New member here

[Otaku]

Hi all,
Just joined, although I've been lurking for a bit. Looks like lots of interesting stuff going on here.
I have a quick question on a SCR/LED array I'm working on. I have four 2N5060 SCRs that are switching four LEDs. Each SCR has a unique 12VDC gate signal going through a 10K resistor. The 12VDC is switched through a relay that is triggered by a 4017 counter. Only one of the SCRs can receive a gate voltage at any time.
Here's the problem - occasionally two adjacent LEDs (1+2, 3+4 etc) will light at the same time while the circuit is running, even if one of the gates does not receive a trigger signal. I know that the 5060 is pretty sensitive so I'm going to try installing 22K gate-to-cathode resistors to see if I can stop the rogue triggers. Would a pull-down resistor from anode to ground be a better solution?
Has anyone else seen this type of behavior from the 5060? I'll be posting a circuit drawing in the morning to give you an idea of what the circuit is doing. All advice appreciated!

 

[DemonicSandwich]

It's likely the fact that you are using SCR's to begin with.
Silicon-controlled rectifier

SCRs will begin to conduct when they receive the gate signal but will continue to conduct (even without the signal) until the current through it falls below it's Holding Current.

In the case of the 2N5060, the Holding Current is 5mA at 25*C.
Why not just remove the SCRs and use the relays to switch the LEDs instead?

 

[Gary B]

An SCR can also be turned on by what is called rate triggering. If the voltage across the SCR K/A rises too quickly it can turn on without a trigger voltage on the gate. If your gate driver circuit is a little leaky or the layout allows for inductive or capacitive coupling between lines, this will also cause false triggering. Since it is always adjacent SCRs that fail, I suspect the latter is the cause. If you are using an etched circuit board, try spacing the gate tracks farther apart and place a ground track between gate tracks where possible. Also look at component layout to maximize the distance between SCRs.

 

[Otaku]

Interesting - I have all of the gate signal wires running right next to each other (30 ga. stranded wire). The SCRs are mounted on a thru-hole perf board with .200" separation. I'll try the resistor idea, but it looks like better physical separation may be the answer. Thanks!
Regarding the reason for using the SCRs, the LEDs show the result of a series of continuity tests. I need to latch the LEDs on until the series is finished and the user resets the system.

 

[Otaku]

The gate-to-cathode resistors (10K each SCR) fixed the problem. Apparently there is some cross-talk between the gate wires.