SparkFun Forums 

Where electronics enthusiasts find answers.

Have questions about a SparkFun product or board? This is the place to be.
By manton
#56347
You can use the individual cathodes to balance the brightness of each LED (or turn them on/off individually) , or you can do what I did and just strap them all together to make life easy. With a supply of 3.5v or less you can just connect the +ve to the anode and the ground across all the cathodes. The LEDs are spec'd for 15mA each, so this is the figure you would use to calculate the series resistor if your voltage is more than 3.3v. If you connect all 6 LEDs together then don't forget to reduce the resistor by a factor of six to compensate (or use 15*6=90mA as the total current for doing the calculation in the first place). You don't have to run them flat out at 15mA if you don't need to, they'll be visible from probably 5mA on up.
Hi James,

Unless there are resistors in series with each LED on the module, you should not connect all of the cathodes together. The LED forward voltage will not be perfectly matched between all of the LEDs. So, if all LEDs are in parallel, they will not be running at the same current level. The only proper way to handle this (for long lifespans), is to put a current limiting resistor in series with each LED.

Running LEDs at a constant voltage, without a current limiting resistor is also a very bad design practice, as there is no way to predict what current they will see. They are meant to be run in constant current mode.

Mike
By mikeselectricstuff
#56348
Connecting LEDs of exactly the same type is generally not problem - any small mismatch will be covered by the LEDs' slope resistance so current imbalance will be insignificant.
By manton
#56352
While this may be ok for hobby use, it would be considered a bad design methodology for a commercial design. Unless you can guarantee that all LEDs came from the same die, their forward voltages with respect to current can vary widely. I see this done all the time in products that are made in China, where high failure rates are very common, and this is the primary reason.

Most backlights that I have seen that use a parallel LED combination, also have at least some current limiting resistors for each set of LEDs.

The same rules apply to putting multiple diodes in parallel to accomodate higher currents. They don't share the current well, unless something is done to accomodate the forward voltage mismatch.

Mike
By JamesK
#56353
manton wrote:While this may be ok for hobby use, it would be considered a bad design methodology for a commercial design. Unless you can guarantee that all LEDs came from the same die, their forward voltages with respect to current can vary widely. I see this done all the time in products that are made in China, where high failure rates are very common, and this is the primary reason.

Most backlights that I have seen that use a parallel LED combination, also have at least some current limiting resistors for each set of LEDs.

The same rules apply to putting multiple diodes in parallel to accomodate higher currents. They don't share the current well, unless something is done to accomodate the forward voltage mismatch.

Mike
Hi Mike,

I take your point, and I would be fascinated to know what series resistance you would recommend I put with these 3.5v rated LEDs when driving from my 3.3 volt supply?

Regards,
James.
By manton
#56354
Well, generally I wouldn't drive them directly from a 3.3V supply. Instead I would use one of the many LED boost circuits that are meant for this purpose. It is just lucky that you can get them to work reliably on 3.3V at all. Even a small resistor would help them share the current better. Try 10 or 20 ohms. Have you measured the actual current through each LED?

This is a bit like asking how one would operate a 5V logic chip on 3.3V supply. It doesn't, without boosting the supply into an appropriate range first.

Good luck,

Mike
User avatar
By bigglez
#56357
JamesK wrote:I would be fascinated to know what series resistance you would recommend I put with these 3.5v rated LEDs when driving from my 3.3 volt supply?
That is either a trick question or a lack of accurate data. What do you mean by "3.5v rated LEDs"?
Last edited by bigglez on Thu Oct 02, 2008 6:42 pm, edited 1 time in total.
User avatar
By bigglez
#56358
manton wrote:While this may be ok for hobby use, it would be considered a bad design methodology for a commercial design. Unless you can guarantee that all LEDs came from the same die, their forward voltages with respect to current can vary widely. I see this done all the time in products that are made in China, where high failure rates are very common, and this is the primary reason.
I thought is was just me...
Not only low end "disposable" products but also hobby
construction articles where the design was not vetted.
I've written to one technical editor recently, but to no
avail (so far - magazine cycles are quite long).
I know of one article that did this, was kitted by a
third party, and failed to work well as multiple
randomly-picked LEDs were installed in parallel.

May I quote your post? Seems to cover all the LED
related issues well.
By JamesK
#56359
bigglez wrote:
JamesK wrote:I would be fascinated to know what series resistance you would recommend I put with these 3.5v rated LEDs when driving from my 3.3 volt supply?
That is either a trick question or a lack of accurate data. What you mean by "3.5v rated LEDs"?
I was genuinely interested in the answer, but the question may have contained a fair amount of 'rope'. I thought Mike responded rather well under the circumstances :)
The "QVGA-Datasheets.pdf" on the SF sales page specifies the LED forward voltage as 3.5V. The leds operate dimly (given there's an LCD on top) from about 2.7V, and are respectably bright at 3.3V. Since I have a regulated 3.3V supply to hand for the controller it was convenient to simply apply that across the LEDs, which is what you've seen in all my screenshots. And I agree with Mike, up to a point. I have my hobbyist hat on and I'm playing with a $1 surplus part - quick and dirty is the order of the day. If I was doing it for real I'd probably include a constant current supply that I could PWM for brightness. I suspect I hit on a pet annoyance, and it was reasonable to pick me up on it.

James.
By manton
#56360
bigglez wrote:
manton wrote:While this may be ok for hobby use, it would be considered a bad design methodology for a commercial design. Unless you can guarantee that all LEDs came from the same die, their forward voltages with respect to current can vary widely. I see this done all the time in products that are made in China, where high failure rates are very common, and this is the primary reason.
I thought is was just me...
Not only low end "disposable" products but also hobby
construction articles where the design was not vetted.
I've written to one technical editor recently, but to no
avail (so far - magazine cycles are quite long).
I know of one article that did this, was kitted by a
third party, and failed to work well as multiple
randomly-picked LEDs were installed in parallel.

May I quote your post? Seems to cover all the LED
related issues well.
Yes you may quote my post.

I'm glad to see that it isn't just me, as I was starting to wonder as well.

The problem as I see it, is that for hobby use, you can get away with a lot, and make a one off work. But this doesn't scale well to commercial products with larger production quantities. With so many engineering types starting to rely on hobby projects for their education, there are some bad practices that are being used. Alas, the individuals have not learned the proper ways to do things and simply do not know any better.

I realize that this forum will be predominantly used by hobbyists, but it is important to understand that there are more correct methods that would apply in a non-hobbyist project. But, I can certainly understand why a hobbyist may make different choices, that are less than ideal.

Mike
User avatar
By bigglez
#56368
JamesK wrote:
bigglez wrote:That is either a trick question or a lack of accurate data. What do you mean by "3.5v rated LEDs"?
I was genuinely interested in the answer, but the question may have contained a fair amount of 'rope'. I thought Mike responded rather well under the circumstances
I was testing the waters, not anyone personally. You might
have been misinformed about LEDs in general, and yes,
it's a pet peeve of mine.
JamesK wrote:I have my hobbyist hat on and I'm playing with a $1 surplus part - quick and dirty is the order of the day. If I was doing it for real I'd probably include a constant current supply that I could PWM for brightness.
Okay. Perhaps you have a higher voltage within reach, even if its
not regulated (guessing that your 3v3 comes form somewhere)?
If you were a newbie (and clearly you are not) I'd have tried to
get you to run from higher voltage and add ballast resistors.
By blboucher
#56421
Thanks for the overwhelming response from everyone! I understand the cautions and mentions of it being bad practice to connect LEDs in parallel rather than in series with individual resistors if at all possible – but then why is it we're dealing with a commercial OEM part such as this with such a fundamental electronics "flaw"?

Maybe that's why they went for $1 each in the first place or maybe the LEDs were matched closely enough for reliable use being powered in parallel?? Any ideas?

So since in this case, it's really not practical to break the thing open just to put a resistor onto each LED, is it safe to say that for "hobby" use connecting it with the proper voltage resist and current limit and run all together from the single + line in would yield a relatively decent lifespan?
User avatar
By bigglez
#56422
blboucher wrote: Maybe that's why they went for $1 each in the first place or maybe the LEDs were matched closely enough for reliable use being powered in parallel?? Any ideas?
If the LEDs in a given unit are from the same wafer their
performance is probably well enough matched to be
directly parallel connected.

In a high volume OEM module there plenty of incentive
to remove components (such as ballast resistors) from
the BOM.
blboucher wrote:So since in this case, it's really not practical to break the thing open just to put a resistor onto each LED, is it safe to say that for "hobby" use connecting it with the proper voltage resist and current limit and run all together from the single + line in would yield a relatively decent lifespan?
Yes, assume the LEDs are matched and capable
of current sharing. Choose a single external resistor
to feed the module, and expect the LEDs to be evenly
lit (and of normal long life).
By JamesK
#56431
blboucher wrote:but then why is it we're dealing with a commercial OEM part such as this with such a fundamental electronics "flaw"?
Hi, there's no flaw in the design of the part. It's normal for the anode to be shared, the key is that the cathodes have separate pin outs on the backlight connector. The current of each LED can be individually controlled as easily through the cathodes as through the anode. A full commercial design would typically use a dedicated multi-channel LED controller like this:

http://www.allegromicro.com/en/Products ... 4/8434.pdf

These are incredibly neat circuits that handle the dc-dc conversion to uprate the supply, multiple channels of constant current sink and an easy to use digital dimmer control. Nerd paradise :)

James.
By JamesK
#56737
Ah well, I think I have to admit defeat on driving the LCD at full speed from the pic32. I got close, and learned a lot in the process. Generating sync was easy, but I can't get in and out of an interrupt routine fast enough to generate the RGB data stream. I tried using DMA to drive the data through the parallel port, using the configurable wait states to set the data rate to 5 MHz. It sort of worked, but the timing seems to be bizarrely unpredictable. This isn't quite the best display I managed, but it shows the problems with lost pixels, and mismatched lines:

Image

It's annoyingly close to working, but I think I could spend a lot of time fiddling and never get this much better. So, time to rip out a load of code, put back the interrupt driven output routine and find a compromise dot clock I can work with!

James.
By Skylark
#56983
I've figured out how to drive the small display over SPI.

Try http://unaligned.org/lcd128x96.c - this is for an FTDI FT232R with the display connected: SCL to TXD, SDA to RXD, CSN to RTS. This source is for Linux using libftdi.

Enjoy :D
  • 1
  • 3
  • 4
  • 5
  • 6
  • 7