Hello!
I am using two optocouplers(BL817) to turn on and off a TC4427 mosfet driver. The driver's input capacitance is 12pF. I am using one optocoupler to connect the input to 15v, and the other one to connect to GND.
Blue is input signal (50kHz PWM)
Red is mosfet driver input
Yellow is mosfet driver output
As you can see the output is quite different to the input signal.
Actually I am trying to switch a N channel mosfet on the high side at 50kHz, using a DC-DC isolated converter, and I just don't know how to do it correctly. I am powering the mosfet driver with the DC-DC, and using the optocouplers to isolate the driver input from the rest of the circuit.
Doesnt matter what drives my opto's. The mosfet drivers driving the opto's is only because my mcu can only output 6mA. I can use a wavegen and it doesn't make a difference. The signal driving opto's is 1.2V, beautiful and same-looking as the input PWM signal.
Capacitors are missing...
Also, if you drive both optocouplers at the same time, you will fry them...
If both go through transient at the same time, you may also fry them...
U3 serves no function in this schematic. The optos are digital, not analog. You got a clean switch to a soggy input, what is problem? Yellow trace is following blue cleanly. What were you expecting?
Yes, if both opto's are on in circuit shown a short will be placed across the power rails.
You've come here with a problem, and have been offered advice. There is clearly a fault with the fragment of the circuit you have shown.
There is no need for U1, in fact if both U1 and U2 are active at the same time there will be a short between VDD and GND.
If you are interested in a much simpler circuit look at using something like Si875x: The version with the logic input (Si8751) could be driven directly by your 6mA mcu pin with no mosfet level translation required. The other with diode emulation (Si8752) requires around 20mA of drive current but has a simpler circuit.
There is your major problem. 1.2 V is not sufficient to drive the optocoupler. Inside it has a LED. This LED will not give any light below 2V (depending on LED type).
It seems it does matter what drives the optos..
What you see in the graphs is a very tiny current through the optocouplers charging and discharging the capacitance of the mosfet on the input side of your mosfet driver.
Don't think so.
The LED inside an opto couler typically drops 1.2volt, because it's infrared.
I assume OP has a 3.3volt processor, and a series resistor that limits LED current to about 5mA.
That resistor, and it's value, should have been on the diagram. 330ohm would be a good value.
Note that common opto couplers are saturated switched, and therefore slow.
I still remember we had to use special/fast optos for 31.25kHz MIDI I worked with many moons ago.
Leo..
I might have misread op's message. I though he was saying that his MCU outputs only 1.2 V.
From that I concluded that he did not have a 3.3V device, but a device running at less than 3.3V.
Did not know that IR LEDS need only that low voltage (but makes sense, as it goes quite steep up from red to blue).
Doesn't really matter as the optocouplers are not a good idea anyway...
Not in this configuration for sure...
And OP said he used 1k, which is too high according to you...
And you say it may be too slow...
