy in description -which might be my case : https://electronics.stackexchange.com/questions/360896/soft-latch-auto-reset-off-hold-on-press-button-is-the-schematic-correct

Here is the original topic (started in november) :
https://www.electric-skateboard.builders/t/soft-latch-switch-single-button-on-off-reset-for-vedder-anti-spark-bom/37746

Here is the current schematic :

I circled the BOM as follow (for up to 60V ability) :

• 1x LM317HVT linear voltage regulator + 1x 270ohm & 1x 820ohm resistors (fixed 5V output)

• 1x resistor between 60 & 100 ohm

• 2x 100kOhm resistors

• 1x 2M resistor

• 2x 1uF capacitors

• 1x CD4011 NAND chip (contains up to 4x NAND gates)

• 1x IRF510 mosfet to switch the load in HIGH / LOW

• 2x diodes for polarity protection

• 1 momentary push button (I got the flatest and smallest one available to me, others might prefer a bigger one)

I've gone back and forth on this already, learnt lot of things but my knowledge is still basic and limits my insight.

So far, I got all the components rounded up imho. Now about the circuit itself, eh I hope you might help me check schematics for good to get this thing working !

Do you see anything wrong ? How would you improve it ? Also, do I need to put the caps specifically in one direction ? I tried various layouts on the breadboard, probed different reactions on the circuit but no successful result yet.

 

[amberwolf]

In the circuit diagram, I dont see any power for your NAND gate chip. Is that just left out of the diagram itself?

As I don't know the diagram or requirements for the anti-spark circuits you refer to (or the rest of your system), I can't tell the answer to the following from the diagram you posted here.

Is this switch intended to pass all the power for the motor/controller system? Or is it switching something else that will then carry the load?

If this switch is in series with the load's V+, then the resistor in series with the drain of the FET and the main V+ is going to have to be huge, and heatsinked, as it will have a very high power dissipation (lots of heat) from the current flow thru it to the motor.

If this switch is not in series with the load's V+, then it only has to be big enough to handle whatever current the switched device requires.


The same is true of the FET itself.



Regarding capacitor polarity, if they are polarized (marked iwth a + or - on htem), then you must follow that polarity in the parts of the circuit they are in (whichever way current flows in that circuit. If you have part of a circuit that current flows both ways in, you must use nonpolarized caps.

Otherwise, it doesn't matter.



When you say you get nothing when probing the breadboarded circuit, do you mean literally no voltage anywhere?

If so, then you may have a defective breadboard that isn't bussing power at the input.

Otherwise you need to explain exactly what does not work as you expect. (what you expected at that point, *and* what you actually got).

 

[Vanarian]

Thank you for your input !

In the circuit diagram, I dont see any power for your NAND gate chip. Is that just left out of the diagram itself?

Yes I left it out of the diagram for clarity but I feed a live 5V straight from the regulator to the NAND chip and ground back to the circuit.

The anti-spark model I use is exactly the following : http://www.antisparkheaven.com/product/vedder-fechter-antispark-switch

It's from the "inrush limiter" module by Fetcher, few tweaks from Fabio (antisparkheaven) to make it more reliable. See the original circuit : http://everycircuit.com/circuit/6094204377235456/fechter-vedder-anti-spark

Is this switch intended to pass all the power for the motor/controller system? Or is it switching something else that will then carry the load?

It is intended to switch HIGH/LOW the module linked, the module itself carries all the power.

If this switch is in series with the load's V+, then the resistor in series with the drain of the FET and the main V+ is going to have to be huge, and heatsinked, as it will have a very high power dissipation (lots of heat) from the current flow thru it to the motor.

If this switch is not in series with the load's V+, then it only has to be big enough to handle whatever current the switched device requires.

Ok can you please elaborate more ?

I see the point, I plan to heatsink the whole PCB to cool down the components ; the resistor on the live "battery voltage" wire should reduce current between 0.2A & 1A if possible, that's what I envisioned. So putting some thermal compound to evacuate heat toward the heatsink will be required ?

Regarding capacitor polarity, if they are polarized (marked iwth a + or - on htem), then you must follow that polarity in the parts of the circuit they are in (whichever way current flows in that circuit. If you have part of a circuit that current flows both ways in, you must use nonpolarized caps.

Damn me ! That's probably the problem in my circuit, I do use polarized caps. Didn't even know there were non polarized ones Cause in the base diagram I worked from there is no cap direction shown.

Thanks for pointing it out, gonna order a set. About the breadboard, I do probe voltages at different points across different pins but it fluctuates or empties progressively down to zero (which might be due to caps eating away the current in some circuit areas).

Do you see anything more to check ?