HELP 12v RGB source to 5v RGB light

Other Hardware General Electronics
1

So not sure if this is quite the right place, but I figured there are a lot of electronics savvy people on the Arduino forums so it would be worth a shot. Not sure if this is simple of not but here's my issue:

I have a 12v common anode RGB source and am trying to power 5v common anode RGB lights on a pc fan. Essentially being able to control color with the 12v source. I know the easy answer is get a 12v RGB light, but unfortunately both Corsair and Thermaltake in their infinite wisdom decided 5v was a good idea and flip the middle finger to Motherboard mfgs just so they could sell stupid proprietary controllers :confused:
Would a voltage regulator work? would I just hook up the 12v+ anode and one of the RGB cathodes? Or would that not work because - is split over the R G and B. Hopefully someone could help me out.

2

Or would I have to split my 12v to 3 buck voltage regulators:
12v+ and R to one
12v+ and G to another
12v+ and B to a third
the connect my 3 12v+ together and run each cathode to its proper color

3

Sorry but you will have to explain in more detail about what you have. What is a 12V RGB source?
Please provide links to the controller and say what sort of LED you have, for example is it a high power lighting LED?

4

You need a NPN or PNP transistor with Arduino PWM pin to driving the 12v RGB LED.

5

Okay so my "source" or signal output is my Asus Motherboard header which provides a 12v 2A RGB signal, which since the leds are dimmable I think means its pwm. The fan has the input 5v+ R,G, and B on its own PCB where (4) 4.5mm SMD RGB LEDS are mounted, I don not believe these are high power LEDS

6

Since a mainboard is potentialy at stakes, let this be verified by someone else, but you should be fine just connecting the 5V pin to some 5v source on yur board and the rest to the header on the mainboard.

7

Unfortunately no easy 5v sources all headers are 12v unless of course I use a usb header but then I'm limited to 500ma and I'm not sure if that would be dicey. not gunna touch the 5v from the Power supply

8

Could you post a picture of that 12v common anode RGB source connector?

9

The red pin of the old molex harddrive power cables is 5V

10

1092×892 401 KB

11

something about the idea of getting power from a 5v 15A molex unregulated by my motherboard and using the RGB header as a ground only on a $275 motherboard seems....yeah....

Im looking at either powering it directly from said RGB Header by somehow stepping down the voltage (but that probably wont work because as I'm looking at voltage regulators most operate BY using PWM to control voltage so any pwm signal input would be scrubbed) or by finding some way to send the PWM RGB signal to a device at 12v and it outputting the same color at 5v (that's why I thought Arduino forums would be a good place to start)

seriously...what idiot designing a 12v fan looks over at motherboard manufacturers who support 12v rgb strips, case manufacturers with their 12v rgb lighting... and decides....5v

12

What gives you the idea that the 5V from molex are "unregulated"? Do you even know what that means? It is exactly what you would feed into a 1000$ 4TB SSD and also what you already send to your MB. Where do you thing the replaceable fuse sits that protects your MB from overcurrent from those 15A?
The voltage on a molex plug is exactly the 5V that powers your computer. The RGB socket is low side switching, it doesn't matter where + comes from as long as it is common ground and does not exceed the voltage limit.

But sure, put so.e cheap China buck converter with questionable EMC and caps right next to your components unshielded. That is certainly a better idea.

13

Whoah there, you could simply correct me instead of being a dick lol. 5v from PSU's can fluctuate by 5% and mine is 5.16v that an issue? My header is only rated for 2A would that be a factor? Would a reputable buck converter even be able to do it without getting rid of the pwm signal?

14

5v from PSU's can fluctuate by 5% and mine is 5.16v that an issue?

No. It will only run your LEDs on about 3% more current (assuming current limiting resistors), which doesn't matter.

My header is only rated for 2A would that be a factor?

No. The current is completely defined by current and voltage. The amp rating is just a max rating of your supply. You could attach a 5V, 300Gigaampere PSU and everything would be fine (well, until you short something ...)

Would a reputable buck converter even be able to do it without getting rid of the pwm signal?

There is not PWM in the supply. RGB controls with one positive terminal and 3 RGB terminal are low side switching. You supply a constant high voltage and ground is switched on and off. That is why you can take any 5V with common ground to feed it. I do not see any reason to put another regulator in there, but multiple (though not very strong) ones against it.

Your 12V header is probably intended for case modding with LED strips. On those strips, 3 LEDs are in series with a single current limiting resistor. That needs 12V. You fan has probably just one LED per color, so it would be totally useless to supply 12V. It would require either further switching regulation on the PCB or, it would burn a lot of energy.

15

So while the buck converter would do it it would 1. needlessly complicate it 2.Introduce another possible point of failure 3. Be an unneccesary expense and 4. reduce the amount of current I can provide to lights because of converter efficiency

I get that the ground is common, so all components share the same ground so no matter what and where it is all are connected to the same ground in the PSU so the only chance for damage is by shorting or drawing too much current

Feel free to correct me if I'm wrong, I'm here to take in what people have to say and to learn at the very least

So say I'm powering only the red led of an rgb strip and want it at 50% brightness
So low side switching is when the 12v is constant and the ground is switched on and off at 50% meaning 50% of the time its powered at 12v and 50% of the time the ground is disconnected and the circuit is interupted and PWM where 12v would be modulated between 12-0v say 50% powering it at 12v half the time and powered at 0v half the time, while the ground is constant and the circuit is uninterrupted?
So (and I'm guessing here) with PWM a buck converter would be able to "smooth" the input voltage with capacitors to a constant non PWM voltage while with low side switching it would not have an effect since the circuit is completely interrupted?

Are PWM and high side switching the same or just similar?

I think you mean power is defined by current and voltage P=I*V so at 12v 2A my header is only rated for 24 watts so there is the possibility exceeding the wattage WILL burn out the header

kalebg:
My header is only rated for 2A would that be a factor?

ElCaron:
The amp rating is just a max rating of your supply.

My PSU itself (as printed on the side) can supply the 5v rail with 15Amps
Since I'm using the header as ground and not using the power I don't need to worry about the wattage?

I mean worst case I decide to be "safe yet ignorant" and get 3 90% efficiency buck converters in parallel splitting the 2A to 666Ma each and add 3 600Ma fuses to the 12v side so if a converter grounds out or I draw too much current I'm safe

16

If you know the current demand for your LED's you can insert a resistor in series with each LED to "soak up" the 10V difference. Example: if your LED's run at 20mA then do the sum 10 / 0.02 to get the required resistance value which in this instance is 500ohms. You will need to get a resistor with sufficient power handling to stop it from overheating and burning out. Calculate resistor power dissipation using the simple formula I^2 x R (current squared x resistance) e.g. 0.02 x0.02 x 500 = 0.2Watts. What ever dissipation value you end up with, you must select a resistor of at least double that capacity to give a good safety margin.

17

kalebg:
So say I'm powering only the red led of an rgb strip and want it at 50% brightness
So low side switching is when the 12v is constant and the ground is switched on and off at 50% meaning 50% of the time its powered at 12v and 50% of the time the ground is disconnected and the circuit is interupted

Up to here, everything is fine.

kalebg:
and PWM where 12v would be modulated between 12-0v say 50% powering it at 12v half the time and powered at 0v half the time, while the ground is constant and the circuit is uninterrupted?

I don't understand that sentence, but from the following, i think it is wrong.

kalebg:
So (and I'm guessing here) with PWM a buck converter would be able to "smooth" the input voltage with capacitors to a constant non PWM voltage while with low side switching it would not have an effect since the circuit is completely interrupted?

Are PWM and high side switching the same or just similar?

No and no.
High and low side switching is just where you interrupt the circuit: Between + and load, or between - and load. This has no influence on the load. A simple load will not even be able to know how it is switches. PWM will remain.
For mechanical switching, usually the connection to plus is interrupted. This has multiple reasons, one of them is that often, a lot of stuff is at GND/- potential, so cutting of plus decreases the danger of shortcircuits, e.g. with the chassis. Cutting off phase in AC makes touching e.g. a lamp socket safe when switched off.
However, in electronics, there are multiple reasons for low side switching. One is that if you want non inverted drive, you need an NPN/n-channel transistor. An n-channel MOSFET has to compare the gate voltage to ground, so it has to be CONNECTED to ground. At the LOW side, the MOSFET source is connected GND and will have GND potential. At the HIGH side, source has + potential (minus a little coltage drop over the MOSFET), so the gate-source voltage is very small, or even negative. So that type of transistor would not work in that place. A PNP/NPN transistor would work, but had inverted gate logic.
The more important reason, however is, that you can freely choose your supply voltage, because your switching component only has to sink current, not provide it and thus have a certain voltage.

I think you mean power is defined by current and voltage P=I*V so at 12v 2A my header is only rated for 24 watts so there is the possibility exceeding the wattage WILL burn out the header

My PSU itself (as printed on the side) can supply the 5v rail with 15Amps
Since I'm using the header as ground and not using the power I don't need to worry about the wattage?

I don't see why a header should be rated for power. That does not make sense. Sending 10A over the header without any other load will require little voltage and not that much power, but it will probably melt the header. 220V @ .5A which drop somewhere else in the circut and deliver 110W on the other hand will probably be fine (but it is of course not advised to used those headers for mains).
Anyway, those 15A are meaningless, as long as you do not produce a short.

18

I'm just going by the specifications and warnings directly provided to me by the manufacturer regarding the header. One part you were unsure about I was meaning the PWM square wave at 50% duty cycle would alternate between 12v and 0v for equal times and unlike low and high side switching the circuit is not interrupted.

Yeah I was completely unsure about that last part just trying to wrap my head around it lol Thanks for the detailed explanation and taking the time to explain it

19

With 0V PWM switching the switching transistor with the source (N FET) or Emitter (NPN BJT) connected to 0V and the Gate (FET) or Base (BJT) being switched by the PWM controller, the transistor of which ever type is being switched on and off. When the gate or base is at 0V the transistor is off. The drain or collector will then be pulled up to the positive voltage rail by current trying to flow in the load (your LED's). This is an inversion i.e. gate or base at 0V drain or collector at +V.
Now, with a positive voltage on the gate or base the transistor will be turned on, connecting the load to 0V allowing current to flow through the load.
All transistors work in this manner. P Fet's or PNP BJT's work in an identical manner but will have their Source or Emitter connected to +V and the Gate or Base will need to negative with respect to +V in order to turn the device on.
All that PWM means is that the 0V connection to you LED's is switched on and off therefore interrupting the current flow in your LED's.

The reason for a power rating on the header you plan to use is because of a) the cross section of the connector pin and b) the contact resistance between the pin and receptacle.
With a cross section of 0.6mm the conductor area is .36mm and will have a single conductor current rating of about 5A at 20deg C and will incur a small temperature rise at that current. with several conductors bunched together the current rating is reduced.

The contact resistance can be as low as 10 milli oms or up to 100 milli ohms depending on the quality of the connector and how often you make and break the connection. As we all know passing current through a resistor, power is dissipated causing a temperature rise. This is in addition to any temperature rise in the pin and receptacle.

Hope that is a bit clearer.