Optocoupler as current measure sensor for uA currents

[avihaio]

Hi,
I'd like to measure currents of ~20uA-2mA using an optocoupler.
The problem is that I never find how the optocoupler works under low uA forward currents.
I need it for an isolated measure of the current in the secondary side of a flyback where the only voltage I have is the output voltage of the converter
I'm looking for suggestion for a linear optocoupler who can do this and how to know from the datasheet if it can measure these currents.
Also suggestions to alternatives can be good.

Thanks

 

[betwixt]

The temperature stability and long term transfer ratio of opto-couplers doesn't really make them suitable for that current range but you might be able to linearize any common type by adding a fixed current to the one you want to measure. For example, add 2mA so the range becomes 2.02mA to 4mA.

Brian.

 

[KlausST]

Hi,

the only voltage I have is the output voltage of the converter

Not very useful information. Can you give any value?

There are optocouplers for linear operation, some of them have a monitoring diode for feedback. They are way more precise than usual optocouplers.

If you could transform the current into a dutycycle (triangle and comparator), than you may use a cheap optocoupler and get very good precision and stability.

Klaus

 

[crutschow]

Why do you need to use an optocoupler?

 

[avihaio]

Hi,


Not very useful information. Can you give any value?

There are optocouplers for linear operation, some of them have a monitoring diode for feedback. They are way more precise than usual optocouplers.

If you could transform the current into a dutycycle (triangle and comparator), than you may use a cheap optocoupler and get very good precision and stability.

Klaus

About 600V - it's a one-shot flyback capacitor charger, so after charging is finished it starts to discharge and I'd like to monitor it on-board in an isolated method and send by some interface (I2C, SPI...).
The problem is that I don't have any stable DC voltage in the high-voltage-side and it makes my system much bigger to bring such a voltage to this side (3.5kV isolation is a requirement).
So triangle and comparator might be problematic without supply.
Linear optocouplers - I didn't find yet one with a graph that shows what happens to the output current under 100uA.

Why do you need to use an optocoupler?

Optocoupler is an example. I'm open to suggestions.

 

[betwixt]

Possibilities:
1. As it's a dynamic current, can you use a transformer for isolation? You do not quote a minimum voltage, amount of charge or rate of charge/discharge.
2. If you are talking about high currents, maybe a Hall effect sensor can detect the magnetic field in the discharge wiring.
3. If the capacitor has a fixed value, the current might be calculated from the rate of voltage change across it.

We really need even more detail on the scenario to find the most appropriate solution.

Brian.

 

[d123]

Hi,

I think galvanic isolation may be a solution, they cater to high voltage on one side and I2C, etc. on the other.

The Mouser link is to see prices, some are "only" €5 to €6.

Galvanic Isolation Amplifiers

You could get an overview as to whether this is really a viable option at Isolation – Technical documents and if it looks appropriate, have a read of this manufacturer's products:
Isolation Solutions: High voltage isolation for robust and reliable system operation.

 

[avihaio]

Possibilities:
1. As it's a dynamic current, can you use a transformer for isolation? You do not quote a minimum voltage, amount of charge or rate of charge/discharge.
2. If you are talking about high currents, maybe a Hall effect sensor can detect the magnetic field in the discharge wiring.
3. If the capacitor has a fixed value, the current might be calculated from the rate of voltage change across it.

We really need even more detail on the scenario to find the most appropriate solution.

Brian.

Thanks Brian.
1. The capacitor charges to the target voltage in ~20-30ms, so I think that a transformer might not do the job. The capacitor discharges to 0V in a slow rate (> 1sec, I can configure it).
2. After some simulations, I seethat the current range can be increased to be 80uA-4mA. Higher maximal current will harm the capacitor charging rate and maximal voltage, due to current limit.
3. The rate of voltage change is not good for my needs BUT, if you have a suggestion how to measure the voltage on the capacitor in an isolated way, without a stable Vdd, it can be very helpful and may solve also my current sensing need.

 

[FvM]

I didn't find yet one with a graph that shows what happens to the output current under 100uA.

Does it mean that there's no output? Unlikely. You may want to make some measurements on your own.

3.5kV isolation is a requirement

Test or working voltage? If the latter, the choice of suitable opto couplers is rather small. There's probably no off-the-shelve type with photo diode receiver.

I'm however missing some essential specifications, particularly relative/absolute current accuracy, bandwidth and acceptable primary voltage drop.

 

[avihaio]

Hi,

I think galvanic isolation may be a solution, they cater to high voltage on one side and I2C, etc. on the other.

The Mouser link is to see prices, some are "only" €5 to €6.

Galvanic Isolation Amplifiers

You could get an overview as to whether this is really a viable option at Isolation – Technical documents and if it looks appropriate, have a read of this manufacturer's products:
Isolation Solutions: High voltage isolation for robust and reliable system operation.

Correct. The only problem is that I can't provide this 4.5-5.5V supply voltage Vdd1 in a stable way, so "passive", or only secondary side supply components are maybe better for me.

Thanks