Compensate temperature drift in optocoupler

[flote21]

Hello guys!

I am have a big trouble with 4N25 optocoupler regarding to the temperature. See my schematic:

Red crosses means not assembly components.

When I switch the power supply and I measure voltage with the multimeter in TP37, the measurement is 5.3V. However as far as I increase the temp amb, the voltage starts to decrease until 4.3V!!! It is like a discharge of a cap effect.

Any idea to compensate this temp error?

Thanks in advance!

 

[barry]

What's driving your LED? What is your temperature, 10000 degrees? How do you know it's the optocoupler that's causing the problem and not the load resistor or the LED drive?

You could probably put some temperature compensation circuit on the base of the optocoupler.

 

[dick_freebird]

I'd back up a bit and decide whether this "error"
affects the outcome where it counts. Everything
drifts; how much, and does it matter?

Presumably this is part of a larger assembly
and maybe the outer loop is driving the opto
output voltage down, for its own reasons
(like going out of regulation elsewhere).

Your comparator ought to be flickering about
5.00V and neither of your voltages (initial,
or after temp & time) are correct.

 

[flote21]

Hi again

I think that the problem is not on the LED side of the opto, because i have been monitoring the current of the Resistor LED and is not changing its value during the complete temp test.

The temp test was made from 25ºC to 85ºC and the voltage in TP37 was decresing until reach the VREF. then it stops and it the voltage in V- of the comparator is not lower than Vref.

Can you give me some advise about a temp compsation circuit to implement in a easy way according to the previous schematic?

thanks in advance.

Regards

 

[dick_freebird]

Why do you think that this is a problem
in the first place?

Why do you think that the problem (if any)
is with the optocoupler and not some other
influence?

From the description (which is inconsistent
between posts) it now sounds like the larger
assembly transitions between open loop
and closed loop ("decreasing, then stops
at V- = Vref"). V- should equal Vref always
if the assembly is closed loop.

"It is like a finger pointing away to the moon"
(student with blank look; slap upside the head)
"Do not concentrate on the finger, or you
will miss all the heavenly glory"
- Bruce Lee, "Enter the Dragon"

That is to say, broaden the scope to include
what is driving the LED, how and why, and
maybe you will see the actual -cause- and
then determine -what- needs compensation
(or perhaps, none as it's a different problem
entirely).

 

[KlausST]

Hi,

Do you want it ON/OFF switching or linear?

When switching, then increase R148 or reduce VRef.
Switching always is with a lot of overdrive.

Klaus

 

[flote21]

Hi!
I want to work in a linear mode.
Thanks

 

[FvM]

I want to work in a linear mode.

The part isn't designed as linear optocoupler. The observed temperature drift is just expectable.

 

[flote21]

mmmm.

And can you give a recommendation of a part which can work in a linear mode?

Maybe there are some commercial optocouplers which a embedded compensation circuit...

Greetings

 

[barry]

If you look at figure 6 of the Vishay 4n25 data sheet you'll see that the collector current varies from 15ma to 10ma over a temperature range of 25 to 85 degrees. That's a 33% change. You can't just run this open loop; as others have mentioned, you need to look at this from a complete circuit point-of-view and compensate it appropriately.

 

[FvM]

High linearity optocoupler are providing two photo diodes with one LED, e.g. HCNR 200.

 

[KlausST]

And can you give a recommendation of a part which can work in a linear mode?

I tried an internet search. "Optocoupler l " automatically added "inear".

Many good results. Parts, selection guides, datasheets...

Why don't you try this?

Klaus