Friday, December 14, 2007

Pong, Monski Pong and Failing Flex Sensors

Made some progress, debugged the Monski Pong program, basically a whole lot of typos from typing it in (copied from the text in the book). The ball runs slow but I think I can get it to go faster by editing the code. maybe my computer is too slow. Had some issues with the flex sensor which I will describe below in more detail. Am setting up for project 2 "Wireless Monski Pong". Will solder a connector (Coaxial DC PowerPlug 5.5mm O.D, 2.1 mm I.D. RadioShack PN 274-1569A) that I had in the parts drawer that happens to fit the Arduino battery connector well enough, it sticks out a little more than it should.

Set up the blue tooth dongle on my PC, a Planet BT-410U. I got it a while ago and it was not setup on my computer, I could not find the install cd and usually that is not a problem if the manufacturer has drivers on their site, planet did not as far as I could tell. Fortunately I found the cd and was able to do an install and test the blue tooth with my Motorola H500 headset. Worked good!

I was thinking about putting a cannibalized PC serial connector on the board to use a set of old atari 2600 paddles to replace the flex sensors. I will probably just do it since the paddles will be neat to use with the arduino anyway. Plus they have a couple buttons, and I could also use an atari 2600 joystick, which is really 4 buttons for the joystick (plus the actual button).






I had a problem with one of the flex sensors, it was acting like it wasn't there. I did the usual "wiggle the connection" to try to get it to work, and indeed it did work for a few minutes. I took a look under the stereo microscope, fully expecting the traces to be cracked or something. They were fine, what I did see was that the metal connector was not really connecting. The connector is attached with barbs that poke through the plastic material. On the surface of the plastic is a silver trace that is a conducting layer. The barbs were mostly not touching it. If you look at the pictures of the barbs, you can see the plastic film is extruded along with the barb making a mostly metal to insulating film contact, not the desired barb to conductive silver contact. I think that the metal barbed part was installed on the incorrect side, it would probably be much better if it had been installed on the silver trace side. In order to repair I used some "Circuit Writer" which is a polymer paint with silver conducting particles in it. So far it works ok according the the ohm meter. I will let it dry overnight under a lamp to keep it warm. Hopefully it will be a good permanent repair.

4 comments:

tigoe said...

Yeah, that's the weakest thing about those flex sensors. It really helps to provide some stress relief on the end of them by the contacts too, it tends to buckle and break. I try not to solder, but to use wire wrap on these sensors these days too,

BTW, if you don't want to type in the code, it's all online at http://www.makingthingstalk.com


-tigoe

matt holland said...

i had the same problem with one of my two flex sensors. since i don't have any polymer paint handy, i abandoned the flex sensor idea and used trimpots instead. trimpot pong is probably almost as fun.

also, do you know how to get around the jittery paddle problem? why do the analog ins values vary when nothing is changing? i tried using moving averages, and also ignoring values within certain tolerances, but found tradeoffs with each. what's the best solution?

robo_t said...

The only problem is where I can purchase a pink fluffy monkey (aka: monski) i can't find it anywhere!

Clark said...

Thank you so much for posting this great article, I am so interested on what I read, good and useful content about "Pong, Monski Pong and Failing Flex Sensors", I would like to invite you to take a look to order viagra, you will find some good stuff there.