Please excuse my long rambling post. Hopefully it makes sense! I have a good idea of the issue now but will present all the data I have in case I'm wrong.
The circuit is as per the diagram up the thread, except I have some indicator LEDS (and resistors) in parallel with the chip, so I can monitor each pin status. LEDs 1, 2 and 3 are on GPIOs 2, 3 and 4 respectively. Ground is pin 6. The big LEDs are fed from a 9v battery positive and ground through the chip. Ground is connected to the pi and 9v battery using a common rail on the breadboard (star connection as suggested!). I'm using GPIOZERO with the pins designated as LED or PWMLED, if that makes any difference. I'd post a picture but I can't for the life of me remember how I attached the last ones..... Running the LEDs on either PWM or just on/off causes adjacent LEDs to glow, when they are not turned on. I'm calling this bleed, for want of a better term.
To figure out what is happening, I put a multimeter between the junction between GPIO4 and pin IN3 (for LED3) and ground then ran a quick programme to turn the LEDs on one by one. As each pin goes high, I'm measuring the following voltages at said junction:-
All off - 0.0v
LED1 (GPIO2) on - 0.03v
LED2 (GPIO3) on - 0.25v
LED3 (GPIO4) on - 3.27v
So there is a small voltage at the junction between GPIO4 and the chip IN3, when the Pin is not high, which increases as the high input gets nearer. Could this be enough to produce an output in the big LED? I know the gain on the transistors is high so a voltage on the input might be the issue?
To see if this is a chip issue or a PI issue, I disconnected the pi and manually put 3.3v on each input in turn. This worked as it should and there was no bleed on the big LEDs. Each one came on on its own. I measured a voltage of 0.8v at IN4 to ground however, regardless of which pin (except IN4) I probed. This can only be coming from the chip as nothing else was connected to either IN4 or OUT4.
So, manually probing 3.3v on inputs works fine but using the Pi results in bleed. Next test... Pi connected to IN1 only, run the test programme to see if there is any bleed? Nothing..... There is only bleed if the GPIOs are connected to the INs on the chip. Could it be some interaction or voltage running through the PI's GPIOs from the chip that is affecting the output?
Final test......Other chips.
I have 5 chips. They generally look ok except one has a visible blemish on one leg. They appear free from contaminants or debris. 4 of them have the bleeding problem. The fifth works perfectly with the 3 test LEDs all remaining completely off until the input is triggered. I did the same test as above on the seemingly working chip:-
Good chip:-
All off - 0.0v
LED1 (GPIO2) on - 0.0v
LED2 (GPIO3) on - 0.v
LED3 (GPIO4) on - 3.27v
The good chip doesn't have any voltage on the junction until its turned on.
So is that it, case closed? 4 out of 5 chips are faulty? I did get them off ebay....
Thanks again...
Phil
The circuit is as per the diagram up the thread, except I have some indicator LEDS (and resistors) in parallel with the chip, so I can monitor each pin status. LEDs 1, 2 and 3 are on GPIOs 2, 3 and 4 respectively. Ground is pin 6. The big LEDs are fed from a 9v battery positive and ground through the chip. Ground is connected to the pi and 9v battery using a common rail on the breadboard (star connection as suggested!). I'm using GPIOZERO with the pins designated as LED or PWMLED, if that makes any difference. I'd post a picture but I can't for the life of me remember how I attached the last ones..... Running the LEDs on either PWM or just on/off causes adjacent LEDs to glow, when they are not turned on. I'm calling this bleed, for want of a better term.
To figure out what is happening, I put a multimeter between the junction between GPIO4 and pin IN3 (for LED3) and ground then ran a quick programme to turn the LEDs on one by one. As each pin goes high, I'm measuring the following voltages at said junction:-
All off - 0.0v
LED1 (GPIO2) on - 0.03v
LED2 (GPIO3) on - 0.25v
LED3 (GPIO4) on - 3.27v
So there is a small voltage at the junction between GPIO4 and the chip IN3, when the Pin is not high, which increases as the high input gets nearer. Could this be enough to produce an output in the big LED? I know the gain on the transistors is high so a voltage on the input might be the issue?
To see if this is a chip issue or a PI issue, I disconnected the pi and manually put 3.3v on each input in turn. This worked as it should and there was no bleed on the big LEDs. Each one came on on its own. I measured a voltage of 0.8v at IN4 to ground however, regardless of which pin (except IN4) I probed. This can only be coming from the chip as nothing else was connected to either IN4 or OUT4.
So, manually probing 3.3v on inputs works fine but using the Pi results in bleed. Next test... Pi connected to IN1 only, run the test programme to see if there is any bleed? Nothing..... There is only bleed if the GPIOs are connected to the INs on the chip. Could it be some interaction or voltage running through the PI's GPIOs from the chip that is affecting the output?
Final test......Other chips.
I have 5 chips. They generally look ok except one has a visible blemish on one leg. They appear free from contaminants or debris. 4 of them have the bleeding problem. The fifth works perfectly with the 3 test LEDs all remaining completely off until the input is triggered. I did the same test as above on the seemingly working chip:-
Good chip:-
All off - 0.0v
LED1 (GPIO2) on - 0.0v
LED2 (GPIO3) on - 0.v
LED3 (GPIO4) on - 3.27v
The good chip doesn't have any voltage on the junction until its turned on.
So is that it, case closed? 4 out of 5 chips are faulty? I did get them off ebay....
Thanks again...
Phil
Statistics: Posted by philmulrain — Wed Jan 24, 2024 4:51 pm