Galvanic isolator using optocouplers

Thread Starter

Dars1988

Joined May 19, 2022
5
Hi everyone,

I have a very simple circuit I need for automation purposes but my knowledge with transistors are very far in the past. I want to galvanically isolate a 0-24 VDC square wave signal using the following optocoupler : https://www.phoenixcontact.com/en-us/products/single-solid-state-relay-opt-24dc-24dc-2-2966595. Otherwise, I want a copy of my input on my output. My input maximum frequency is around 3 kHz, no big deal.

I searched and found out I can't use this optocoupler alone and I tried the following schematic which does not give me a square wave. I put a random resistor of 47k :

Schematic.png
That does not output a square wave. Suggestions??

Thank you!
 

Thread Starter

Dars1988

Joined May 19, 2022
5
From the SSR technical data.

Transmission frequency300 Hz
I noticed that but I also noticed

Typical response time20 µs (at UN)
Typical turn-off time300 µs (at UN)

If turn-off time is 300 us, should I be able to sustain 3.333 kHz? I'm not sure what transmission frequency means in this case.
 

nsaspook

Joined Aug 27, 2009
9,733
I noticed that but I also noticed

Typical response time20 µs (at UN)
Typical turn-off time300 µs (at UN)

If turn-off time is 300 us, should I be able to sustain 3.333 kHz? I'm not sure what transmission frequency means in this case.
Think about the harmonic content of a clean 3kHz square wave. If your switching time is 300us you're low-passing to the sine-wave fundamental of a square wave 3kHz signal. At 300Hz the switching speed allows for most of the harmonic content to remain intact for a much cleaner reproduction of the original transmission frequency. You need a faster opto-isolator.
 

michael8

Joined Jan 11, 2015
287
An optocoupler and a solid state relay are two different things. The solid state relay likely contains an optocoupler
plus additional power switching (FET?) device(s).

Also the 47K "random resistor" is likely not allowing enough current to flow to maintain the speed the device is capable of.
I'd aim for at least 5 mA so 24 V / 5mA -> more like 4.7K.
 

nsaspook

Joined Aug 27, 2009
9,733
You will need current limiting resistor on the LED side. (input)
Not really. The technical data says typical input currents at 24v is 7mA. It's internally limited like a typical SSR designed for 24V operation. The root problem here is the device and his frequency of operation, not the circuits driving it.
 

Thread Starter

Dars1988

Joined May 19, 2022
5
Ok, I understand and I found out why I didn't have a square wave. Oscilloscope problem. I indeed see a decay of the square wave at 300 Hz and more. I'll look for something better. Thanks.
 

Papabravo

Joined Feb 24, 2006
18,426
This shows the effect of a low pass filter on a square wave input. To avoid distortion, I would say that the corner frequency of the lowwpass characteristic needs to be about 10 times the fundamental frequency of the squarewave, which will allow for harmonics 3, 5, 7, and 9 to pass through.

1652979232142.png
 

crutschow

Joined Mar 14, 2008
29,799
Below is the LTspice simulation of the circuit using a faster 4N25 coupler:
I connected the output resistor to the emitter so the output is in phase-with the input.
For reverse polarity, connect the resistor in the collector, as your circuit shows.
(R3 is just for simulation purposes and would not be in the real circuit).

Will the somewhat slow fall-time of the output be a problem?
If so, there are faster opto couplers available such as the 6N136 (bottom).

1652982233131.png

1652982837303.png
 
Last edited:

AnalogKid

Joined Aug 1, 2013
10,057
What is the source of the 24 V input data stream? Specifically, how much current can it source and/or sink? Depending on the optocoupler chosen, the source might have to supply several mA for reliable operation. If it can sink more than it can source (a common thing), a simple re-arrangement of the opto's input component(s) can use this to an advantage.

ak
 

Thread Starter

Dars1988

Joined May 19, 2022
5
What is the source of the 24 V input data stream? Specifically, how much current can it source and/or sink? Depending on the optocoupler chosen, the source might have to supply several mA for reliable operation. If it can sink more than it can source (a common thing), a simple re-arrangement of the opto's input component(s) can use this to an advantage.

ak
The current sunk and sourced are the same. I want to isolate the input from the output because the input is fed with grounded 24V and I want the output to be floating.
 

Irving

Joined Jan 30, 2016
2,669
Not sure how to interpret the data sheet.
What is the difference between the max supply voltage and the max output voltage (below)?

View attachment 267663
Hmm, on the 2019 Toshiba data sheet for the ones I have in my parts bin both are stated as 15v absolute Max and the device is only specced at 5v..


The other issue is the output drive capability is tiny, certainly no good to drive more than a couple of inches of cable; it needs a low output impedance buffer on the end to be real-world useful..
 

Attachments

AnalogKid

Joined Aug 1, 2013
10,057
The other issue is the output drive capability is tiny, certainly no good to drive more than a couple of inches of cable; it needs a low output impedance buffer on the end to be real-world useful..
On page 7 the output current is characterized to above 8 mA. At 30% CTR, that's more than enough to drive a 1 K load if the source can supply 17 mA.

ak
 

Irving

Joined Jan 30, 2016
2,669
On page 7 the output current is characterized to above 8 mA. At 30% CTR, that's more than enough to drive a 1 K load if the source can supply 17 mA.

ak
True, on the 6N136 though that doesn't go far on a capacitive cable.

My bad however, the quote was intended to refer to the AN25, but editing the post moved it around and lost the reference.
 
Top