LED array sequencer

[MrWild]

I made a circuit with discreet componants that works, but I have far too many parts to make a compact system. I don't have enough knowledge on PICs, but they seem like the ticket. I'd be willing to pay for code and circuit that works for my simple needs.

The particulars are:

six LED arrays times two for a total of twelve (12) arrays.

Each array has five LEDs.

power is 12v.

first input causes five arrays of each group to light

Second input takes precedence over the first and causes one group of five arrays to sequence. first time period #1 on, second time period #1 and #2 on, etc. until all five arrays are alight then off for three time periods.

The sixth array is on for four time periods and off for four when the second input is activated.

Third input does the same thing only to the second group of LEDs.

I need the package to fit on a 1" x 4" PCB.

Any ideas, thoughts, or suggestions?

[One of Many]

That looks like 60 LED's?

You might look around this site for LED Array hints and solutions that could apply to your needs.
Discover Circuits

There are some high drive current PIC's, but most will limit the outputs to 20-30mA each, if not in total. That would require external components like BCD 7 segment display drivers to handle the current and utilize 4 BCD coded lines of PIC outputs to control 7 LED's.

It would help to post a diagram of the LED array layout. Then a diagram of each LED output condition expected in the array based on the inputs in the previous desription. Whether they be banks of LED's, all lit, sequenced or individually controlled. I couldn't make sense of the written description for lack of a visual reference in what it should look like when functioning.

My limited experience is with Parallax products, as in Basic Stamps and their SX48 chips.

DC

[Al_The_Man]

For compactness you may want to look into 10 element LED bargraph display in a DIP package.
Al.

[gar]

061119-1627 EST USA

MrWild:

If you multiplex the LEDs you can do this with an Intel 8279 and peripheral circuits. The same concept can be done with a micro.

When you multiplex LEDs an illuminated LED is pulsed at high current for a short time at a high repetition rate, above the eye flicker frequency.

If the normal steady-state current was 10 ma, then if the pulse on time is 1/16 the drive current might be 160 ma or higher.

Multiplexing reduces the number of wires and drivers to the LEDs.

.

[pastera]

Do all five leds in a group come on at the same time or do they sequence on?
If you need to sequence individual LEDs, I don't think you will have the room on a 4 sqr-inch board for all the connections needed.

If you are not in a hurry, I could help you out with a layout and some code but I am busy for the next few weeks.

How are your soldering skills? The best way to make this fit is with small pitch SMT parts that take a good steady hand to solder.

Aaron

[Mariss Freimanis]

Ideas? Yes; multiplex them. That means 12 columns, 5 rows. Only 17 led drivers (5 + 12) to manage 60 leds.

Mariss

[lerman]

Originally Posted by Mariss Freimanis Ideas? Yes; multiplex them. That means 12 columns, 5 rows. Only 17 led drivers (5 + 12) to manage 60 leds. Mariss

Well, since it is all being done in software, you might as well go all the way and use 8 logical rows and 8 logical columns. Save a driver compared with Mariss's scheme and be able to manage 64 leds. Also, lot's of drivers come in groups of eight. And lot's of processor ports are 8 bits wide.

Ken

[gar]

061120-1716 EST USA

As Mariss suggested the 5 x 12 matrix is probably the better approach because the layout of LEDs probably falls in this arrangement, and also the logical maping in memory is easier to work with if this is the case.

.

[pminmo]

If there are patterns of LEDs use a matrix as suggested above in conjunction with a constant current sink that is switched.

[MrWild]

Thanks and WOW, totally lost. Not a bad thing if I eventually learn something. What my idea is for, is to create custom turn signals with brake lights incorporated into them for motorcycle use. Many of the custom high brow bikes only leave their turn signals on long enough to pass inspection. Once through the DMV/DOT mess and their bike is titled, off come the ugly, only there for the moment, signals.

I've created some housings that look good enough to leave on after inspections, but the electronics are just way too cluttered when doing it with discrete componants. Because of how fast ideas are stolen and production sent over seas by some of the unscrupulous aftermarket folks, I was leary of looking for help in a public forum. My health is deteriorating though so I decided to bite the bullet and try to get it done while I can.

For this design, the rear signal lights have five arrays of five LEDs. Three of the arrays in each signal will light the center LED as a running light (six total). For brake lights, all the arrays will light. For turn signals, the right or left signal will sequence. If the brake is applied at the same time a signal is operating, the signal takes presadence over the brake light function.

For the front signals I have rear view mirrors with 4-5 (in design phase) LEDs that are centered in the mirror housing.

My circuit works, but in order to hide the electronics in as small a package as possible for uncluttered wiring, I have only the spaces (.625 tall X .980 wide X 4.5" long) in the rear signal housing(s) for the PCB(s). My current design has a PCB in each housing taking care of it's side. Even then it is silly crowded. My circuit uses a 555 timer, decade counter, transistors, diodes, and resistors along with a few capacitors for the 555. For the brake light/signal I'm using a 12v DPDT selenoid and a 7805 for the low voltage requirement of the ICs. Very crowded circuit board with many of the resistors and diodes standing on end.

Another design I'm working on has hidden signals in the stop/running light that motor out when the signal is switched on. Lots of ideas, but my health is going to heck in a hand basket. Docs now say I need the Gamma Knife or Cyber Knife for neuromas (tumors) pressing on my brain stem. This is on top of other problems. I'd like to try to get this all done this winter in case the sh!t hits the fan, so instead of learning all about PICs and the code writing to program them, I decided to ask for help and scratch up the coin to pay for a working PIC design. Don't expect to get rich, but don't think I'm asking for a hand out either. I just want to finish a bike I've spent ten years slowly building. No not a HD. Those were sold long ago to cover med expenses.

[Al_The_Man]

What became of your AcroMill?
Al.

[MrWild]

Sold it at a loss to a guy in Omaha Nebraska. I listed it on eBay and told the truth of the mill's many problems. Because of that, I only recieved one bid and it sold for just $1,000. At least I can look at myself in the mirror without cringing and sleep a peaceful sleep knowing I'm not scum like the bum that sold it to me.

The new Supermax has a much smaller memory and will only load one program of just 1,000 lines at a time, but it is much more robust and has no mechanical problems while holding very tight tolerences. It also has a working auto drawbar and air operated brake. I'm almost done with an older Dell to use for drip feeding through the RS232 port. I'm much happier and haven't had a hiccup with the new machine at all so far. It really makes the old BP look like the junk it was. I do miss the acroloop though and came close to gutting the controls and selling the mill itself as scrap.

The new standing wheelchair and the supermax seem to go hand in hand. Life is better.