Analog Optocouplers Beginner Questions

[pop]

Hi,

I want to connect an analog Optocoupler (e.g. a phototransistor or a photodarlington), to isolate a high voltage AC from my TTL circuit. I want this optocoupler to give me a proportional output signal, 0-5V (active mode), as variable high voltage is applied to optocoupler's LED.

I understand how a 4-pin optocoupler (2 pins for Diode, and 2 pins for transistor's collector and emitter, no base pin) works.

But what do you do about those 6-pin optocouplers (e.g. 4n32, 4n25) where you have a base pin?

What is the purpose of the base pin in a phototransistor, considering the input signal is provided by the LED?

It's been a while since I worked with transistors in active mode, sou could you tell me which configuration would be best for my application? (e.g. common emitter; common collector; are +ve,-ve power supplies needed when attached to collector and emitter respectively?; etc.)


Any information is welcomed.

Thanks

 

[Styx]

analogue opto's are not that good (ence why Airbus disallow them in aircraft use).

However, I recently asked a similar question abt the 6th pin while using such an opto in discrete use. It was recommended to tie that pin to emitter-GND via a 100k resistor.

This worked very well with propper diode driver.

For linear use, I would recomend using that pin to bias the photo-transistor into a quessent conduction poitn (you will have to look at the datasheet to calculate such a point)

you will have to ensure that the diode side is driven with enough current (~40mA) for full demand

 

[pop]

Thanks Styx

I'll give it a try!

 

[Sebi]

Also existiert the real analog opto with monitor diode...

 

[Styx]

Sebi, that is alot better cct then using a plane old opto-coupler setup

The worst thing abt using opto-couplers in their linear region to isolate an analogue signal is over time as they age their linearity severly deteriorates.

With the method there is a feedback of such a deterioration. However, the downside of this is, as the opto starts to deteriorate (and it will after not that long), this closed-loop will over-drive to compensate, and thus age it faster.

Admitidly it will last alot longer (as in with usable characteristics, plus general use) than no-feedback.

That leaves only two other methods for anaolgue isolation.
1) AM over an isolation XFMR
2) digitise it and serial it over an opto (in its descrete mode, where it lasts alot longer, perfromance wise)

with 2) it is worth looking at magnetic isolation chips IL71# range
https://www.nve.com/webstore/catalog/product_info.php?products_id=58

extreamly high datarates with isolation on par with optos, thus you can still get the resolution. One warning, be careful with the power supplies to these chips and also fields - I tried these chips (out of interest) in a gatboard to switch 800A at 350V and they kept being influence by the fields.

 

[Dean Huster]

Clairex used to make a line of analog optocouplers that used LDRs as the "receiver". I believe they were bought by Perkin-Elmer and I don't know if the product line was continued. LDRs as the output device can allow you to do some interesting things with voltage dividers that are sometimes more difficult using a transistor.

Dean

 

[jbeng]

EG&G Vactec also makes (made?) a line of LDR/LED based optocouplers they called a vactrol. A Google search of "vactrol" came back with several hits ...
JB

 

[pop]

Thanks for the responses.

I did not mention that I will be isolating a current transformer (with a load resistor and tied into optocoupler input) and a TTL circuit (that measures CT's secondary current). With that said I wouldn't like to have any TTL on the high voltage side.

Is feedback required on the optocoupler's transistor amplifier? In general, when would you have a feedback on a transistor amplifier?

Thanks

 

[pop]

I've just about had it, with trying to get the phototransistor to operate in active region. I have tried biasing the base with 100k ohm resistor but, the 4N26 still remains in saturation region. I have also tried to make my base voltage smaller than my collector voltage (common-emitter config) but I was still stuck in saturation region.

Does anybody have a schematic or any suggestion of how to operate 4N26 or similar phototransistors in the linear (active) region?

Is there any other way (other than analog optocoupler) that I can isolate my current transformer completely from my TTL circuits?

Links about analog optocouplers are also welcome as I have found that majority of sites contain information about on/off aplications of phototransistors.

Thanks.

 

[Styx]

biasing with 100kOhms?
Using 100Kohms to GND is good for saturation mode, for use in the active you need to use the base pin as a voltage divider, to "bias" the output photo-transistor

 

[pop]

Yeah, I did try to bias using a voltage divider such that Vc>Vb. Also I used a negative voltage on a base once I got frustrated with saturation mode.

I heard that not all phototransistors are made to operate in active mode. Maybe 4N26 is like that. Certainly, its datasheet never mentions the active mode, but the application notes AN3006/04 mention the possibility of it.

Anyway, it was suggested to me to try to isolate the current transformer from TTL by using another current transformer on the first CTs secondary.

Thanks for the responses.