The plot sickens......

[chuckd]

Further investigating my 4.4 firmware box and why it doesn't fade my LED's properly, I've found something that is a bit disturbing.

I brought the box inside, and put it on a single network. I then hooked up just one of my 50 count C6 LED strings (CDI, new version) to channel 1, and tried to dim the light strand through the hardware utility console. It dimmed just fine, but...... it had no terminator (47K 1 watt resistor), so the fade got very blinky around 30% or so.

I slapped on the terminator (which we all do to fix this blinky fading problem with LED's), and lo and behold, the strand will no longer dim. The waveform on the scope looks quite strange, and the lights erratically flash all over the place.

Using any other box I have (older V5 or less boards with 4.3 or 4.01 firmware), fades work just fine with the terminator.

So, the new (V7 blue board 1602 with 4.4 firmware) fades like the other units, UNLESS I've put on a 47K terminator to the strand. Once I put on the 47K terminator, everything goes to heck.

Can anybody else verify this with their new V7 boards?

Thanks,

Chuck

[chuckd]

OK, now I have a picture of what's up.

I noticed as I slide through the intensities, the lights flash all over the place. Looking at the scope, I noticed occasionally the 'dimmed' waveform occasionally had parts of a regular sinewave in it. After a couple of attempts, I caught it on the scope. Here's a picture:



So, what the heck?

Something look loose to you?

Every channel (1 through 16) does this.

Chuck

[jaosh]

wish someone would find a fix to this problem. got everything wired up with about 80% LED and the only issues is with the new controller with 4.4 fading in/out. im not an expert by no means but all other LEDs fade just fine. thanks in advance

[Guest wbottomley]

Send LOR an email at support@lightorama.com and explain the problem.

They usually write new firmware when a problem arrises.

Give them time to get in the office Monday morning.

[chuckd]

Wow, so I'm not the only one?

Have you also noticed that the unit randomly flashes the LED's, even when all channels are off? You may have to wait a bit, but it sure happens to me (about once or twice a minute).

[jaosh]

sent to dan hopefully it gets fixed soon. hey it dont look so bad if you custom your fades with shimmer at least they all "flicker " at the same time ha ha

[chuckd]

I sent it with the picture of the scope trace as well. You can imagine how weird it was about 1:00 AM last night, seeing a couple of my trees randomly flashing with no show running. Very spooky!

I have no idea why a terminator makes a difference with this, however.

[Jeff Sand]

Looks like the triacs are definitely not firing right (or being triggered correctly) on the bottom half of the waveform. A couple of questions for you:

1) Have you tried duplicating this experiment using incandescent lights instead of LEDs?

2) Have you caught the full wave occuring on the top half of the wave or only the bottom half?

Based on that scope graph I'd say you're more likely looking at a hardware problem than a firmware one.

BTW, Anybody know what the hardware differences are between the V5, V6, and V7 blue boards?

[chuckd]

The same problem happens when I simply plug in an incandescent table lamp. The waveform for the table lamp example looks the same as the LED strand with a terminator.

Also, there are upper waveform problems. Here's a couple of examples.



And another:



I thought other people might want to see what it's 'supposed' to look like. This is a shot from one of the older working units:



An interesting trace is what this looks like WITHOUT a terminator on the LED strand. You can see what the capacitive effects do to the zero crossings:



You can see why there would be problems without terminators at the low end of fades, because the zero crossings are so messed up.

The REALLY weird thing is, this new V7/4.4 unit handles this far, far better than the good looking trace with nice level zero crossings.

The scary part, though, is that I don't seem to be the only one with this problem!

[Jeff Sand]

Chuck,

Can you find a place to attach a 2nd probe between the optoisolator and the triac gate? Likewise, somewhere further upstream of the optoisolator? It would be useful troubleshooting to identify where source of the misfiring is. You probably know this already, but the idea of a triac-based dimming circuit is basically thus:

1) Some analog circuitry sits on the AC line and generates a pulse when the voltage crosses zero
2) Zero-crossing signal gets sent into a microcontroller where it starts a timer
3) Some microseconds later, microcontroller generates a signal to tell the triac gate to open and let current flow. How many microseconds is what determines the dimming level
4) This signal usually goes through an optical isolator or other buffer circuitry for protection
5) Triac is signaled, opens gate to let current flow until next zero crossing when it automatically closes
5) Repeat 120 times per second

Looking at the evidence so far, and seeing as how this might not be an isolated incident, I'd say the problem is likely at either step 3 (software [firmware] problem) or step 4 (hardware problem betwen microcontroller and triacs). In any event, what you want to find out is where it's breaking down -before the triac, before the optoisolator, or before the microcontroller. I'm as curious about this as you are.

[cmoore60]

Jeff Sand wrote:

BTW, Anybody know what the hardware differences are between the V5, V6, and V7 blue boards?

Jeff I recall reading where Dan stated the new firmware 4.4 was due to the use of a different processor and signals being on different pins.

Chuck

[-klb-]

cmoore60 wrote:

Jeff Sand wrote:

BTW, Anybody know what the hardware differences are between the V5, V6, and V7 blue boards?

Jeff I recall reading where Dan stated the new firmware 4.4 was due to the use of a different processor and signals being on different pins.

Chuck

On the PC series at least, I believe that the V2 hardware, and 4.4 relate to changing some of the microcontroller pins being used on the header, not necessarily any changes elsewhere.

[shfr26]

I'm not sure,but that looks like a sinus rhythm with occasional PVCs

[rwertz]

Chuck,

I had similar a problem a couple years ago. I purchased the two controllers startup kit and one of the controllers had serious flicker / inconsistent control when dimming.

If I had scoped the waveform I would have expected a similar response. I say this because I was running Sylvania StayLit strings (the kind with bypass diodes in the base) and sometimes I was seeing half the string dimming and half the string full on.

I tried all the usual efforts (eliminate fluorescents, isolate comm. lines, etc.) and nothing worked. As a last resort I updated the firmware and the problem went away. There was no documented flicker issue with the original firmware, and the other controller with the same version ran fine.

I don't know if it's possible to get a bad firmware image loaded onto the processor. But that's what I suspect was wrong. Maybe you could download the same firmware version and see if it goes away.

[chuckd]

The problem is that the V7 board at this point only accepts 4.4 firmware. None of the others can be used.

[JonB256]

Chuck, I don't see in your posts that you "reflashed" the firmware. I think that is what Reg Wertz was suggesting, and so am I. Your TRIACs and Optos must be good or you would never get the correct waveform, so I'd suspect either microcontroller chips not solidly seated or a bad firmware flash. Look very closely at the PIC pins to see if perhaps one of them has a "bent under" pin instead of going into the socket properly.

[chuckd]

I'd try that if I could download 4.4 firmware, but so far i don't see where it can be downloaded. The support site only has 4.3, and the 2.60 release doesn't have it either.

I'll yank the processor and reseat it just to make sure.

Any thoughts that this might be a zero crossing detector problem?

[Jeff Sand]

chuckd wrote:

Any thoughts that this might be a zero crossing detector problem?

I would tend to doubt it, because if it was missing the zero crossings, in theory, then the processor wouldn't be firing the triacs at all and you'd see a flat half-wave rather than a full half-wave. Then again, without the schematics, it's hard to say.

[chuckd]

Jeff, that's a comment dear to my heart. I've been pleading here to get schematics for the units to no avail. I do realize that most people out there won't use them, but quite a few of us are technical enough to put them to good use.

With 53 LOR units, I've been having one or two fail each season. I would have no problem troubleshooting them myself, but without schematics, I just don't bother. I do think LOR would prefer they not be returned for repair, as they seem swamped this time of year. Getting support is pretty tough for me, for some reason.

[chuckd]

OK, pulled the processor, and every pin was just fine. Reseated it, and the problem is still there.

I won't have time to poke around until later tonight, but I'm not sure why I'd poke around the triacs/opto's, when all 16 channels do exactly the same thing. That was why I suspected the zero crossing circuitry.

Anybody know if that's changed in the V7 board?

And, do the processor I/O pins go directly to the triac controllers?