Interrupt glitch filter troubles

Archive
1

I changed my glitch filter to an interrupt-type (on rising or falling edges).
I set the time to 35ms

 static InterruptPort rdButtUp = new InterruptPort((Cpu.Pin)FEZ_Pin.Digital.ButtonDown, true, Port.ResistorMode.PullUp, Port.InterruptMode.InterruptEdgeBoth);

          TimeSpan gfilt = new TimeSpan(0, 0, 0,0,35) ;//35ms glitch filter
            Microsoft.SPOT.Hardware.Cpu.GlitchFilterTime = gfilt;

However, I put in 2ms wide pulses every 800ms & the interrupt triggers every so often (using debug print to see it). The wave is clean from a generator.
Even a 10ms or 1ms pulse will occasionally trip it…what’s wrong? I also tried setting the filter to 20ms, same thing

Is this a priority thing? Can I change the priority? I’m not familiar with netmf interrupts

2

In addition to your problem I also encountered CPU-overflow (I think…):
See the discussion in this post:
http://www.tinyclr.com/forum/topic?id=2546&page=2

I never got a answer and closed the case…

3

Try making the pulse greater than the glitch time?

4

I’ am afraid I dont get what you mean: if you want to stretch the original pulse then you need a hardware device called a mono-stable multivibrator (not what you think it is :slight_smile: ) , or a RC-filter (resistor and capacitor) on the input.

My alternative would be to use your own RLP interrupt-handler…and implement the filter-algoritm yourself.

5

if the original signal is clean then you do not need a glitch filter. a glitch filter is u squally used with a mechanical switch.