Does anyone know of a FET module capable of driving a common cathode RGB LED using 12 VDC.
I recently got a 7FET but it works with common anode - yes I know that common anode is more popular.
I'm exploring the possibilities - although replacing the LED's is way way way down on the list of desirable activities.
Can anyone suggest a board or components - I believe some sort of P channel Mosfet will do the job but I'm no electronics whiz.
I need to drive 4 RGB LED's, using single channels and wiring each and every LED all the way back to the Arduino (a Mega 2560) is 'silly' and not happening - and 5V can only drive one LED per circuit - 12VDC can drive all 4.
You'll need NPN or N-channel MOSFET to drive the gate - it has to go to 12V to turn off, while the gate needs to be pulled low to turn on.
In that case you can use a Standard MOSFET, it doesn't need a logic level gate.
NTD2955 would be okay, and less costly
Another option is mic2981 to source current to the LEDs.
"The MIC2981/82 is an 8-channel, high-voltage, high-current
source driver array ideal for switching high-power loads from
logic-level TTL, CMOS, or PMOS control signals."
Can you explain this "5V can only drive one LED per circuit - 12VDC can drive all 4."
You can't wire RGB LEDs in series like you can discrete LEDs.
As I calculate it given a forward voltage of 3.4 VDC that leaves no headroom so each and every LED would need a 5VDC wire - instead of one per pair (green and blue forward voltage) and one for all 4 reds (2.1 VDC forward) - maybe I'm calculating it wrong - like I said I'm no electronics whiz.
PPS - I haven't wired them in series but I have +VE on 2 blues via a 270 ohm, 2 greens via another 270 ohm, repeated for the other 2 blue and green, combined 4 reds via a single 180 ohm - well nearer 190 since I paralleled a 470 and a 330 ohm since I didn't have 180 ohm available. There are three 12V +VE - Red, Green, Blue and one -VE that goes to all LED's. Kind of hard to explain but it works just fine for RED, GREEN, BLUE or WHITE - but I now need a way to apply PWM to the +VE's so that I can mix the colours,
Sorry if I'm rambling a bit - it works but knowing my luck I did it totally wrong ...
?? - there are 4 individual RGB LED's, all are common cathode - that's -ve to the long pin, surely the diode won't work as common anode. I can switch all 4 to Red, Green, Blue and White (all on), what I can't do is control them individually because the FET module I have expects to do things the other way - common anode.
I proved the theory by connecting the LED negative - all 4 leds share the same ground (common cathode) and PWM works fine on my FET module but it affects ALL the colours in each LED - so I know it wants to 'sink' the voltage.
How do I wire a common cathode LED as a common anode ?
I've been reading all the data sheets for the recommended devices above and haven't a clue where to start - all part of the learning process I guess.
BenKenobi:
I have +VE on 2 blues via a 270 ohm, 2 greens via another 270 ohm, repeated for the other 2 blue and green, combined 4 reds via a single 180 ohm
That part confused me. It sounded like you had wired the reds, blues and greens in series with each other, which you could only do if they were discrete leds, or RGB leds with 6 terminals.
If you have wired them in parrallel with each other but only used a single series resistor, then that is a bad idea, you should have a resistor in each separate current path.
Perhaps you should draw us a diagram. Hand-drawn and scanned or photographed will be fine.
I already use this board to provide dimming on two channels to some single colour LED's but it can't handle my common cathode RGB's as independent channels - had my LED's been common anode it would have been fine - but they aren't and changing them out is way too much hassle where they are fitted.
Again, you are confusing me. Your diagram shows 12 separate leds, but you say they are common cathode. If you cannot draw an accurate diagram of your circuit, it may be because you are not clear about it in your own mind, and could wind up damaging your components as a result.
If your leds are genuinely common cathode, you should have separate series resistors for each anode. 12 in total for your diagram above.
Also I am almost certain you have calculated your series resistors incorrectly. Red leds have a lower forward voltage than green & blue, so your red series resistors should be a larger value than the green/blue ones, not the other way around. I think you will damage your leds like this.
Have you calculated your series resistors as though the leds were wired in series, whereas they are in fact wired in parrallel?
I will assume the forward voltage of the red leds is 2.0V and the green and blue are 3.5V, and that the max current is 30mA for all 3 colours.
For red, series resistor required = (12 - 2.0) / 0.03 =333R (use 330R)
For green & blue, series resistor required = (12 - 3.5) / 0.03 = 283R (use 300R)
A 12V supply is very wasteful for this circuit. Most of the power will be wasted as heat in the series
I'm not so sure that I'm the one confused, I showed 12 separate because it was easier than drawing a 'combined' RGB package and avoids any need for lines to be crossing. Each coloured box represents a colour within a single LED, 4 coloured boxes per colour = 4 LED's
The lower resistance on the RED is because there are 4 'diodes'
So for RED :-
Forward voltage = 2.1 volts, 'If' = 20ma
the 180 ohm resistor dissipates 72 mW
together, the diodes dissipate 168 mW
total power dissipated by the array is 240 mW
If I did it your way using a separate resistor for each I'd need a 560ohm on every one and be dissipating 260mW per LED - this is a waste of components and energy.
for BLUE and GREEN
Forward voltage = 3.4 volts, 'If' = 20ma
each 270 ohm resistor dissipates 108 mW
together, all resistors dissipate 216 mW
together, the diodes dissipate 272 mW
total power dissipated by the array is 488 mW
If I did it your way using a separate resistor for each I'd need a 470ohm on every one and be dissipating 256mW per LED colour again this is a waste of components and energy.
...which assumes you are free to wire up your leds in series. You don't have that freedom. You can only wire your common cathode leds up in parrallel.
For example I assume you got this result:
Solution 0: 4 x 1 array uses 4 LEDs exactly
+----|>|----|>|----|>|----|>|---/\/\/----+ R = 180 ohms
The wizard says: In solution 0:
each 180 ohm resistor dissipates 72 mW
the wizard thinks ¼W resistors are fine for your application
together, all resistors dissipate 72 mW
together, the diodes dissipate 168 mW
total power dissipated by the array is 240 mW
the array draws current of 20 mA from the source.
Please explain why I can only wire in parallel, my resistors are on the positive side not the negative, all LED's share a common -VE, each gets its voltage requirements via an appropriately sized resistor.
I see no logical reason for treating an RGB LED any differently to 3 logically / physically separate LED's.
Even if I had common anode I'd do the same thing except my resistors would be in the -VE side of the circuit.
What your suggesting to me currently makes absolutely no sense, granted done your way I could use 5VDC but I'd need significantly more resistors and connections.
If you think I don't get it you're correct, I see nothing logically wrong with my circuit, I need the 'justification', I want to know and understand why I'm incorrect - assuming that I am.
Its not your leds you need to hack out, its those resistors. And you need to put more resistors in. Can you un-solder the leds and re-use them? Sounds like they may be ok (which surprises me a bit).
And you can pwm common-cathode leds. Crossroads suggested some components earlier. However, they may have been overkill in terms of current capacity.
You only need to switch 80mA per channel. Too much for an Arduino output, but an ordinary pnp transistor like bc327 can do that easilly.
Question is do you want to stick with 12V supply or change to 5V?
If you stick with 12V, you will need some npn transistors (like bc337) to drive the pnp ones (otherwise the Arduino pins will get exposed to the 12V supply).
If you switch to 5V, the Arduino pins can drive the pnps directly.
This is what you needed for the common cathode RGB LEDs.
Paralleling the resistors for same color LEDs is not ideal, but does work.
Single PNP per color with 5V as mentioned, or P-channel MOSFET with NPN to allow 12V on the gate to turn the P-channel off, and NPN that arduino can drive.
I have just Dremel'd the LED's out - they were epoxied in and not much else was going to remove them. Where they were fitted - and how - made resistor changing impossible - that and I'd created a harness with shrink sleeve etc - no space to work where they were. A days work in the toilet but that's all part of the learning curve I guess. I can salvage the resistors though.
I've just ordered some common anode RGB - just doing a new drawing, I'll post it in a bit for 'scrutiny'.
I could use either 5 or 12 VDC, doesn't really matter now.
I'll be using the 7FET module to handle the switching - it can handle up to 1A ... I won't be any where near that.
As far as the component bit goes CrossRoads I'm still keen to explore the PNP / NPN - N Channel / P Channel FET thing but I need to learn how to build a circuit to use them - the data sheets I find unreadable mostly. D S and G might as well be Greek although I figured that the load goes onto S and the switch is done on G - what I need to figure out is how you know whether a high on G or a low does the business ... I also read that PNP is a little slow for PWM - is this true ?
Standard PWM frequency is 490 Hz or 980 Hz - P-channel MOSFETs can handle that.
If you use logic level N-channel FETs, then all need is this for a design.
This N-channel FET works fine when driven from Ardiuno
I use 32 of them as LED strip drivers (or other high current loads) on this board.
BenKenobi:
A days work in the toilet but that's all part of the learning curve I guess. I can salvage the resistors though.
The moral of the story is built it on breadboard first, before you solder/glue anything.
BenKenobi:
I could use either 5 or 12 VDC, doesn't really matter now.
Go for 5V then, keep it simpler.
BenKenobi:
... I also read that PNP is a little slow for PWM - is this true ?
No. Ordinary pnps will be orders of magnitude faster than needed for pwm. pwm is slow - only a few tens to a few hundreds of hertz, enough to fool the eye. Most ordinary transistors can switch up into the megahertz range.
