Looking for LED Driver

Go To Last Post
22 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Greetings folks -

 

I could use an LED driver that drives up to 9 individual LEDs (3 RGB chips, probably). One that pretty much does what I would like is the TI LP55231. It uses I2C interface, uses PWM current control individually, and has a programmable sequencer for different flash patterns. 

 

What I am looking for are alternatives to this device. It does not need to match feature by feature but that has a similar feature set would be really helpful. Low shut-down current is essential. Cost is an important factor. It would be nice to NOT have it in a leadless package.

 

The idea is to drive 9 LEDs, 3 red, 3 green, 3 blue and offload the management from the MCU. Minimum would be 6 channels, 3 red and 3 green.

 

Any suggestions?

 

Thanks

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

Last Edited: Tue. Nov 12, 2019 - 05:53 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Do you need the charge pump?---or do you always have plenty of volts avail to drive the leds?

 

perhaps this (the idle current may be higher than your allowable)

https://www.onsemi.com/pub/Collateral/LV5237JA-D.PDF

 

https://media.digikey.com/pdf/Data%20Sheets/Toshiba%20PDFs/TB62781FNG.pdf

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Tue. Nov 12, 2019 - 06:36 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Jim,

I use a Ti TLC5940, which has 16 channels and is good to 120 mA per channel.  But the clock and management are all done by the MCU.  A little easier chip to use is the TLC5947 has 24 channels good to 30 mA and its own internal clock, so there is not much load on the MCU, but is also probably more than you need.  Both are SSOP, so you have some leads to work with.  East Coast Jim was giving the 5947's away a while ago and has code for driving them.  There are other Ti TLC594's with different number of channels, but they are mostly 16 or 24.  Communication is SPI.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Are they large LEDs? Otherwise you could Neopixel type LEDs.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

LEDs are probably about  0805 package size, depending on whether 1, 2, or 3 diodes are inside. Probably 20ma max rating and will probably run at 10mA or less, depending on their brightness. They are used for occasional status indication for a battery powered instrument. 

 

My hope is  to mount the driver and the LEDs on a little piggyback board, say in the range of 2cm x 2cm to maybe 3cm x 3cm. An SO16 package would be ideal.

 

Supply voltage will be 3.3V so there is not much headroom. I don't care whether the LEDs are on the high side or the low side. For  each "channel", there will be one LED (e.g. no series string) but that diode might be any color.

 

JGM's TLC5947 is DIP32 and would bigger than the board I would like to mount on!

 

Jim   

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I was thinking along the lines of these or similar https://www.adafruit.com/product...

 

5x5mm so you could easily fit at least 4 in a 2x2cm board. No drivers required

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For a quasi-"production" unit, that is awfully expensive! But, your point is well taken. 

 

A lot of the use is with a small set of flash patterns. With those NeoPixel and similar units, all of that still has to be managed by the MCU. I was hoping to find something where i could transfer a pattern to it, then go away.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Jim, I use an RGB led like this : digikey 516-3280-2-ND  in a battery monitor project.

I measure the battery voltage and flash (0.1 sec) different colors depending on what I want to convey about the battery.

with long sleep times and short flashes, I can keep the battery drain to less then 200uA average current.

The blue led is the one that will be hardest to make work with 3.3v, I was using 5v for my battery monitor.

With some tricky programming, you can use one of the LED's to measure ambient light and vary the flash brightness, more in Sun light, less in darkness, helps with battery drain...

Jim

 

Click Link: Get Free Stock: Retire early! PM for strategy

share.robinhood.com/jamesc3274
get $5 free gold/silver https://www.onegold.com/join/713...

 

 

 

 

Last Edited: Tue. Nov 12, 2019 - 09:23 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A lot of the use is with a small set of flash patterns

What is needed as a pattern?  Morse code?  Or just blinking...Is the pattern independent per led , or linked between them (more like a sequence of some sort)? 

 

Is the avail voltage always at least 3.3V minimum?  If you have plenty of volts, you only need a driver, not a booster driver...makes a lot more parts avail.

 

Are you completely against a larger QFN package?...those aren't too hard to work with (compared to some of the microscopic ball types).

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That is the sort of LED I  was thinking of. Both green and blue say 3.1V. It may need a charge pump, which the TI LP55231 I originally referenced does have. The idle mode on that device is 10uA which is a bit large but probably acceptable. But, it only comes in a leadless package which does NOT thrill me.

 

Patterns may include N green flashes from one of the LED packages. Or it might be flash 1x green from first package, 1x green from the second, and 1x green from the third in sequence. "Flash" in this context might be 100ms-200ms on.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

Last Edited: Tue. Nov 12, 2019 - 09:33 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You can get the s2812 (neopixel) from the likes of LCSC for a few cents in smallish quantities. It takes a bit of cpu time for the AVR to talk to neopixels due to the timing. Luckily there is another series of similar devices which use a clock and data arrangement. The part number escapes me but it might be something beginning with APxxxx. Much more relaxed timing.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:
The part number escapes me but it might be something beginning with APxxxx.
Diodes Inc

AL3159 (Low Voltage DC-DC LED Drivers) | Diodes Incorporated (9 LED, 300uA typical quiescent, 1uA max shutdown)

AP3156 - 6 LED, 50uA typical quiescent, 1uA shutdown

AP572x series - 1 LED, boost instead of charge pump, PWM, 500uA quiescent, 1uA max shutdown

via Low Voltage DC-DC LED Drivers | Diodes Incorporated

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For a quasi-"production" unit, that is awfully expensive!

Correct but I wasn't saying to get them from Adafruit for production wink

 

As Russel says

You can get the s2812 (neopixel) from the likes of LCSC for a few cents in smallish quantities.

https://lcsc.com/products/Light-...

 

HOWEVER we may have a problem with the 3.3V supply , never mind a bit more noise.

 

EDIT

Luckily there is another series of similar devices which use a clock and data arrangement

The APA104 and APA102C are of this type as well as the TAX5050xS-00 Series from UBLeds (I have some samples of these) unfortunately they all suffer from the need of a voltage above 3.3V.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

Last Edited: Tue. Nov 12, 2019 - 11:50 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks, folks. 

 

Great suggestions!

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Why not just program some ATtinyxxx as a flasher peripheral?...it can sleep to draw low current when all leds are off, can likely drive 9 reasonable current leds directly.  You can use a 65KHZ timer IRQ to make a 256 Hz 256 level brightness PWM, with each led settable at its own brightness.  Has the advantage in that you can form whatever patterns you desire & tailor them to your needs.  You can even save the patterns in EEPROM ! 

 

You can even add things like ADC to  perhaps change the pattern or brightness based on batt voltage, light sensor levels, etc.  You can even get in an TQFP or SOIC package.

 

I feel like I should start makin & selling 'em!  I used to see "temperature controllers" & "security alarm chips" being advertised & sold that were obviously some preprogrammed PIC chips.

 

 

 

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I got the same suggestion from jgm. I like it, actually.

 

The one drawback I can see compared to the TI LP55231 mentioned in msg #1 is the need for current limit resistors. Well, also the fact that low current blue LEDs can have a forward voltage that is awfully close to the 3.3V supply. That, right there, might be enough to kill it for a 3.3V system. The LP chip has an X1.5 charge pump that raises the LED supply voltage just enough so that their constant current drive system can work.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

 

the fact that low current blue LEDs can have a forward voltage that is awfully close to the 3.3V supply

no worries (other than needing a cap & diode)...give it your PWM & away you go!  You can use diode arrays for multiple blue leds  (if space is an issue)

Note they don't mention Vbb should come from a pin (pin might be common to all blue leds)...if you want to be able to turn off the led...otherwise it would always have at least Vbb volts.

 

As far as resistors---You might be able to get "resistor leds", but I recently found I could find almost none in SMD packages--with extensive searching!  The handful I found were $$$$

Of course, you can easily get an array of 4, 8 ,9, 16 resistors packaged the size of a large ant...use a common gnd (bussed) style

https://www.ctscorp.com/wp-content/uploads/2015/11/752.pdf

 

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Wed. Nov 13, 2019 - 09:05 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The application that I am thinking of does not use PWM. Flash is just that, flash: maybe 250ms on, 5000ms off for a few cycles.

 

But, point taken. I will have to think about that circuit structure.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It uses I2C interface, uses PWM current control individually, and has a programmable sequencer for different flash patterns. 

  The application that I am thinking of does not use PWM

 I thought you wanted to use PWM to set the brightness...anyhow, it certainly is available (IRQ with individual channel software dividers). Since leds require only a low freq PWM (few hundred Hz), it's easy to do. 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ka7ehk wrote:
is awfully close to the 3.3V supply.

what is dictating the vcc 3.3v?  could it run at 3.5 or 3.6v?  

Would that make it easier to drive the LEDs?

 

Jim

 

Click Link: Get Free Stock: Retire early! PM for strategy

share.robinhood.com/jamesc3274
get $5 free gold/silver https://www.onegold.com/join/713...

 

 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Vcc is dictated by an ST accelerometer. But, there may be a way through. Thinking cap deployed.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ki0bk wrote:
With some tricky programming, you can use one of the LED's to measure ambient light and vary the flash brightness, more in Sun light, less in darkness, helps with battery drain...
If the ambient isn't too large then 250bps bidirectional at a few cm (1Kbps at 1m possible) though does have an issue for high humidity [maybe NeverWet (spray silicone) would aid]

Bi-directional LED communication | Tools and Tips | The Embedded Muse 385

...

Bob Paddock submitted this idea for bi-directional comm using just an LED. Part of the secret sauce? Reverse-biasing an LED makes it a photoreceptor:

...

The Embedded Muse 386 | Tools and Tips

...

In the last issue I referenced an interesting article about bi-di communications over an LED by reverse biasing the LED. Siva Govindarajan wrote:

...

 

"Dare to be naïve." - Buckminster Fuller