[Solved] Why is it that when a PWM signal is applied to the module, full voltage appears at the output?

Hello.
Has anyone perhaps encountered the case that when I feed a PWM signal to a module (or MOSFET), I always get the full voltage at the output regardless of the PWM fill given? The PWM signal is applied correctly - I checked with a multimeter. The voltage and frequencies agree. This happens regardless of the amplitude of 0 - 3.3V or 0 - 5V. To make it funnier, when I constructed my father an aquarium lamp controlled by PWM via current drivers it all worked correctly. I am attaching an overview diagram and would appreciate any hints. .

Mos fet at high PWM frequencies may not be discharged specifically its gate . Simply turning it off may not be enough because of the capacitance of the gate and the leads. Special mos fet drivers short-circuit the gate to ground or even polarise it with a negative voltage when the transistor is switched off. This discharges the gate faster and the transistor switches off more quickly.

sigwa18 wrote:

Mos fet at high PWM frequencies may not be discharged specifically its gate . Simply turning it off may not be enough because of the capacitance of the gate and the leads. Special mos fet drivers short-circuit the gate to ground or even polarise it with a negative voltage when the transistor is switched off. Then they discharge the gate faster and the transistor is switched off faster.

Yes, a possible reason that it does not close, but I doubt that in this case we are dealing with frequency in MHz.
To the author, it would be good to show a wiring diagram including the type of transistor, if you don't have an oscilloscope then instead of a MOSFET connect a bipolar one without a lamp with some resistor in the hundreds of ohms to the base and a similar one in the load and then check with a multimeter if it responds to PWM.
You can also reduce the frequency in the ESP for testing.

Added after 4 [minutes]: .

Aaa, one more thing: see the schematic of your ESP board, if there is no filter on the PWM output in the form of a capacitor with a resistor.

In my case it is a frequency of 1kHz. I use a ready-made module consisting of 2 MOSFETs designed for just such things: .
At first I thought there was a ground problem but I checked everything and it is ok.

Ok. Now it's still a matter of finding the schmat of your NodeMcu. There are several versions of this module, so only you know what you have. If there is a filter built into the PWM output, then that is the cause of your problem.
1kHz is a very baaarry low frequency for the effect mentioned in the second post to have an effect.
Still, do a trivial test that the two mosfets are working. Just give any pin from the controller a signal with f<1Hz, the lamp should blink at that frequency. Check, we'll rule it out at least.

Ok thanks gentlemen for your answers. Everything was generally fine only not knowing why without load the multimeter shows such miracles. When a load is connected to the output everything is fine.