I'm currently working with a sensor that is a open collector output flow sensor. The manufacturer states I need to use a 12V supply and a pull up resistor connected the output of that supply (R1). This generates a roughly -6V to 6V square wave with the frequency changing based on the flow rate. I need to read this data with an Arduino and based on the recommendation of a coworker I'm trying to use an optocoupler with the hope of getting a 0 to 5V square wave out. I've made a crude drawing with the circuit but the issue I'm having is protecting the optocoupler from the negative voltage from the sensor, the optocouplers I've looked at have a reverse voltage of 6V and I believe this means any voltage exceeding this spec will damage the optocoupler. Based on a little research I think others use a reverse biased diode to protect the internal LED of the optocoupler but I haven't been able to find any info on how these parts are selected. Any guidance or info is appreciated.

 

The signal from the open collector output is 0V to +12V.
It would only be ±6V if the signal goes through a coupling capacitor.

 

The signal from the open collector output is 0V to +12V.
It would only be ±6V if the signal goes through a coupling capacitor.

I've attached an oscilloscope capture of the output from the sensor, perhaps I've made a mistake in setting up the oscilloscope? Any ideas on easy things to check?

 

perhaps I've made a mistake in setting up the oscilloscope?

Looks like the input of the scope is set for AC not DC from the appearance of the waveform.
Don't see the need the protection diode or R2.

 

I've attached an oscilloscope capture of the output from the sensor, perhaps I've made a mistake in setting up the oscilloscope? Any ideas on easy things to check?

View attachment 308776

Thank you, the coupling was the issue and actually explains so much of what I saw today.

 

10k ohm with a 12V supply will only provide a little more than 1 mA to the optocoupler. The optocoupler better have a very high CTR. I would lower the resistor's value by at least half.

 

Seems one could eliminate the opto if the 12 volt supply is stable using two resistors. The two supplies share a common ground as seen in post #1.

 

Seems one could eliminate the opto if the 12 volt supply is stable using two resistors. The two supplies share a common ground as seen in post #1.
View attachment 308787

Yes, this is what I'll be doing. I didn't really understand how the open collector output works and the manufacturers info was a little confusing so I didn't think to check the issue with the oscilloscope.

 

I also noted that when the sensor out is LOW the input to Arduino is HIGH with the schematic in post #1. Is that your intent as well or does it matter?

 

I didn't really understand how the open collector output works

Typically it's an NPN transistor with the emitter connected to common and the collector unconnected.
You add a resistor from the collector to a positive supply to complete the circuit.
The output voltage then goes from ground to the positive supply value.

 

Why can't this just be done with a single pullup resistor to +5 Volts?
I have never heard of an open collector output that requires a pullup to any specific voltage?

How many leads does this sensor have? Do have a part number?

 

Why can't this just be done with a single pullup resistor to +5 Volts?

I agree. If the sensor is open collector, just hook a pullup from the output to +5V and you should get a 0 to 5V square wave signal.
You only need an opto if you want electrical isolation.

 

Here are a couple of options.

 

Why can't this just be done with a single pullup resistor to +5 Volts?
I have never heard of an open collector output that requires a pullup to any specific voltage?

How many leads does this sensor have? Do have a part number?

Finding info online about the part was kind of challenging so I had to resort to contacting their sales support team and was specifically told that the pull up needed to be connected to the operable supply range provided in their document with a particular comment saying don't go below 9V. This is the only document I've been able to find/was provided.

 

Should work with 12 volts on the BLUE and the GREEN going to ground.

A simple pullup of 2.2 K ohms to +5V going to WHITE, the signal.
This is the totally bog-standard open-collector solution.

(post 13 shows correct wiring)

 

Use the "Isolated" circuit in post #13 since the arduino's input is 5V. The ardunio already has an internal 22k pull up (if enabled).
If you use an external pull-up, then disable the internal pull-up.

Edit: added comment
In addition, it seems the sensor and circuit may be used in an industrial environment, so I recommend the "isolated" circuit with the opto.

 

I'm currently working with a sensor that is a open collector output flow sensor. The manufacturer states I need to use a 12V supply and a pull up resistor connected the output of that supply (R1). This generates a roughly -6V to 6V square wave with the frequency changing based on the flow rate. I need to read this data with an Arduino and based on the recommendation of a coworker I'm trying to use an optocoupler with the hope of getting a 0 to 5V square wave out. I've made a crude drawing with the circuit but the issue I'm having is protecting the optocoupler from the negative voltage from the sensor, the optocouplers I've looked at have a reverse voltage of 6V and I believe this means any voltage exceeding this spec will damage the optocoupler. Based on a little research I think others use a reverse biased diode to protect the internal LED of the optocoupler but I haven't been able to find any info on how these parts are selected. Any guidance or info is appreciated.

View attachment 308774

Why not replace the R2 with 1 (or 2) diodes in series that allow the +ve level to pass and block the -ve part? Of course, you will not get galvanic isolation, but is it necessary?