This thread was started as a result of an issue that Toodle Pipsky highlighted in this thread:
http://www.doityourselfchristmas.com/forums/showpost.php?p=64024&postcount=54
Testing a Ren-C/Grinch combo and using LEDs to monitor the Grinch outputs.
Testing at 60Hz (ZC at 120Hz) did not show any flickering during any test scenario.
Using a pulse generator to create a ZC signal at 100Hz, testing shows flickering of channels (some not all) that should be OFF when other channels are ramped UP or DOWN. When other channels are full ON or flashing full ON/OFF, I get no flickering.
To see this flickering I had to run the test in the dark because the flicker is so faint.
The attached image shows what is happening to the channel that should be full OFF when it flickers. The blue trace is the ZC signal (expanded to catch the flicker) and the yellow trace is the flicker. The flicker pulse always seems to occur at the same point in the ZC signal (if that helps Phil).
Good idea starting a new thread. I'll pop my meagre contribution here.
Phil asked: Which channels are you using for that LED board? Also, how was that video sequence generated? Was it using a Vixen sequence or the Vixen channel tester? So is that flickering occurring when Vixen is sending data or does it happen when vixen is just sitting idle and not sending any serial data to the controller?
Channels 1 -3 (first block of 4). I'll try putting them on "further up the chain". The video (
) was created using both a sequence and channel tester. The worst flicker, the most noticable one at the start) was using the channel tester with channel 1 on solid. But then that flicker didn't happen the next time I put that channel on solid or ramped it up and down. The flicker can happen when Vixen is sending data or when it's sitting idle. I can stop a sequence, it'll be sitting there idle for a few minitues or seconds and the flickering will start.
The trouble is it is so random - it just won't do it in the same spot or do it with any kind of regularity. And for example it was really bad when I first fired it up, but then when I went to film it, it had calmed down considerably.
I'm afraid I don't have all that fancy gear, so I can really only offer observations. But I hope that's at least helpful! I'll keep running it and see if anything else pops up.
Toodle...for now, my questions for you are moot, since RL has some scope traces.
RL, do those same glitches occur with a 120 Hz zero-crossing? BTW, I assume that the top trace in that picture is the output signal from one channel of the Grinch/595 board.
i can confirm that this is basicly what i have seem on multiple ren-c/grinch combo's.
as i said in the other thread i think it is to do with the chips not having a off clocked out in time when the ZC pulse goes positive. thus keeping the led turned on once the clock is stopped.
Im wonderiing if its possible to change the code to clock for 1 more cycle once the ZC publse goes high so as to clock out a 'off" signal to all channels.
Also clocking out an "off" signal would probably help the random flickering i get sometimes when the ren-c/grinch is turned on but there is no data yet.
i hope this makes sense.
edit: ill try and post the scope pics i grabbed on the weekend
Peter
Phil,
Yes, the yellow trace is a single channel output of the grinch going to a LED/resistor.
At 120Hz ZC, I get no flickering or glitch pulses.
Just like TP, if I pause the sequence during a dimming ramp (down or up) it will randomly flicker. The flickering is not regular or predictable, it just happens when it wants to and (for me anyway) only when a dimming value is used.
Also dimming multiple channels causes it to happen more frequenctly and over more channels. If I dim only 8 channels the flickering only seems to affect 8 other channels but if I dim 16 channels then all other OFF channels flicker.
I also found out that while running at 120Hz ZC and having no problems, if I change the ZC to 100Hz (an no other change) then the flicker/glitch comes back full time at the same time in relation to the ZC signal (approx 35usec from leading edge). If I go back to 120Hz ZC the signal clears up and works properly again.

