Hi,

I've been building a test/proto perfboard for game buttons. Concept is simple: 4 'action' buttons, 1 reset button, a buzzer and some indicator LEDs.
The original design is (src:elonics.org):


The modifications I made is replacing the LED+resistor on the 555 output with an IRLZ44N connected with gate to the 555 timer output to turn on a piece of 24V LED strip, and made it a 4 button circuit.
Reason for this is that the circuit is powered by 12V (VCC in next drawing) and the LED trip runs on 24V.
So the 'new' design is:


So I put everything together:


So, left photo: top bar and middle 'bar' are negative.
Right photo: bottom line is VCC (12V). Brown is negative rail, blue is reset rail and green is the status rail.
Soldering wise it was easier to connect all 24V's going to the buttons in a cluster near the first (right one, on the 2nd photo) 555 timer, so I did.

This all works though due to some weird issue with the first button (right on 2nd photo) that when one other button is pressed, eventually the led strip connected to the first (right) 555 timer suddenly started to turn on slowly. Even when the 555 timer was removed.

While trying to find out the cause, I'd read somewhere that between Gate and Source, a MOSFET acts as a capacitor. This in my opinion explains the slowly powering on of the MOSFET/ledstrip. So I placed a 10k resistor between the trigger pin and ground.
This helped as the first button doesn't turn slow on anymore when another button is pressed. BTW: only the first 555 timer MOSFET slowly turns on, the others just work as expected. I can't measure the voltage rise on the gate because as soon as I either connect my multimeter or oscilloscope, it turns off. Which also made me think of adding the pull-down resistor.

So my question is why is this happening? Why is that single MOSFET turning on by itself?
Is it because I placed all 24V+ lines near it? And is the first MOSFET 'charging' due to inductance? Should I add pull down resistors to the other MOSFET gates as well?

I did triple check all traces/connections and there are no bridges etc that shouldn't be there.

any thoughts are welcome.

[edit]
Edit's to complete initial post based on additional questions in comments.
* IC has been swapped, 'issue' stays with the same 'position'. To add to this, I've also replaced the FET and the diode.
* with the connected pull-down the led strip does turn on when the button is pressed.

KR,
Onno.

 

Does that first LED strip come on correctly when its button is pressed?

 

Swap the ICs around and see if the fault follows the IC. Removing the IC completely won't tell you much as the associated FET gate is then left floating (assuming the status diode is unconnected) and so the FET could do pretty much anything, so just swap the ICs and see what happens. You could have a 555 where one of the output transistors is failing.

 

Does that first LED strip come on correctly when its button is pressed?

Yes, with the pull-down resistor everything works as expected.

Swap the ICs around and see if the fault follows the IC. Removing the IC completely won't tell you much as the associated FET gate is then left floating (assuming the status diode is unconnected) and so the FET could do pretty much anything, so just swap the ICs and see what happens. You could have a 555 where one of the output transistors is failing.

The IC was replaced and the issue stayed. The IC also did move around and the issue stayed at the first position.
When the IC was removed, the status diode was left connected.

 

With everything in place as it should be, check the resistance between the MOSFET gate pin and the '555 output pin. It should read less than 1 ohm.

 

Double-check the pin-out of the FET and ensure the drain and source connections haven't been transposed.

 

Sorry for the silence, got caught up in some work. I'll triple check the pinout and measure the resistance as soon as I've the setup back. Some people were really eager to test/use it. haha.

They did however report back an interesting bug though. While they were using it for more than an hour, they noticed that at least one of the other LED strips also started to turn on while no button was pressed. So same behavior as the first one, but this just took longer. Feels like some induction induced charging is going on .. Though I'll recheck everything as soon as I've the set returned.

 

Casual look-through...the maximum voltage on the gate of the FETs is +/-10V. Vcc=12V...? Voltage dividers stepping that down to around 5V may do it, and will also make sure the gate is positively either ground or +5V.



Paul
KI5VNH

 

The whole circuit is very susceptible to noise.
Because your power supply (wall wart) is not earth grounded, there will be a lot of common-mode noise on the whole circuit. It would be best to have a grounded power supply.
Because your pushbuttons have long cables on them, they are a good antenna. Did you try put a cellphone near them?
It would be customary to have a small capacitor about 10-33nF from pin 2 to GND, and on the STATUS_RAIL to GND.
NE555 must have local decoupling capacitors like 0.1uF because the IC output stage is known for making spikes due to shoot-through when switching.

 

the maximum voltage on the gate of the FETs is +/-10V. Vcc=12V...?

Good call. The 555 output voltage driving the MOSFET gate will be close to or above the maximum rated gate voltage.
Gate breakdown could account for the LEDs turning on unexpectedly.

 

The problem that stays with the position is definitely due to some leakage current raising the gate voltage. That is why connecting the meter switches it back off. The leakage might be due to a bit of solder flux between the gate connection and some positive voltage, or it might be caused by internal leakage, which might be within the specification of the device. So definitely the pull-down resistor is the correct fix.