I’ve noticed that when I connect a APA102 strip to Clock, Data and Ground (so no 5V connection) the LEDs also blink. Not with the signal I send to them, but they are not off. If I connect 5V it works fine.
I’m wondering if other people have the same experience and if this is normal. I would expect that they were of when I don’t add 5V power.
I’ve seen some chips attempt to pull power over data or data/clock line.
There’s always some leakage from the data/clock lines.
In general, it’s not a great idea to let that happen.
It happened because I was testing a Mosfet to switch the 5V off, to prevent current drain in idle mode.
So of course I can stop sending clock and data. Or would it be a better idea to switch the 5V that goes to the level shifter?
It might be better to switch the Output Enable pin on the level shifter instead – you’ll likely get a faster response time from the shifter if it’s not being powered off entirely.
Are you using an SN74HCT245 like the Teensy OctoWS2811 adapter?
I use the 74AHCT125 which I think is identical only Quad instead of Octal. Now I have the OE pins standard to the ground (active low). So you suggest pulling them high to turn the outputs off? Would that be 3.3V or 5V then? The A pins are connected to the 3.3V outputs and the Y pins deliver 5V.
What would be the benefit of this, instead of send nothing to data/clock? I assume that the APA102 chips can’t draw power from them if you don’t make those high?
The datasheet [1] says that the logic input voltage must be between 0-5.5V (Output Enable is just another logic gate internally). It also says that it the chip recognises anything below 0.8V as LOW and anything above 2.0V as HIGH.
You can just avoid sending data if that’s how you’d like to do it – FastLED leaves the pins LOW after data transmission is finished. It’s also less wiring.
Thanks! I did a test and can conform it works fine. Since the mosfet and OE both are active low I can combine them on one output pin.