LEDs go solid instead of flashing

Other Hardware General Electronics
1

It will come as no surprise to y'all that my recent success with this circuit has revealed another deficiency. For those who have not followed some of my other threads, the short version is that I am building an LED distribution amplifier, to take the output of a "sound and light machine" and allow multiple people to use it. You can see the current state of the circuit in this video: Sound and light machine distribution amplifier - on the perf-board - YouTube

The issue that I am now trying to solve is that when the frequency of the source signal gets to a certain point, the LEDs light solid instead of appearing to flash. If I plug the glasses that came with the machine into the machine itself, I can see that they are flashing. If I plug the machine into my circuit and use my home-built glasses, they light up solid.

I'm not sure the exact frequency at which they light up solid. The machine has a display that shows a number that increases as the flashing frequency goes up. By comparing the lower-frequency flashing to a metronome, I have tentatively concluded that the flashing frequency is 1/15th the number shown on the display. The LEDs go solid once the display shows 240, which would suggest a flashing frequency of around 16 Hz.

One troubleshooting step would be to swap the my home-built glasses with the machine's original glasses. This isn't possible for two reasons. First, the machine's glasses use the barrel for positive and the tip/ring for negative, whereas I have wired mine using the conventional wiring (barrel for negative, tip/ring for positive). Second, the machine's glasses work at about 4v and don't seem to have a resistor, so maybe the machine has a simple CC driver. My glasses have resistors and are designed to run at 12v.

I have tried taking the PWM circuit out and driving the glasses directly off of 12v. This didn't fix the problem.

I have tried reducing the value of the pull-down resistor on the FET's gate. This didn't fix the problem.

As usual, thanks for any suggestions you may have.

circuit.png
circuit.png1099×859 8.92 KB

2

You appear to have an opto-triac in the circuit. Why?

3

MarkT:
You appear to have an opto-triac in the circuit. Why?

I'm sorry. I couldn't find my exact optocoupler, so I just grabbed one from DipTrace with a four-pin pattern. I've updated the circuit to use what I believe is a correct symbol.

4

I've continued to research this on my own and found a helpful thread on electronics stackexchange. Two things that I will try once my son goes to bed are: reducing the value of the gate resistor and adding a diode to bypass the gate resistor on turn-off.

EDIT: Installed diode. No change. Reduced gate resistor to 10 ohms. No change.

5

Okay, problem solved. It was unrelated to my circuit. :slight_smile:

6

Don't leave us in suspense! What was the problem?

7

MarkT:
Don't leave us in suspense! What was the problem?

Sorry for the delay in responding. The problem was this: I had observed that the left and right channel of the glasses appeared to flash in sync with each other. Therefore, I concluded that they probably shared a common driver, and so I bonded the tip and ring of the TRS jack together. There was little justification for this; I just hate to have copper in the cable that is going to waste, so I figured, why not use both lines to carry the signal instead of just one.

After hours of troubleshooting, I finally thought to pull out the junmper that was bonding the channels, and the problem immediately went away. Upon further investigation, it turns out that the left and right channels are actually distinct, and they are flashing slightly out of sync with each other. To the naked eye, it often appears as if they are flashing in sync, but they sometimes aren't. My best guess is that when the frequency of pulses got high enough, the fact that both left and right pulses were being combined causes the LEDs to see enough current to light up solid.