Help with SCR latching

[jcmpix]

Hi all, I'm new here. I built a simple SCR circuit similar to the one on this page:
http://www.circuitstoday.com/how-to-test-an-scr
The differences are
1. I used an LED instead of a lamp.
2. The resistance at the gate is 1k ohm.
3. I placed a 500 ohm resistor between the +12V terminal and the LED.
Now the circuit triggers properly, when S1 is pressed the LED lights up. The problem is the SCR won't latch. It's a C106 and according to the data sheet here
http://www.ee.ctu.edu.tw/material/data sheet/C106B.pdf
the latching current is only 5 mA. The current through the LED is probably around 16-20 mA (haven't actually measured it, don't have the circuit right now). What could be the problem? Is the gate current too low? I've tried three SCRs with the same results. Thanks!

 

[Resqueline]

It seems you've fulfilled all criteria, except the datasheet says it used a 20mA gate current for the 5mA latch test.
For a 50mA anode current however it says that the neccessary gate current is less than 0.2mA.
I haven't tested these exact limits & relations in practice for myself, but the first thing to try is to increase the gate (&/or anode) current.
Of course also double-check connections/polarities, making sure you connected the SCR the right way around.

 

[shrtrnd]

Stick an incandescent bulb in there, and see if maybe the LED (which IS a diode), is
hanging-up your SCR.

 

[janagyjr]

I was going to make some comment regarding the LED being a diode and seeing if the diode drop and current isn't being taken into account, but even with a 500Ω resistor, that leaves .018A of current which should drive the LED, though if the SCR requires 20mA, you might want to bump that up to 30mA (I don't know where you got the LED from, but superbrightleds.com states in their datasheets that their LEDs have a peak continuous forward current of 30mA).

 

[jcmpix]

What effect could the "diodeness" of the LED be having on the circuit? I'll try the bulb, and also try changing the gate resistor to 330 ohms.

 

[janagyjr]

The LED has a voltage drop that you have to subtract from the supply voltage when you are figuring for the resistor value for a specific current. Both the LED and the SCR will have a diode drop. The LED's drop is nominally 2V (though I've measured 1.8V in practice). The SCR in question (for which you provided the data sheet, see page 3 of the PDF for electrical characteristics) has one of 2.2V. Add that to the LED drop and subtract that from your supply voltage and then use that to figure for your current (since your latching current and holding current is way smaller than the LEDs, you shouldn't have any problem figuring for 20mA to drive the LED and the SCRs maximums are high enough that you won't damage it, either).

This is all supposition on my part (based on quite a bit of recent experience with diodes in general and LEDs in particular), though, as I do not have an SCR to monkey around with.

 

[jcmpix]

Thanks, that's what I figured too. The 500 ohm resistor gives 16 mA which should be enough, but apparently isn't. In addition to the above-mentioned suggestions I'll try dropping that to 200-300 ohms.

 

[janagyjr]

That should be enough to drive the diode. The max is 20-30mA, that's not a minimum.

 

[jcmpix]

Sorry for the very late reply. I finally got the circuit to latch. The original "SCR"s I used were in a TO-92 casing. I replaced them with TO-220s and the circuit worked, even with the old resistance values. Seems like the first "SCR"s weren't SCRs but BJTs.