August 2005 Entries

My laptop detects the board ... sweet.

I also got some 'assorted LEDs' from Radioshack for the power / TX / RX LED spots so I can tell if everything is peachy on the boards.

2.50 per box, got 4 just in case ... really only needed 3

Lets see if my c++ skillz are where they use to be and lets see if I can port this code over to the eviler side of computing ... c#.net and winAmp.

I did do a little winamp plugin app back in the day for the betabrite LED sign but that was straight C++.

All is good, all is well. the first board I tried didn't like me which tells me it has a bad solder joint some where. Flashed 2 other boards just fine ... the only other two boards that are assembled.

HELP

I have no clue what I'm doing nor where to start. STK 500 is what I have to program the firmware. Even after reading the manual ( I feel dirty for reading a manual ), still no clue.

Anyone that can tell me how to properly hook this up into the board, set up the STK500, and have it program would be my bestest friend in the world and if they are in Chicago, you'll be hooked up with some serious beers / drink of choice. Can comment or just email me at clint@rutkas.com

67 of those LED SOB's are completed.

Now on to figuring out how an AVR programmer works. Not going to lie, this thing looks kinda freaky.

Currently have 29 built. I've figured out a good assembly line method for building these. Do one bend on on a ton of them, then move on to the next bend. Testing to be sure the LED isn't fun. They are so bright for a solid five minutes afterwards I still have spots in my eyes.

I'm not totally sure if I'll cut down the leads or leave them as is. Not even sure i'll solder them in since the sockets do a very good job of holding the LED in.

using 22 gauge wire, 1.5" exposed on each wire in a looping grid pattern, with 8 layers of plastic wrap, I successfully built a force sensitive resistor. Freaking sweet.

Using the same grid and wire, I built one using the foam. Functional and works.

I'm not really sure which will be better in the long run. The plastic wrap will run the risk of becoming damaged by extended use. The foam sensor is a just bigger since the foam is 1/4" in height.

I still haven't attempted to daisy chain them together also to test out my baseball diamond idea.

I have 14 but the white stripes are getting threats from the rock gods and I must be present to see the outcome.

You can however check out the directory here. The names are fairly user friendly.

Here is my new favorite picture ever.

Purchasing will be on hold for a bit. Spending about 1150 on LED's isn't chances are a wise decision.

After a quickie test with the default packs, it produces a phatty white light. Pics to follow.

Still need to whip a test sensor.

Emailed Grant couple days ago.

I actually built a few sensors using conductive foam before I stumbled upon the saran wrap solution (which yields a smaller sensor, costs less, and is easier to build). We sandwiched about 6-8 layers of saran wrap with copper wire bent back and forth to form a square of copper about 1 cm^2. The resistance between the wires drops about 5 orders of magnitude (from about a MOhm to about 10 Ohm) when you step on it. This is quite a hack job and probably not the most reliable solution, but it is dirt cheap.

Best of luck. Your cables look amazing, by the way.

- Grant

My test sensor was just using a wire, betting the increased surface area of a grid will make it actually ... you know ... work. Now that I have foam, may try both with a grid. I need to get finer wire however. Looks like a trip to radio shack for me on the way home. Now only if they didn't have such retarded employees.

LEDs along with my Jameco order came in. Looks like I may just have a module done this weekend if my old resistor packs will work. Breaking 1,450 SIPP sockets in half will suck.

Also the anti-static foam is totally different from jameco than it was from Radioshack. Bouncer and not as see through. May just work as the lego site claimed with this foam.

Looks like I get to test out LEDs though and see what resistor packs I need to buy. Time for major soldering after work.

ordered some stuff from jameco. Going to give their static foam a chance, maybe it is better than the radioshack stuff. Watch it come before my LED's. I may just start soldering the SIPPs on before the resistors to keep up progress.

SOCKET,SIPP,10PIN,MACHINE
TOOLED PINS,SOLDERTAIL Jameco#:102200 Qty: 1450 $261.00
@ HEADER,.1"ST MALE,1RW,2PIN,
(10) .025"PST,.21"GOLDTAIL Jameco#:108337 Qty: 10 $1.60
@ HEADER,.1"ST MALE,1RW,3PIN,
(10) .025"PST,.23"GOLDTAIL Jameco#:109575 Qty: 10 $1.70
@ HEADER,.1"ST MALE,1RW,4PIN,
(10) .025" PST,.23" GOLDTAIL Jameco#:117559 Qty: 10 $1.70
FOAM,CONDUCTIVE ANTI-STATIC,
24"x12"x1/4" Jameco#:13864 Qty: 1 $9.29

Subtotal of Items: $275.29
Shipping: $11.34

Total for this Order: $286.63

Attempted to use the static pad I got from radioshack and mimic the sensor from the lego page. For the list of me, I couldn't get it to work.

Attempt 2 using a plier wrapped with electrical tape I can get a drop of resistance. I'm not happy with the foam's ability to jump back however. I'm not sure that it actually work due to this.


Some phatty PSU's. One is still at work. Biking with that much weight hurts.


resistor hack for the PS-On pin to ground.


My PSU connector clipped.


What a semi unassembled power routing block looks like.


The connector assembled and ready for some rocking.


fully assmebled power routing block.


4 of the 10 boards that are as finished as possible

To turn on a ATX power supply without a power supply isn't too hard but still a ghetto hack. All I need to do is PS-ON to a ground. I opted for the less amount of power feedback.

For my Kingwin 350W power supplies, I need a 1.5k Resistor.
For the In Win 240W power supply, I needed a 3.3k Resistor.

Since I'm stupid and don't really know how to calculate resistor values, I used this site (google rules).

I'm not totally sure if I'm going to solder them directly on in the PSU or figure out something else. I doubt I'll ever reuse these power supplies so this may not be a bad idea just to do a direct solder.

Something I just found out, a power supply won't turn on unless it is hooked into a computer. Now to figure out how to turn this on.

the issue of the day. The power supplies will be in today. I want to drink this weekend. Can I do both?

Power supply hookin will be all of 30 minutes max. Between work and the bar, I think it can be done ... unless I just go straight to the bar. Then wait until tomorrow.

I'll post some detailed pics on how I'm going to actually wire this up with a nice image on how each power block works done via Illustrator (If I'm lazy, MSPaint).

My one of my bosses at Sonoma Partners, the company I work for doing c# for MS CRM, found out I was building this monstrosity during lunch today. He claimed in a year period I won't get 3 girls to dance on the floor.

I beg to differ. I claim 5 girls. 1 being the token 'larger than normal' girl but still, 5 girls. While many may call me a girl, I dance like Bill Cosby so I don't count.

after being forced to watch baseball, I think a diamond shaped pressure sensor would be a good idea for total accuracy for being on the tile. One pad in hte middle of each side. Since the sensor is basically a switch, you need all 4 shut before it will turn on.

Who knew watching baseball would give me a good idea.

Now to fight with PDF generation code at work. Come on c# and .Net ... who's loves you like a red headed step child.

$30.00 - 2 primary for the main board link (+5V@30A)
$19.99 - 1 for the display in my room / testing link (+5V@20A)

$2.19 - 4 power Adapters link

I got the adapters so I won't destroy the power supplies.

Total after shipping is $122.46 (shipping was 33.71 for 3 day air ... it was only 3 dollars more than balls ass slow)

I've gotten as much as I can right now completed. The LEDs won't be in for another 2 weeks.

In the lull of not being able to build, I'm going to start planning out on bar napkins on the calculations for building the floor itself. Along with an improvised pressure sensor. The lull is happening in the middle of a hell week at work so I doubt I'd do much work anyways.

Legos may provide a solution. I knew having a RCX would come in handy. Their solution (at the bottom of the page), is simple and appears to be effective. MIT did the same thing. It would just be a matter of creating the nooch not too deep, not too shallow. If too shallow, the weight of the Lexan will trigger it. Too deep and won't get triggered. Here is the foam they used. Still need to see if it is cheaper on Digikey.

Speaking of Lexan, I still need to figure out that too. This is the last remaining major cost in the build ... other than the LEDs. I would count the power tools too but I'll use those for more than this.

Changing the Power, I2C, Serial parts from a SIPP socket to a ISP header.
.100" Male Headers (Gold)
ISP headers 2 - 108337CJ - .16 cents per
ISP headers 3 - 109575CJ- 0.17 cents per
ISP headers 4 - 117559CJ- 0.17 cents per
total cost will be five dollars, figure buy an extra just in case of each.
http://www.jameco.com/Jameco/catalogs/c253/P119.pdf

SIPP socks will be needing 2 10 length PER ribbon. This will mean at a minimum I'll need 1280 (64 spots on a board). In this quantity they are 18 cents per.

Figuring I'll buy 1400 just in case some of the breaks I dislike. Looks like I'll be spending about 250 dollars on just SIPP sockets.

As I'll need some shrink wrap too. From digikey, I was looking at 1/16" colored tube from Digikey. I picked up a 3/32 and just slightly too big in my eyes. a 1/16 will bring it to 1/32. They do have 3/64 but that is took small I think from the start. Figure I'll use about 5 and a half feet per board for each LED color. So total I'll need lets say 60 feet to be safe. Total I'll use 600 feet of shrink wrap tubing ... damn. I can get RGB and figure use yellow to the 5Volt line. Maybe white, maybe black. Per color it will be about 16 dollars from digikey. So what is 4 times 16 children? 64

And since i was dumb ass and screwed up so many damn female power plugs ... I need to buy 5 (going to say 10 for safety).

Now to figure out how to build a ghetto pressure sensor that can be hidden in the wood.

test LEDs won't be shipped until 9/7. This will make my Halloween attempted deadline a bit hard. Not just the fact it gives me a reduced building time but also to make the actual full order if I liked them. Plus I still need to buy the resistor packs ....

Guess I can always have a Disco Christmas Party. But damn it, I was hoping to be dancin with a kilt on.

Gives me some time to get my power supplies / figure out how to build the frames / build frames / buy tools / build cables / buy replacement (and additional) parts / solder on the the QPOS's on the remaining 5 boards.

Bought some RGB LEDs for testing. I got 70 for the 'learnin' board. If I dislike them for the main board, I'll still use them for the board that is going on my wall in my room.

http://www.ledtronics.com/ds/dis1024/dsdc0334.pdf

I completed one power block also. I messed up a rather large sum of the male power connectors before I figured out how to do it easy. More so, having multiple types of solder came in handy for doing this. I have a 60 tin / 40 lead solder that dominated. I'll post some pictures on how I did it soon as I get replacements ordered for the ones I messed up.

Things I learned also, twisting wire by hand sucks. Use a drill. (in my case, the power screw driver my mom gave me for christmas). Now starting my DIY for wire twisting.


clamp one side (both wires) down


connect the other end of both wires to the drill


turn on the drill and twist til you're happy.

and here is one of my two "power blocks" completed

I haven't updated costs in a bit even though I've bought additional stuff.

Namely a wire stripper, the 22 gauge wire (red and black), 16 gauge wire (orange and black), 2 terminal blocks, some terminal block connectors, ummm, a PCB for testing, a AC to DC adapter for testing ... and I think that is all.

Soon to buy three 200W PSU's from newegg. Two for the main floor. Interesting thing I noticed, the 5 Volt outs on the PSU's have an Amperage limit. Good thing I noticed that.

This one has +5V @ 28A so with the 4 amp per board, I could power 7 boards ... but I'm only going to power 5 per PSU.

Why not buy a massive PSU to power everything? Bigger PSU's produce more noise and heat. duh. Plus to pull the Amps needed, it seemed as if I'd need a massive 600 watt PSU. This way is cheaper too.

One of the two power blocks I have. Each will be wired into the PSU from a computer supply. Each block will need 20 Amps going in to it.

Each board needs 5V and 4 Amps. So I need 20 Watts per board (5V * 4 Amps = 20 Watts ). Not totally sure if my baord will be 8 or 9, but figure I'll need 160 to 180 Watts worth of power.

So it has been a bit since I've posted but now that i have internet access at my apartment and stuff isn't as hectic at work, I can do a little bit more.

I've created 1 of my ghetto DC power splitters. I'll whip up a nice photoshop version of it. I really should rewire it to it is a bit more pretty and user friendly with color coded wires. Plan on using a computer power supply for this also.

On other news, i think I may have found some cheap RGB led's that will work with the board. 1.80 for each with the possiblity of a volume discount. 60 degree viewing angle claims a full color range. Going to see if I can order 10 for testing and then if I like them, go for the mass order. This will also let me figure out the resistor packs. This will be far cheaper than if I have to buy each red / green / blue as separate LEDs. I'll link it soon as I hear back from them.

I have 5 boards pretty much ready to go soon as I solve the LED issue.

Well, looks like tomorrow between 1 and 5pm central, comcast should roll in with my internet access again.

about damn time. been down for a week.