144 LED Games Console Project

[mtripoli]

If enough people are interested I can layout a PCB for this and upload the schematic and Gerber files. You could then send the files to your favorite PCB house for manufacture, or if someone wants to make a run of boards you're all set. Though I'd use SMD personally I'd think a through hole board would be easier for first-timers to solder. It would be a two sided-plated though hole board. I can knock this out in a day or so...

Give me the word and I'll go ahead. I would say though add any switches, connectors, headers etc. ideas before I do it (layout is no big deal, rip-up and re-route sucks)...

Mike Tripoli

[Art]

Give me the word and I'll go ahead.

If you're waiting for the word from me, then go ahead
If SMD makes the LEDs cheaper that's something to think about.
I paid about $30 Au in LEDs per unit, and they were fairly ordinary LEDs.

The only consideration I can think of is the joystick buttons should be
arranged like a joystick, and remember you have to turn the unit vertical to play tetris,
so I don't think the joystick can be ideally located for both games.
I've posted an image of how I'd try to lay it out now so the joystick could be
used for both games without wiring it seperately.

It also might be an idea to add the EEPROM as it is in the first schematic.
If you're using a 40 pin DIP for the 16F877 then the ICSP header is optional.

In the schematic I have given, two of the direction buttons switch pins to gnd,
and the other two switch pins to +5V. This could be changed to make
the buttons all uniform (and the software altered), or left as is.
Cheers, Art.

[mtripoli]

I'm glad I was sitting... I know things cost more depending on where you live, but this surprised me... I don't know what you have available for buying components there but they are nowhere near that much here, buying from an "expensive" distributor (Digikey). Here's a link for red 5mm TH parts: http://search.digikey.com/scripts/Dk...me=160-1678-ND. As you can see 100 pcs. is $8.60USD ($9.60AUD) A comparable SMD (http://search.digikey.com/scripts/Dk...=160-1463-1-ND) is $13.00/100. This is a 1210 package; I don't know that I'd go smaller than that. Again, for beginners SMD would be hard enough to solder, let alone smaller. Smaller devices "typically" don't have a round lens as well.

The "joystick" buttons are not a problem. I didn't see how many pins are used on the uC. How about a couple of pads for a rolling ball tilt switch (one pin on the uC)? These are about $1.50 here (so that'd be $6.00 there) If you tilt the board 90 degrees the joystick pad (and display) can "rearrange" to the correct orientation?

Resonators with built in caps are cheaper (about $0.60) than a crystal.

It looks like any NPN will do. 2N5210 parts are about $0.15USD. For whatever reason, "BC" parts are slightly more than "2N" parts here.

I'll start drawing the schematic... I have a fresh martini in front of me so I'll be on cruise control... the package type can wait until more people have weighed in...

Mike Tripoli

[mtripoli]

I've looked at the footprints for the components. I've wanted to do this in the past, but frankly when you are getting paid to do something you can't screw around. I'm going to design the PCB so that it can take both through hole parts or SMD parts in the same footprint. What you care to populate is up to you...maix and match! Theoretically it *shouldn't* take any more board space than one or the other...

What was your spacing on the leds and do you think it optimum? The "aspect" ratio obviously effects how the display looks. It *looks* like 0.100" centers to me...

Mike Tripoli

[Art]

Sorry, I was away for two nights. Lucky I took my laptop and pic programmer with
me so I saved a lot of chip space again

I don't know what my LED spacing is. As long as it's even, it should work.

The EEPROM in the circuit def shouldn't be needed now, but I don't think a tilt
switch is needed either. The orientation of the joystick is determined by the game you're
playing, so it can be switched with the game.

One of the pins that was connected to the EEPROM should now be used for a
serial port I think.
I was using one of the LED input ports, but without the EEPROM, the serial ort could have it's own pin.
The other EEPROM pin could now perhaps be connected to a PCB speaker.

[Art]

Hi again Guys,
I've done a few updates including support for a 250 character display message
for the scroller. This was done by making space for tables in program memory again.
Now the only thing other than a message on the on-chip EEPROM are game high scores and message length.

Removal of the external 12C EEPROM chip provided extra pins.
The serial port now has a dedicated pin, and a PCB speaker has been added.

Both games now have basic sound, and the unit clicks as you type characters into your terminal program
when using serial message programming mode.

I thought I'd do more with this one, but I've moved on to another pic project.

Cheers, Art.

[Art]

Thought I'd better post that there is an error in the LED matrix schematic.
The 18 transistors should all be connected to + 5 Volts, not Gnd.
Cheers, Art.

[Mike, K8LH]

..... I paid about $30 Au in LEDs per unit, and they were fairly ordinary LEDs.

Hi Art,

Nice project!

I know you're pretty far along in the design so please forgive me for jumping in with suggestions at such a late date.

Have you considered using some of the bargain "surplus" or "new old stock" 8x8 matrix displays out there? You could probably replace $30 worth of discrete LEDs with three 8x8 modules (8x24 = 192 LEDs) for less then $5. Use SIL (single in-line) machined pin headers for module sockets and you can utilize the space under the matrix modules for other circuitry (notice how much room I have under those large 2.2-inch modules in the pictures below).

Have you considered scanning rows instead of columns? This would increase the duty cycle from 1/18th to 1/8th which should improve brightness by increasing "average" LED current. Using "direct" I/O drive on the rows is limiting you to about 1.3-ma "average" current with your 1/18th duty cycle. That would go up to about 3-ma "average" current with a 1/8th duty cycle.

Kind regards, Mike

[Art]

Have you considered scanning rows instead of columns? This would increase the duty cycle from 1/18th to 1/8th which should improve brightness by increasing "average" LED current.

That would be nice, but I'd then have to use 8 transistors for the rows.
Bit late for this hardware, but if someone wanted to add that to a PCB design
I could certainly update the software.

The size of the display was really determined by the proto PCB I was using,
and it had to be a certain height to scroll letters of the LCD font I wanted.

[Mike, K8LH]

Hi Art,

Yes, you'd need 8 PNP or P-FET row driver transistors but then you could get rid of the 18 column driver transistors --> fewer transistors, more current limiting resistors on the columns now instead of the rows, and more than double the "average" LED current (and brightness).

If you've already committed to a PCB design then I would agree it's a bit late for this hardware change, but then again, if you're designing a new PCB then why would you limit yourself to the design that just fits on the prototyping board that you happened to have?

Here's my (unsolicited) idea for a modular system that will support a single 8x8 module or up to eight daisy-chained 8x8 modules. Start by designing a single 2.1"x2.1" PCB for those large 2.2-inch 8x8 matrix displays which can be used either as a "master" 8x8 display module (with cpu) or as a daisy-chained 8x8 display module. Install a PIC, column driver IC, and row driver transistors on the first 2.1"x2.1" PCB for a stand-alone "master" 8x8 display module. Install just the column driver IC on an additional 2.1"x2.1" board and daisy chain this new 8x8 module to the first module for an 8x16 display. Likewise, add additional modules as desired. The modules connect via a simple 10 pin powered buss (plus +5V and GND lines) and the 1/8th (12.5%) duty cycle remains unchanged wether you're using a single 8x8 module or all eight 8x8 modules.

In summary, a single relatively inexpensive 2.1"x2.1" PCB can be used as the basic building block for a high performance full brightness mono-color display of 8x8 up to 8x64 in size.

Kind regards, Mike

[mtripoli]

Hey Art,

I got your message. Clear some of your mailbox so I can reply. The system is saying you've exceeded your inbox amount.

Mike Tripoli

[mtripoli]

Hi Art,

Yes, you'd need 8 PNP or P-FET row driver transistors but...
Kind regards, Mike

Hi Mike, Mike here...

I agree with you 100%. However, I think *part* of the appeal of this approach is it's "brute-force". I'm in fact a big fan of the "smart drivers" from TI (and others, see my previous post) and have designed systems much as you've described for commercial use. However, I think that users, beginners in particular, are more comfortable with an "all-in-one" board like this.

It does have it's shortcomings but that's not really the point. When you look "under-the-hood" there are a lot of basic lessons (no pun intended). This board grows (by an additional 143 leds) on the beginning "blink an led" ideas. Other than generating sound (not so easy) flashing leds is probably up there with the first things that people want to do when starting out with microcontrollers (other than those that have never touched a soldering iron, never done any programming of any kind, buy a "system" and then want to "design" a USB based robot control system with 100 servo's, motion feedback, PID, voice control and image recognition. With one PIC. In Basic. Canned routines. Yes, that is a dig and you know who you are).

I volunteered to layout this PCB not because I am a nice guy (I am actually, just ask my dog), but because contrary to what some people think I do believe in helping out people that genuinely want to learn and see this is as a platform that would (hopefully) make them pull out a soldering iron, buy some parts and go through the process and learn something. My previous post about "not doing rip-up" is out the window awhile ago...

Mike Tripoli

[Mike, K8LH]

Hi Mike,

I realize now that Art is pretty deeply entrenched in this design (seeing several copies of the same prototype on his web page) and so I suspect I am clearly out of line by suggesting that he switch to scanning rows instead of columns to improve display performance and brightness. That change BTW would not require column driver ICs.

The more advanced modular design with sinking column driver ICs that I suggested was intended to stimulate the synapses but is probably out of place here and so I apologize...

Kind regards, Mike

My previous post about "not doing rip-up" is out the window awhile ago...

This went over my head Mike. Did I miss something Sir?

[mtripoli]

Ok, after way too many emails from way too many people I've posted the files again. Here they are: http://www.scarydesign.com/144led/144LEDGERBERS.rar. Get them, save them, repost them somewhere else; this will not last forever.

To the dude that sent me a personal email asking me to not only repost the files, but explained in broken russian/english that the design didn't suit him and he wanted it changed to support USB with new code and would I get "right on it" YOU ARE THE REASON I PULLED THE FILES. All info is there to have a PRO PCB HOUSE make the boards; these were not designed for kitchen sink laser print transfer so forget complaining that the traces are too thin etc. and if you haven't figured out that a wire soldered on the top and bottom of a drilled hole is the same as a plated though hole, I just don't know what to tell you. Happy Halloween, Thanksgiving, Merry Christmas, Super Rosh Hashanah Kwanzaa and grandma's birthday. Please stop sending me emails asking for a favor "just for you". I do love to do "favors"; my rate starts at $75.00/hr. with a 10 hour minimum.

[funnydrummer]

Ok, after way too many emails from way too many people I've posted the files again. Here they are: http://www.scarydesign.com/144led/144LEDGERBERS.rar. Get them, save them, repost them somewhere else; this will not last forever.

To the dude that sent me a personal email asking me to not only repost the files, but explained in broken russian/english that the design didn't suit him and he wanted it changed to support USB with new code and would I get "right on it" YOU ARE THE REASON I PULLED THE FILES. All info is there to have a PRO PCB HOUSE make the boards; these were not designed for kitchen sink laser print transfer so forget complaining that the traces are too thin etc. and if you haven't figured out that a wire soldered on the top and bottom of a drilled hole is the same as a plated though hole, I just don't know what to tell you. Happy Halloween, Thanksgiving, Merry Christmas, Super Rosh Hashanah Kwanzaa and grandma's birthday. Please stop sending me emails asking for a favor "just for you". I do love to do "favors"; my rate starts at $75.00/hr. with a 10 hour minimum.

Thanks a lot mate.

In attachments you'll find my version of the project, enjoy!

[funnydrummer]

can someone tell me how to edit my post?
I've edited schematic, and here it is http://www.2shared.com/file/EZ6gNZQ5/SCH.html , can someone tell me where to connect that R19?
If someone can edit that schematic, I would be greatfull.

I used R19 for an extra power LED, which you haven't connected yet. There were some errors in that schematic.
I suggest you download the latest version in the post above... ;-)

[mtripoli]

thanks, but I have 17years, and in school we are making SCH and PCB with tango, and I won't buy displayes, I buyed 144 LED diodes.
Can you just connect that R19 on my schematic that I uploaded, and tell me if I have connected LED diodes correctly, because I edited your schematic

Your schematic looks a little odd for the displays. As shown it looks like the leds being used are offset in different directions so if the they are physically pushed together there would be rows and columns not lit. The resistor you are so worried about (R19) originates from one of the "d-pad" switches. These switches are pulled low and are active high as drawn in the schematic I did. I believe Art changed the code to reflect that. In his original code he had it as drawn with two pulled high and two pulled low for no reason whatsoever. Why you have another low value (120 ohm) resistor tied here I can't imagine. It looks to me like you simply copied the schematic wrong. Look at the original schematic in the RAR file I posted. This is not a difficult project. You've admitted "editing" the schematic; don't make changes to a circuit if you don't understand the circuit in the first place and then demand that others fix it for you or explain it to you.

[FernandoA]

hello, art
well I am new to this and would like to know how to connect the pulsating since I have not clear yet, you could help me. And this other your work is great.

[Pichai]

Cool Project