PMW output oddity expansion box

[rufessor]

Hi -

I am not going to post my PDE here unless it becomes necessary as I am working on a new one that will replace the existing version. But, what I have going on is odd... I am curious where to look trouble shooting wise...


So, I have mostly untouched the progressive sun phase PDE implemented on my PWM expansion board, and it was all working fine for over a month when one day the lights didn't come on (I DID not change or otherwise alter the program). So I reset the controller when I got home and it all came on correctly, i.e. PWM output by DC voltage was correct for each string.

Then I noticed that the lights stayed on too late and never dimmed... I presume they reset at midnight as the next morning they were not on.

What I have discovered, is that apparently the PWM slope is working, since if unplug and replug the PWM controller from the main-about when it should be in the middle of sunrise, with only Blues on, the Blue strings come on. If I let it sit, it will never turn on the Whites and the blues do not ramp in intensity as measured using a digital volt meter from the PWM board output. If I then power it on and off later in the morning when the whites should be on, they turn on, and all channels are set to the CORRECT PWM intensity, i.e. 70 for channel 0 and 2 80 for 1 and 2 and 3 and again in the evening powering it off and on will hit the sunset blues only, then finally the violets only etc.

So- OBVIOUSLY, my PWM slope function is being fed the correct intensity values, and the times seem about right. Where should I look to see how it is that the PWM board on a power cycle is outputting the correct value, but when left alone gets locked.

Its almost like PWM slope output is executed only on reset and not from the loop (recall this is running on the processor on the PWM board)... but since its a function call to a library and I did not change anything from when it was working fine, to when it started this... I am wondering if it would be more likely a memory thing, like maybe I screwed up the sunrise/sunset part or the cloud effect and memory is going weird as the progressive sunphase calculation lengthens the day.

I kinda want to figure this out even though I am almost done and ready to test the other PDE I am writing using the sun/moon library I found.. (I will likely have Q's in a new post tonight before I try this new PDE)

Really curious where to look or even how to trouble shoot this. I will post PDE if necessary when I can do so...

[rimai]

I went over this one with Deckoz through the weekend and discovered there was a bug in his code.
PM him if he doesn't pitch in.
It's very easy fix, but he may have gone and made some other adjustments that I'm not aware.
Let's wait for him to say anything.

[rufessor]

Dang.... I was going to PM him but decided I would just post here.... I kinda wondered if it was something with the sunrise/sunset as the season is changing fast now so I would have possibly triggered a new line to run in the code. I will probably cause more than a few issues with my new INO I am writing trying to use the library from SWFLTEK. SHould be fun, really curious about this as it may be something I can learn from! Undoubtably, that.