LED spikes when transistor shuts off ??

[nandoanalog]

 On the picture, you can see a mosfet gate probe (yellow), and another probe on an IR transistor (blue).
 The mosfet is on the the low side of the ir leds. Why do the leds brighten up so much when the mosfet turns off? Some weird capacitance thing?

[Dave]

Draw a schematic and show exactly how your probes are connected.

[alsetalokin4017]

That must be one heck of a  mosfet, that you are supplying with 100V gate pulses.

[danadak]

Looks like the Probe x setting in scope on x 1.

Regards, Dana.

[bktemp]

My guess is your power supply has a slow regulation and overshoots whenever the load is switched off or there is a voltage spike due to the wiring inductance.

[bktemp]

Did you try using a seperate power supply for transmitter and receiver with the only connection between both beeing ground from the scope?

[bktemp]

That confirms my guess:
There is a voltage spike generated when the load is switched off. If the receiver is connected to the same supply, the spike propagates into the output signal.
Adding a capacitor (100nF + 100-1000uF depending on LED current) between +12V and GND should remove the spike.

[alsetalokin4017]

Now let's work on using the scope better.

1. Assure that both the Probes (switch on the probe body) and the Scope Channel input attenuation ("Probe" submenu) are set to the 10x position, so you get realistic voltage values on your readout. I really don't think you are providing 100 volt pulses to your mosfet Gate. Adjust the Channel Vertical Scale settings appropriately.
99 percent of scope measurements will properly be made with the probe and channel set to the 10x position. Using 1x is appropriate only in certain situations like very low signal levels, relatively low frequencies, direct cable connection to a signal source without using the probe,  &etc. Almost always we want the probe body switch and the Channel "Probe" settings to match. In the present case using 10x on both the probe and the channel is appropriate.

2. Adjust Trigger level so that you are getting a stable set of traces rather than the jittery things in your photos. This may involve setting the trigger to the very top of the CH1 trace. Alternatively you can try a Single Shot capture, where the scope will automatically stop after a single horizontal scan.

3. Use the built-in screen-saving process to make your screenshots rather than photographing the screen. I see that you have a USB storage device inserted in the front-panel USB port already, so simply press the "printer" button underneath the Help key at the top right corner of the scope's panel to save the screenshot to the USB storage device. The scope defaults to the .png format but you can select different formats in the Storage menu should you desire, but .png is fine. To speed up saving screenshots to USB storage you may want to Stop the scope with the Run/Stop button before saving. Then you can simply post the image file from the USB stick without having to go through the photography process, and you'll get a much better image. With perhaps 1/8th or 1/10th the file size in bytes... and no dust!

[alsetalokin4017]

Like so?

Beautiful ! Now to get more cycles of the waveform on the screen you can move the trigger point to the left using the Horizontal Position control. 
A slight further refinement, to make the vertical values easier to read by eye, would be to use the Channel Vertical Position control to put the channel baselines right exactly on a graticule division line. The pointy markers on the left side of the trace area tell you where the baselines are located; putting them directly on a line simplifies reading the voltage values of significant parts of the waveform.