Driving LED with Solid State Relay?

[ryanmills]

I seem to be missing something fundamental with solid state relays. I'm trying to use a AQY211EHAX thats rated for 1 amp but I'm shorting out the transistor side trying to drive a tiny LED. Am i just implementing the transistor side wrong? Or am I missing something basic?

[jmelson]

You need current limiting resistors on both the input and output of the SSR. 

Jon

[ryanmills]

Forgive me, I'm feeling like kind of an idiot today. Are the resistors to actually limit current or like a pulldown resistors. Examples that may or may not be correct below.

[TheHolyHorse]

The first image is the correct one, the other one is the same as your first design except it wastes power for no reason.

You want a resistor in series with a LED to limit the current, since the LED is just a diode so it's gonna start conducting if the voltage is higher than the forward voltage, basically just a short circuit and the LED blows. 
The LED should have a forward current(IF) parameter in the datasheet, chose a resistor to achieve that current.

[ryanmills]

Thank you, the wrinkle I have is that the input voltage varies from 1.5 to 3 volts. I picked that SSR because it has a reverse voltage of 5 volts so I had hoped I would not need a resistor on the input. Would something small like 27 Ohm on the input and output be enough?

[Nerull]

The reverse voltage tells you absolutely nothing about the forward current. Reverse voltage is the voltage required to cause breakdown when reverse biased.

If you try reading the datasheet, you will see a absolute maximum input current of 50mA, but you only need at most 3mA to turn it on. You'd also see a nominal forward voltage drop of 1.25V.

82 ohms would be about ideal - enough current at your minimum voltage, and not too much at your maximum.

[TheHolyHorse]

What LED are you using?

NVM

[ryanmills]

I quote the datasheet then hear "If you try reading the datasheet"  I do love EE... Not exactly dealing with FPGA's here but lord help you if you can't grasp everything even in the Beginners Forum.  For the record I'm learning as I go and there is plenty to learn.

In my head I was more worried about voltage than current since the input voltage in my case can vary I can't just use any old current limiting resistor. I had assumed as long as my voltage was within the tolerance of the LED input side it should work as expected. I was not sure on the output but assumed the same thing.

I know this is a stupid question and every expert will tell me to just RTM or go to collage but if someone has a moment I'm still not grasping why I would need an input resistor, I can hook an LED across the pos and neg without any issues, why in this case would I need it if the LED is 1.5 volts and I apply 1.5 volts?

[jmelson]

Thank you, the wrinkle I have is that the input voltage varies from 1.5 to 3 volts. I picked that SSR because it has a reverse voltage of 5 volts so I had hoped I would not need a resistor on the input. Would something small like 27 Ohm on the input and output be enough?

This is a serious problem.  Depending on the LEDs chosen, 1.5 V may not be enough for them to conduct.  Probably the input resistor could be a bit higher, and the output one might need to be lower, depending on the type of LED1.

Jon

[Nerull]

A LED is a diode. It is incapable of regulating its own current, and requires some form of current limiting. An ideal diode becomes a short circuit above its forward voltage. A real diode has some resistance, but this should never be relied upon.



That line is the current going off to infinity and the diode burning up after Vf is reached.

In your first post you are complaining that the diode seems to be shorting out the output. This is why - because you are shorting out the output.

[rcbuck]

Red and yellow LEDs forward voltage drop is around 1.5 to 2.2 volts. Green, blue, and white LEDs forward voltage drop is about 3.3 volts. So 1.5 volts "may" work for the first three without a dropping resistor. However, you need to measure the current flow through the output LED to make sure you are not exceeding its limit.

But 1.5 volts is not enough voltage for the second three. Also, 3 volts applied to the first three would result in too much current that would destroy the output LED. The MOSFET ouput of the optocoupler is rated at 1 amp so you would probably not destroy it.

The AQY211EHAX uses an ultraviolet LED to turn the MOSFET on. The UV LED requires about 3.3 volts to turn on. So a 1.5 voltage source will not be enough for the optocoupler to turn on. In other words the circuit will not work with a 1.5 volt supply.

You can use a 500 ohm pot in series with your DMM in current measurement mode to adjust for the desired current flow at 3 volts. You would do this for the optocoupler LED first. Measure the pot value and replace it with the nearest fixed resistor value. Then do the same thing for the output LED. Once you have the correct resistance values for 3 volts you are good to go.

[ryanmills]

A LED is a diode. It is incapable of regulating its own current, and requires some form of current limiting. An ideal diode becomes a short circuit above its forward voltage. A real diode has some resistance, but this should never be relied upon.

That line is the current going off to infinity and the diode burning up after Vf is reached.

In your first post you are complaining that the diode seems to be shorting out the output. This is why - because you are shorting out the output.

Thank you, learning as I go there are holes in my education. I fully understand diodes but I could never intentionally short one like you would with an LED. It also just realized the cheap amazon LEDS we use in bulk have a current limiting resistor built in so I have never needed them. Just never occurred to me that an LED would need one.

Today "a dim bulb got just a little brighter"

[Zero999]

I answered your question in the other thread. Can you please keep it to one topic per project/question.
https://www.eevblog.com/forum/beginners/do-ssr-require-current-limiting-in-some-fashion/msg2540685/#msg2540685