LED Matrix Display

I'm having a problem with an LED matrix that I built from Radio Shack LEDs. The matrix has 3 rows by 3 columns, where each column shares a common power and each row shares a common ground.

G G G P P P | | | | | | | | | | | | | | | | | | | | |--------G------LED(B)--LED(Y)--LED(R) | | | | | | |--------------G------LED(B)--LED(Y)--LED(R) | | | | |--------------------G------LED(B)--LED(Y)--LED(R)

In the first column are identical blue LEDs that are 20 mA and 3.2-3.8 volts (350 mcd). In the second column are identical yellow LEDs pulling 20 mA, 1.9-2.4 volts (80 mcd). In column three are identical red LEDs that have the same specs as those in column two. When I built this, I had to use what I had lying around. ;) I knew that there would be a difference in brightness between the blue ones and the rest of them.

When three identical LEDs (the columns) come on together, there is no difference in brightness. But when I light them by row, something strange happens. When I light a blue one by itself, it's okay. When I light a yellow one next to a blue one that is already lit, the blue one dims slightly. And when I add a red LED in column three next to a blue and yellow that are already lit, the blue one dims to practically nothing.

As I said earlier, this does not happen when I light the columns. Nor does it happen when I light all nine of them.

I thought that there wouldn't be a problem in running them in parallel because they all have the same current specs. But apparently the voltage difference affects the current? Can somebody explain to me what is happening, and tell me if there is a way to fix it (without replacing all of the LEDs with identical ones)? Thanks.

Just thought of one possible explanation… lighting the other two is causing a voltage drop to less than the 3.2 volts required for the blue LEDs??

Looks to me like you're running them in series, not parallel (which is fine).

lighting the other two is causing a voltage drop to less than the 3.2 volts required for the blue LEDs

Sounds about right to me :).

yes

maybe. do you have resistors somewhere to limit the current draw? if not, try 150 ohms on the column pins.

You will still see a bit of a dimming for more leds lit but not too too bad.

This actually is a parallel configuration, and not series. Each row of LEDs is wired in parallel to ground. Each column is wired in parallel to power. By wiring them this way, you can individually control nine LEDs with only six wires. Try it! To turn on any one LED, turn on its power and its ground.

I actually have this made on a LilyPad soft circuit sewn with conductive thread. The LilyPad puts out 5 volts. The conductive thread has built-in resistance of about 14 Ohms per foot, so I didn't add any---for now it's just a test circuit that I'm playing with.

A voltage drop could be the explanation for the dimming, but then I thought... why don't they dim when they're all on? As far as I recall, they don't. (I'll have to put that programming back in and try it again, just to make sure.) Maybe there's some property that changes when the ones in the same column are all lit, such that whatever happens when only one row is lit does not occur??

Resistors?

This actually is a parallel configuration, and not series. Each row of LEDs is wired in parallel to ground. Each column is wired in parallel to power. By wiring them this way, you can individually control nine LEDs with only six wires. Try it! To turn on any one LED, turn on its power and its ground.

Facepalm

No idea what I was thinking ;D!

It's okay. We all have those moments :D.