Raspberry(3b+)-PCA9685-12V Fans

[Ch3fk0ch]

I will try other liberies to create the pwm pf the raspberry.

[Brandon92]

Did you do the measurements that I stated in my last post?
And what frequency are you currently using?

[Ch3fk0ch]

Yes i did it on your way.
It was set to 25kHz as i messured the 5.xkHz output.

I read something that the original Pi library is bad for PWM.
Pi-Blaster or WiringPi library now to see if that makes a differnt.

[Brandon92]

Okay. Can you show the scope image? So, that I can check the signals.

[Ch3fk0ch]

Look at Post 47.

[Brandon92]

Look at Post 47.

Yes. I saw that. And that is not the signal were I'm looking for. See my post below that one

[Ch3fk0ch]

Thats what i did.
Placed one at the Emitter and one at the Collector.

[Brandon92]

(I put the black Cabel on GND and the Red on the Gipo)

I believe you measured at the output of your arduino and your Rpi. And not across the transistor. If you have two input you could also include the base signal as well.

So, the black one on the ground. And one channel on the base and the other on the collector.

[Ch3fk0ch]

So Ok,
Standart Gipo library set to 25khz:

Yellow= RPi Direct
Blue= Transistor Emitter and Collector

IMG_4603.JPG (155.4 KiB) Viewed 2046 times

Yellow= Emitter and Base
Blue=Emitter and Collector

IMG_4602.JPG (144.32 KiB) Viewed 2046 times

Greetings

[Ch3fk0ch]

I think first have to find a solution for the frequency.
I tested it and i cant get the 25kHz.

Unbenannt.png (12.73 KiB) Viewed 2039 times

Same results when i use a Python script instead of NodeRed

[Brandon92]

The blue line is not what I was expecting. I expect the green and light blue one in the picture below to be present. Are you sure that you connected the transistor in the right way around. And do you have a other one for testing?

[Ch3fk0ch]

Hi.
i will Build it up new with a new Transistor and test it again.

[Ch3fk0ch]

Hi,
i build it up new with a new Transistor:
Same result.

[Brandon92]

Strange, this should work. Did you measure by the way the voltage of the PWM pin of the fan. So, whith noting connected to it. It should be 5V.
Could you show a clear image of your setup, including the measurement.

And did you use the pwm pinnen of the Rpi?

[Ch3fk0ch]

Hi,
i use BCM 18 (PWM0).

I found a strange way of the Full PWM control at 25kHz with PIGPIO.
So i make this in the Terminal:

Code: Select all

sudo pigpiod
python
import pigpio
pi=pigpio.pi()
pi.hardware_PWM(12, 25000, 1e6*0.25) # 25000Hz 25% dutycycle(normaly)

And get this curve (not exectly):

Unbenannt.png (59.62 KiB) Viewed 1973 times

I think this rounded pwm signal is also not the best for it:

IMG_4610.JPG (167.75 KiB) Viewed 1973 times

I also tested the same transistor at the Ardruino there the signal looks normal.
Maybe its cause of the PWM Voltage form the Raspberry

THe PWM Pin of the raspberry only had 3.3V or not?
I will make new Pics tomorrow

Greeetings and thanks

[Ch3fk0ch]

The the Voltage of the Rasperry PWM Pin is 3,11V
Here are some photos of the circuit.

IMG_4612.JPG (112.85 KiB) Viewed 1945 times

IMG_4613.JPG (109.86 KiB) Viewed 1945 times

IMG_4614.JPG (112.55 KiB) Viewed 1945 times

[Brandon92]

Hello,

Is the last scope image with the fan connected? Because all the signals are not what I would expect. And by looking at your images, the connection looks good.

If your Rpi signal is 0V, so the yellow signal on the scope is at the same level as the "B'. If that is the case, I would expect the blue channel to be at a 5V level. Like the the image that I posed with the simulation.

So, you got it to work if I see your table. Its is kind of correct. Because when you are using one single NPN transitor in this way. Your PWM signal is inverted. So, this means that when you send a 0% duty cycle, the fan see it as a 100% duty cycle. And when you send a 100% duty cycle, the fan see a 0% duty cycle.

[Ch3fk0ch]

Hi,
no it was without Fan.
The Raspberry just have 3.3V on The GPIO's not 5V.

I not really know what u mean with:
"So, you got it to work if I see your table. Its is kind of correct. Because when you are using one single NPN transitor in this way. Your PWM signal is inverted. So, this means that when you send a 0% duty cycle, the fan see it as a 100% duty cycle. And when you send a 100% duty cycle, the fan see a 0% duty cycle."

I know why the Signal is inverted^^

What i dont know is why the Signal fromt the Transisor at the Raspberry looks so strange.
With the Arduino i get a good looking Signal (ardruino uses 5V)

[Brandon92]

Ehm okay. Let's go back to something simple.
Can you build this schematic:


And connect the one probe to the base and the other to the emiter of the transistor. Amd the gnd to the gnd.

And use for r2 the same resistor as you use write now. And in this case is vcc the 5v rail. And use for r1 1000 ohm.

By the way, what is your value for r2, and did you measure it?

[Ch3fk0ch]

Hi,
1. for r2 i use the 470ohm resistor
2. Connecting GND and Emitter ? I think u mean Base and Collector or?

IMG_4618.JPG (93.45 KiB) Viewed 1869 times

Channel A= Collector
Channel B= Base

[Brandon92]

I hope this diagram is more clear for you, about how to connect it correctly. If you have quistion about this, or I need to check it. Let me know.

pwmled_con_scope.png (101.62 KiB) Viewed 1863 times

Okay, your resistor is correct.

[Ch3fk0ch]

Thats the way i make it.
Just that i connect Channel B to the Base and A to the Collector.

[Brandon92]

Okay. That is good then. I also see that you have enabled the ac coupling option. And you need to turn it in dc coupling.
You can see this because the ground level of your signal is not at the same level as the channel label.

But despite that, your signals are looking good

[Wirepower84]

Did this ever get working? I’m in the process of doing the same thing with my raspberry pi.
Need to know what material I require, resistors, transistors etc