Opto-coupler, slow turn off
I have an opto-coupler circuit where is very slow turn off.
I can't figure out what is the cause of this problem.
It's for UART isolation, with speed 9600 baud.
Datasheet: EL357NC
Do I have some error in my schematic?
Or is actually opto-coupler IC faulty?
This is schematic:
This is input signal (METER_RX_IC), which goes from MCU through 220 Ω:
This is output signal (OPTO_RX_E):
Opto turns on as expected, but the fall-time is very slow.
I tried to remove Q301, no change.
2 answers
You are accessing this answer with a direct link, so it's being shown above all other answers regardless of its score. You can return to the normal view.
Some optos are really slow. What off time is yours specified for?
Your circuit could be better. The transistor is turned on thru 1 kΩ, but off thru 11 kΩ. You are also giving it way more base current than needed, slowing turn-off.
Move R318 to the other side of R317 and adjust both values.
Do some actual calculations. With a 10 kΩ pullup to 3.3 V on the collector of Q301, the transistor needs to pass only 330 µA. 1/10 of that is plenty to keep a 3904 in saturation, so 33 µA into the base when the opto is on.
Let's decide that R318 (which is now directly between base and ground) should be 2 kΩ for reasonably fast turnoff. At 750 mV on the base, that would draw 375 µA. To provide some margin, let's make sure everything works even if the B-E drop is 800 mV. R318 would now draw 400 µA. Plus the current needed to turn on the transistor, 433 µA needs to come thru the opto when it is on.
You didn't provide a link to the opto datasheet, so I'll pick 300 mV as the maximum guaranteed on-voltage of the opto. With 800 mV on the base and the 3.3 V supply, that leaves 2.2 V across R317. (2.2 V)/(433 µA) = 5.08 kΩ, so 4.7 kΩ it is.
So in summary, move R318 to be directly on the base of Q301, change it to 2 kΩ, and change R317 to 4.7 kΩ. I think you'll see a significant decrease in off-time that way.
Again, all this assumes the opto is fast enough to support your desired switch-off time in the first place. Since you didn't provide a link to a datasheet, I'll leave it for you to check.
1 comment thread
I changed value of resistor R318 from 10 kΩ to 1 kΩ, and that alone significantly improved off-time. From 200 μs to around 25 μs. That is much better.
So I looked into datasheet again. I found the answer on page 5, figure 8:
Test circuit:
It's surprising for me, that load resistance has such significant influence on the off-time.
0 comment threads
0 comment threads