I will say though even when it when full output, reef-pi still was unable to control it so I think its part of the problem but maybe not, bit is broken right at the crimp point, hoping that is why things are flaky....so I need to fix that first and then go from there but how to quickly fix it, that is the question.
Well not sure this is the issue but that cable is messed up on one of the PWM channels, would that be enough to mess it up, I dunno...Wow, and there I was figuring out all the internals
The flicker you see is likely a beating signal between your input PWM and the internal PWM. The closer your PWM frequency to the internal frequency, the more visible the flicker.
Whats going to be the easiest, lol. THe circuit I am running is the one listed on the lighting guide, couple of resistors and npn. The flickering seems to be worse at the lower voltages vs the higher ones, so just need to remove the flickering and still figure out the other board, lol.
I do not think I have any diodes at least not small ones, what kind of diodes am I looking for. The only caps I have on hand, the smallest ones are 100uf, does that matter?
Basically like this. The diode charges the capacitor over the original 1k resistor, the other 1k resistor limits the discharge when the transistor is activated. That way charge and discharge are about equal and you get about 0-10V from your 0-100% PWM.
A 1microFarad capacitor is minimum, 10 would be better to reduce residual Flimmer.
that gives you 2 channels of 0-10V „DC“.
If you do that you can, just for fun, measure „0V“. I imagine that should actually be something like 0.6V ^^
General purpose diodes for more than 10V. I think 1N4001 should be fine. You don’t need those diodes, they just improve linearity between 0 to 90%. Without them 50% will give you more like 40% intensity.
Hm, now that we know the problem you might test continuity of the cable lines during flexing.
I just want to make sure I understand the fritz update above, I assume I keep all the existing resistors in place and I am just adding to the original design, correct?
Basically like this. The diode charges the capacitor over the original 1k resistor, the other 1k resistor limits the discharge when the transistor is activated. That way charge and discharge are about equal and you get about 0-10V from your 0-100% PWM.
A 1microFarad capacitor is minimum, 10 would be better to reduce residual Flimmer.
that gives you 2 channels of 0-10V „DC“.
If you do that you can, just for fun, measure „0V“. I imagine that should actually be something like 0.6V ^^
Well, if both your PWM‘s 5V and your 10V come from an LM2596 each that both share the same 12V supply, I think you technically don’t need to connect these „two“ GNDs (because they are technically one and the same).I just want to make sure I understand the fritz update above, I assume I keep all the existing resistors in place and I am just adding to the original design, correct?
Here is my current layout and trying to figure out how I need to correct it in order to match what you have (no laughing, lol.
From the top comes in PWM and ground from reef-pi, down to the right comes in 10v. One question is the long ground wire that I have from the top to the bottom, I can shorten that and have it start at the very top, is it going to matter if that path travel through the cap? My other question would it matter if I have the diode and resistor coming from the same hole? I do not think that matters but if not I would need to move my npn down, trying to avoid as much rework as possible.
Sorry for the dumb questions, just trying to figure out how to move things around to match the diagram.
The buck converter supplies is stepping down 12v to 10v, I have a separate 5v supply feeding the rpi. So I have a ground coming from buck and the rpi side of things if that make sense so they all tie together. I think that is an interesting idea, I will swing that one ground wire across then down and that should give me the space to lay it out as pictured.
