Again wrong - you will never have 1W in both Shottkys. But good luck anyway.
The point of the two diodes is to prevent the power supply that is suppost to power WW to power CW and vice versa via this current path:
If those diodes were not there, the two supplies would be completely in parallel and the one button for one light temperature scheme would not work.
Okay, I am intrigued. Up to 2.5A will flow through the diode.
The diode has a certain forward voltage that I expect to be in the range of 0.4-0.6V, looking at a few datasheets of 3-5A Schottky diodes.
What am I missing here?
That never both diodes will carry 2.5A at the same time ( that would be 5A through the Led strip ). If both buck converters are on, only the one with the ( slightly) higher output voltage will source your LED strip. If the voltage would be really ( nearly) the same, the current will be split, so that the sum of current through both diodes would be 2.5A. Which again leads to a loss of 1W overall.
Or will your LedStrip really draw 5A with warm and cold on?
If GND2 is connected to ground of power supply 2 then no current would flow through either diode
Yes. So? It's 23m( 2 color temperatures) in parallel fed from both sides, on aluminium heatsinks.
Look, this combination of condescension and wild assumptions is really frustrating. There is a reason I chose 2.5A power supplies. I measured 2A for each color temperature, and I will most definitely have that current run over the diodes, which will also most likely have a higher forward voltage than 0.4V under those conditions.
GND2 is the label of the low side of power supply 2. I have no idea what you are talking about. If I leave those diodes out, the two power supplies will be in parallel and either will power both color temperature (while being spectacularly overloaded if the other one is off).
I have even given you a link to the simulation of this aspect of the circuit at the very beginning (which does not take into account the internals of the power supplies, but those are not relevant for the diode question).
You never told us the specs of your LED strip, so we have to do assumptions. Bye
I literally told you, or however however suggested it first, that I am dealing with 2.5A and it would dissipate more heat that I am willing to dissipate at that point, and you plainly told me that I am wrong, because you decided from "YoU DiD nOt GiVe Us ThE SpEcS", you can derive that the actual currents that I did mention are wrong.
I don't need that. Yes, bye.
@ElCaron
What I would do.
Use Switch 1 to power the AC to the power supply.
Use Switch 2 to select High or Low on the Micro to select WW or CW.
Just 1 power supply.
Not true. For example, Q2 is only connected the the GND of power supply 1. If power supply 2 is on and 1 is off then there can't be any current flow through through Q2 back to power supply 2.
That has already been suggested, but having both on is also a desired mode. If it was XOR, I also would not need two power supplies.
I really do not not know what you are talking about.
If you ask "what is the point of the diodes", are you asking
a) why I don't replace them with a wire or
b) why I am not leaving the respective connection open?
If I replaced the diodes with wires (case a), then the low sides of the power supplies (labeled GND1 and GND2) would be directly connected, just as the high sides are, so they would be in parallel and the blue current path would exist.
If I left the connections open (case b), then the MCU would have no low side connection.
You ask a lot of "educative" questions, but I have answered all of them (with things I already knew) without any constructive reply from you. I am always willing to accept that I am on Dunning-Kruger's mount stupid, but in that case you are not a good educator to get me off that mount, and my suspicion rises rapidly that it is actually you who did not understand the circuit.
C'mon, if you get digital input from the switch, it's up to you to make the logic. One press warm, next cold, next both, long press for dimming etc...
1 Like
Noted, but as I said, I want one switch for cold and one for warm, and not add a code book to my bathroom.
Up to you.
So leave out even the microcontroller if you can't do dimming...
"The bathroom user cannot do dimming with the light switches" is not the same as "You can't do dimming".
Well, you asked if this will work and I'm telling you no but you seem absolutely certain there is nothing wrong, so just build it.
3 Likes
Up to you.
With motion sensor you can leave out the switches as well.. 
Simuation behaves exactly as I expected in every single node.
- DPDT switches top left and right: I have modeled switching off a power supply as cutting the connection on both sides and a 5kOhm resistor in parallel (I measure 5kOhm on my switched off power supply).
- Switches bottom left and right: I have modeled the MCU with switches to the MCU high and low side and added a resistor to read the potential.
- The diodes on the bottom prevent exceeding the maximal Vgs of the left MOSFET when the drain is weakly pulled to the high side potential of the right side power supply when the left is switched off and vice versa. Wouldn't do that for an IC, but the MOSFET has a specification for +- max Vgs.
- On the top right there is single terminal 24V. That is to pull the absolute potential of the circuit to 24V on the high side of the power supplies, else the simulator tends to make everything negative. It can go without affecting the functionality.
If the Zeners I ordered arrive until then, I'll physically test it next weekend.