Can I drive individual channels of a common anode/cathode RGB LED with this?

[sallz0r]

Hi all,

I'm making some LED driving boards for RGB LEDs, and I want to try and drive this RGB LED:   http://www.dealextreme.com/details.dx/sku.4530

with 3 of these ICs (one for each colour):  http://www.supertex.com/pdf/datasheets/HV9918.pdf

I'm pretty sure it can't be done, as it's a common anode (or if you read the comments, someone said that the silkscreen was wrong and it's actually common cathode, so I'm not sure) -- but is there any way to individually drive the channels of that LED with that IC? If not, would it be possible with any other LED driver IC that people know of? (with the emphasis being on 'cheap'... ;-)  )

Thanks for any help and time anyone can give.

[Psi]

I bought some of those exact ones from DX about a year ago.
The comments were correct, the ones i got are mislabelled and are common anode.

However i wouldn't trust that it's the same now, they may have corrected it, or it maybe just random depending what type they have in stock when you order. Get the leds first so you know which type you have.

Looking at that IC i dont think you'll be able to drive common anode or cathod leds from 3 of those chips.
Neither the led anode or led cathode for that controller IC are connected to the supply rails.

The DX LED has a common anode so if you connect one of the leds up to 3 of those ICs they will be all linked at that anode point.
So unless you plan to run each IC from an isolated supply it will create a path where the current through the sense resistor for one IC can supply power to the leds on another IC's. So the IC current sense system will not work properly if at all.

You might be able to redesign the circuit a little so the current sense resistor is on the SW pin thus making the led anode connected directly to vcc.
That would solve the issue, but you'd need to amplify the current sense voltage somehow so it's was what the IC expects.



DX have other RGB leds that have seporate anode and cathodes for each led which makes things much easier.
But they are brighter and so more expensive.
They also have many leds in series internally, so you have to be aware of their large forward voltage, 10-36V etc..

here are some examples

http://www.dealextreme.com/details.dx/sku.44043



http://www.dealextreme.com/details.dx/sku.39960

[Jimmy]

sallz0r

Can you post a description of want you are trying to achive? Are you going to buy more than one of this a want to make 3 of them show independent colours. You could use a darlington array ULN2803 and resistor in series to limit current and voltage or if u are using a microcontrller with a serial output you can buy a TLC5940NT for $3.50 and have 16 output channels all with the same current and you can daisy chain them to give heaps more outputs

[sallz0r]

Thanks Psi -- that's sort of what I was expecting. :-)  hrm, that's OK then, I might have to see if I can find some other drivers that will be happy with common anode LEDs.... hmm.....

[sallz0r]

Hi Jimmy,

Can you post a description of want you are trying to achive? Are you going to buy more than one of this a want to make 3 of them show independent colours. You could use a darlington array ULN2803 and resistor in series to limit current and voltage or if u are using a microcontrller with a serial output you can buy a TLC5940NT for $3.50 and have 16 output channels all with the same current and you can daisy chain them to give heaps more outputs

I'm going to be using many of these RGB LEDs (as well as some of the other higher-power ones DX has), and I'm making common driver boards for them; I would ideally like to use a common board that I can just change the constant current output for the different LEDs that I'm using.

I've looked at some options (like the darlington array and resistor) to do it all manually, however I'd prefer to stick to an IC that does it all-in-one -- just a current source with a PWM input that I can drive. The TLC5940NT looks good, however can only drive up to 120mA -- I would need at least 400-500mA, which is why the HV9918 was appealing.

I mgiht look at some other drivers, though, see what I can find....

Thanks for your ideas!

[Psi]

Just out of interest, where were you going to get the HV9918 IC?
It isn't stocked at RS or digikey.

[sallz0r]

Just out of interest, where were you going to get the HV9918 IC?
It isn't stocked at RS or digikey.

I thought I found it at digikey.... but it must have been mouser, now that I think about it. I'll see if I can find it....

EDIT: aah, yep, here it is: (no idea if this link will work or not): http://au.mouser.com/ProductDetail/Supertex/HV9918K7-G/?qs=WQ7GE9i949B6m2sbw7RjWw%3d%3d

[Psi]

here it is: http://au.mouser.com/ProductDetail/Supertex/HV9918K7-G/?qs=WQ7GE9i949B6m2sbw7RjWw%3d%3d

Thanks

[Jimmy]

Common anode you want to sink the current not source it

[sallz0r]

Aha! this one would probably work, wouldn't it?:

http://www.supertex.com/pdf/datasheets/HV9910B.pdf

I'd need a few more external parts, but it would do the job, I think.... hmm....

[Jimmy]

HV9910B looks like an expensive way to drive a fet to me

[sallz0r]

HV9910B looks like an expensive way to drive a fet to me

Hmm... what would you suggest?

[Psi]

yes the HV9910B should work fine, it has positive going to the led string, so you can use a common anode rgb led with 3 of the IC's

It also has a PWM input for current control, which might be handy.
It doesnt appear to have an analog current control input tho, like that first IC had.

I also like the fact that that IC can handel 8V to 450V dc on its input pin due to an internal regulator.
It makes it good for led room lighting projects where you might want to run the chip from mains voltage.

[sallz0r]

Ack! I've just realized a flaw in my plan! :-)

Some of the lights I'm planning on using are common cathode, which means using something like the HV9910B doesn't work (as it appears to be a constant current sink), which doesn't let me have a common cathode for the RGB LEDs. In fact, pretty much every single LED driver IC I can find is a current sink (or at least, has the current sensing resistor or other components on the cathode side of the LED).

Does anyone have any idea if there's any LED driver ICs that would support a common cathode for multiple LEDs, acting as a current source? Or failing that, can anyone point me in the direction of what I would need to do to design something like that myself?

Thanks!

[arcom]

You can easily design your own driver circuit using mosfets and LM317 configured as current source. You will need one LM317 and one P-channel (or even N-channel) mosfet for each color. Efficiency will be low but it works well. Another option would be to use current limiting resistors instead of LM317.

As a last resort you can always get 6-pin RGB LEDs with separate anode/cathode for each color and use some driver ICs.

[Jimmy]

http://www.electronics-tutorials.ws/transistor/tran_7.html

Are you going to make these bords with a vision of being able to sell them?

It is cheeper to sink current than source it