Reading PWM signal in 4.3 framework on fez cobra II

NETMF & Gadgeteer
1

Hi there,

Does anyone have example code that shows how to read PWM signal in 4.3 framework and measure duty cycle and pulse high time ?

Thanks

2

I am interested as well. PWM support has changed. My code is crashing when the freq and duty are set to zero. I guess I’ll try to turn off the port instead, but an example would be great.

3

Below is example of PWM in 4.3 framework using signal capture to measure distance by ultrasonic sensor.

class PWM
    {
        public double final_distance;
        public uint[] signal;
        public SignalCapture pin; 

        public PWM()
        {
             signal = new uint[sample_rate];
             pin = new SignalCapture(GHI.Pins.G120.P1_5, Port.ResistorMode.Disabled);            
        }

        public void distance_measurement()
        {
            pin.Read(false, signal);
            final_distance = median(signal, sample_rate);
            Debug.Print("Distance: " + final_distance.ToString());
        }
}
4

I have a working pwm methods in 4.1, but when I make minor changes so that it will compile with 4.3 the pwm class does not work anymore.

Here are the methods:


public void Grab()
        {
            TurnOff();
            pwmA.Set(  20000, 100  );
        }

        public void Release()
        {
            TurnOff();
            pwmB.Set(  20000, 8  );
        }

        internal void TurnOff()
        {
            pwmA.Set(0, 0);
            pwmB.Set(0, 0);
        }

I have updated it to this for netmf 4.3


 public void Grab()
        {
            TurnOff();
            pwmA.DutyCycle(100);
            pwmA.Frequency(20000);
            pwmA.Start;
        }

        public void Release()
        {
            TurnOff();
            pwmB.DutyCycle(  8  );
            pwmB.Frequency(  20000  );
            pwmB.Start;
        }

        internal void TurnOff()
        {
            pwmA.Stop;
            pwmB.Stop;
        }

Any ideas why the 4.3 code is not working? I guess I will need to put a scope on it to see what looks different about them.

5

What is the equivalent call to pwm.set under the new implementation for 4.3?

6

https://msdn.microsoft.com/en-us/library/jj610646(v=vs.102).aspx has all the details you need - and you should have a link to that in your kitbag of tools, for sure. PWM specifics https://msdn.microsoft.com/en-us/library/microsoft.spot.hardware.pwm(v=vs.102).aspx

7

Thanks for the links. They are nice, but seem to me limited to what intellisense in visual studio provides. I need to know why code that works in 4.1 does not work in 4.3.

Thankfully in GHI Code Share ChrisO has posted a replacement class for PWM

https://www.ghielectronics.com/community/codeshare/entry/932

and the description of the Duty Cycle is as follows

‘Duty Cycle expressed in Percentage EG: 0.5 gives 50% Duty Cycle of the Frequency’

This led me to realize that the Duty Cycle threshold changed from 0 to 100 in 4.1 to 0 to 1 in 4.3. Ah the smokin guns appears!

It was pretty cool to see the 4.1 pwm signal via an oscilliscope and after I realized the duty cycle change I was able to reproduce them in the 4.3 code.

Thanks for the feedback it kept me going.

8

I am using the PWM output of the EMX to grab and release a magnet. I have the solution, but would like to know if it can be explained.

When the following runs a short pulse is generated. This is not expected. I would expect a transition from low to high and the signal to stay high, since the duty cycle is 1.

        public void Grab()
        {
            pwm.Stop();
            pwm.Frequency = 20000;
            pwm.DutyCycle = 1.0;
            pwm.Start();
        }

When I watch the signal with an O-Scope, and step through the code. When I hit the pwm.Frequency = 20000; line the pulse rises and then when I step through the pwm.Start(); line, the signal goes back low. So I commented out the pwm.Start(); line and now the signal stays high.

Any Explanations???

9

I assume you use actual PWM elsewhere ? Otherwise, PWM duty=1 is just a digital output and you should use it that way…

One other test you might try is not stopping the PWM and just changing the duty, as a stop/start will always tend to give you an unexpected pulse length that could cause issues (I’ve not ever tested this though, give it a whirl and check if it works)

10

I don’t think we should use the PWM as a digital output either, but I am just updating code to 4.3 now.
Just changing the duty cycle is a good idea, I will give that a try, I will need to build a small circuit with an led so I don’t have to keep setting up the o-scope.
thanks

11

I’d always want to make sure the 'scope didn’t show a glitch as well :slight_smile: