. m i n . . . i t p i n g .

circuits

Doria Fan - RFIDs

Discrete tags assigned to readers as opposed to bar code system
small radio tower + tags

passive : antenna + circuit (caps charges and discharges electric charges) gets charged as signal is sent to it, based on this, it gets a different radio ID

HF: can read multiple tags and info
LF: gain chip

serial communication/ interface: once up and running, this is all code level not electronical anymore

Theresa - Presence, distance sensor

Posted by min at 03:18 PM | Comments (0)

March 23, 2006

DriveBy Prep　03.23.06

Current P&I preamp Module

Datasheet of K1A6280H

Experimenting with new amps, SSM2019_Self-Contained and TC7660S, Voltage Convertor.

View in IE for ppt
PlasmaSpeaker film cutting instructions
PlasmaSpeaker FilmSpeakerAmpRelated

Posted by min at 05:59 PM | Comments (0)

March 22, 2006

UBi.ach　03.20.06

Our UBi.ach Receiver in action

FlowChart

Posted by min at 03:44 PM | Comments (0)

March 21, 2006

Latest Circuit using the Voltage Divider

Posted by min at 11:56 PM | Comments (0)

March 20, 2006

JYM in Seoul until May

JY is continuing putting together the piece for Daum and will be working in Seoul on his exhibition at Bitforms, Seoul for April.

He will be part of the Bitforms show in June in NYC. More updates to come.

Posted by min at 11:09 AM | Comments (0)

UBi.ach Update

G finally returned from Texas and we met up on Sunday 03.19.06.
From G's blog.

The php file will:
a. parse the new email messages and place them in a database (SQL).
b. place the subject of the message on the clipboard
can use the - GtkEditable::copy_clipboard command:
void copy_clipboard(void);

copy_clipboard() copies the current selection to the clipboard.

It also causes the "copy-clipboard" signal to be emitted.

c. The TTS application will run automatically (from the startup), and take as argument anything that's on the clipboard. Then needs to erase the clipboard.

For the midterm, we will have a stuffed toy with a walkie talkie inside and the sound out will come from the ubiach@gmail.com account through the TTS. More to come.

Posted by min at 10:58 AM | Comments (0)

March 17, 2006

MOMA, Architecture in Spain

Went to the Moma with Fino to catch James Clar's latest. The model was really well made. Probably the most out-there model that could easily distinguished from the rest. I was able to see Zaha Hadid's latest work also.

Posted by min at 10:10 PM | Comments (0)

March 14, 2006

The Latest Circuit

We need to fit everything on one board.

Posted by min at 10:16 PM | Comments (0)

March 11, 2006

UBi.ach Hardware

The plan over break
Details of easy walkie talkies
Gilad: ubiach@gmail.com perl script into clipboard
To check every minute

Tracy and Min: Hacking into walkie talkies
Hack input and output
Output of the computer into the walkie talkie
Output into out speaker

Today, Tracy and I met up at ITP at noon to start our walkietalkie hacking.
Instead of using a MINI ITX, the EasyRadio, we figured out from Todd H that all we'd need was a walkietalkie. This part would be the part that would be needed for the doll to become wireless from the computer.

Went over to the big bad Kmart at Astor Place to find these cheap $20 WalkieTalkies, Model FV200 made by Motorola.

As soon as we opened up the walkietalkies, the first thing T says, "I HATE these switches!"

What we've been able to accomplish today
- hacked the power button on the walkie, 2 pins act as switches
- the audio input from the mic in the walkie #1 gets hooked up into audio in on the laptop, listens to the TTS software and transfers the sound into walkie #2
Diagram

During the SpringBreak week, Tracy and I will be working on the rest of the buttons on the walkie and try to fit in into a stuffed toy just to test it out.

Posted by min at 05:13 PM | Comments (0)

Location and Identification　03.09.06

Class Notes

RFID tag from Texas Instruments

The reader costs about $70.
- pets in NYC are to have these embedded in between their shoulders by law. Recently have become a law
- there are readers that only read certain tags
- can do w. multiple readers to sync with each other. the RFID reader also has a serial connection to receive data
- set to different frequencies so not to let interferrance with other readers (keep them paired up)
- powered RFIDs can travel much further
- passive tags gets a burst of energy when reader comes by it (32bits on the tag)
- semi passive
- active passive : its own power source, more data, reads further
- invasive technology
- look into Sklar, Dorkbot

Mini ITX

GPS
- need a point of reference
- satallite reference. government has control over them
- question in accuracy, not too sure of the location

Biometric
- knowing exactly who a person is (finger prints etc)

Posted by min at 10:04 AM | Comments (0)

March 09, 2006

UBi.ach Revised FlowChart

UBi.ach FlowChart02

Automated gmail check to php checks then brings into the clipboard, into Text to Speech. Sound out.

The Ace High, Text to Speech App reads out whatever is on the clipboard of the computer.

Posted by min at 10:59 AM | Comments (0)

March 08, 2006

PlasmaSpeakers Breakthrough　03.05.06

Our first try of trying to rebuild the preamp module did not work out so well. We attempted at using different capacitors throughout the circuit.

We met up this past Sunday and reworked the Analog Devices SSM2019 Audio Preamp Chip. We were constantly getting crackling sounds from the chip until Chunxi played around with the resistors and got some great sounds out of the audio out. This chip alone can replace the preamp module that Plasma had provided us with. The volume is not as loud, but we can include a transformer to this to get a louder volume.

Posted by min at 09:20 PM | Comments (0)

WiPort　03.04.06

Class Notes

Need to use serial for first WIPORT configuration

Inverting serial, instead of talking to the PIC, talking to the WiPort

Through zterm, hit 0 to connect
Hold down to X key

Gateway IP address : 192.168.1.1

8 bits of the host port netmask

WLAN :
N
Find network name (netobj)? Y
Enable Ad Hoc network ( letting other things to connect as an access point) N
Data rate : set to 6

Power management : radial battery cycle N

Save prefs

Fix the antenna to a certain point that needs to be pointed towards a direction
Try and reload web based interface

192.168.1.102

If one is trying to connect to the wiport, it can only handle one at a time.

Once set up, pic can talk to the WIport using hex invertor

Wiport over Bluetooth. Wiport is easier than Bluetooth. Wiport covers a much larger area. Bluetooth is for personal space areas of about 2ms or so. Less battery power also. Wiport has 300 ft range. Robust connection and range of signal.

MINI ITX boards:

Full blown 17x17 motherboard like the computer. Much more powerful than a PIC. DO need the power to do multimedia.

Mac minis are mini ITXes (nano ITX) can run Pentium, windows

Wireless servers : ITP sandbox
To get access to sandbox, email tom about accessing sandbox.

Then set it up to make it talk to the outside world

With the mini itx, we don’t need one.

MINI ITX

IDE 2.5” 1.8 harddrive, or compact flash (difference like video ipod n nano) power is the issue here- battery usage 1gig CF= 60 bucs

Depends o nthe OS that is used,. Windows xp more than linux.

Have to test it out to see power usage
Mini ITX sites : email and ask Patrick before ordering them

Computer GATE computergate.com
Mini ITX link

800 mhz processor MINI ITX board $100 : EPIA -800
128-258 RAM $20
HD :CF or mini hard drive $20-200 , connector for the Compact Flash
Mini hard drive – adaptor = $10
Power $30 ATX connection power

ATX : Big computer desktop boards
ITX : mini motherboards

Mini ITX: working with the computer, scripting languages are allt here, C C++ java, can run servers on it also, graphical environments as well.

To connect to motors, etc, use serial, usb or connect to an XPORT to talk to eachother. Sensors hooked up into the Xport to talk to the computer to get data in.
Knobs etc. on the XPort

XPort with serial / Ethernet into the mini itx

We have to build our MINI ITXes

For the wiport, need a good fanning/ cooling system- heat sync etc

WIport and Serial RF:

Same range,

Wiport has everything bundled together. No protocol needed, has web access.

RF: buy a transceiver, receiver 2 serial connections going on. Have to do our own protocol.

Posted by min at 09:09 PM | Comments (0)

Building Your Own Circuit Boards, PCBs　03.07.06

Build your own Printed Circuit Boards
- perf boards
- wire wrap : to create connections but pins have to be longe
- battery: mAmps / hour

Printed Circuited Boards
- express PCB: 2-3 days
- soldering padding on holes, so do not have to apply extra solder
- thicker line on a pcb = more current
- can drill extra holes to give yourself more options

Eagle software
- gerber file, simply a bitmap form to print pcbs
- voltage, ground (supply)
- pin head libraries for headers, jumpers etc
- netline to make real connections
- use "eye" tool to check if right connections are made
- name the nets- double click to end net
- wire: only in drawing, not a real connection between components
- route tool: to make connections, links and angles onto each component
-
Manufacturers
4pcb : student pricing
gold phoenix pcb, china : cheapest but don't look through, quite slow
batch pcb: average pricing, late

- put SMD baords into toaster oven to allow solder to sink into place

Posted by min at 02:39 AM | Comments (0)

Space Annotation & Smart Home, Midterm Project　03.02.06

Class Notes

Tristan's friend from Columbia U came to demo a neural controlling interface. Simply freaky. As one would control the buttons on a mac, his arms would twitch without his control. Only about 0.025mAs going through these electrodes, but the layer of the skin with H20 abundant tissue cells allow for the current to move through rapidly, affecting the muscle movements.

Midterm
- concept, theory, execution (actual physcial thing)
- presentation of

Ideas
- Single board computer
- SD chip has 60 sec chips SMD
- TMS :cuts out certain parts of the brain functions to knock out parts of speech, visual field, etc

Spatial Annotation
- wifi systems, gps
- we don't really need to have the tech to get this idea of spatial annotation, ex. street signs, tagged environments around us
- who, what, how is it used?
- creating tags onto spaces (open node), chalked tags on streets (temporal quality)
- visual, audio tags
- podcase : museum audio guidance
- storage= flat
- memory = texture to it
- the idea of "tagging" has changed
- tagging is storage not memory (experience)
- needs the experience and memory
- retrieving into not creating exprience or mem
- bodies moving things that are tagged, not necessarily static tagging
- changing community with a wifi boombox

Domestic Annotations
- smart homes: predicts, manages life at homes, controlled environments
- controlling environments outside of the home and managing appliances
- knowing personal prefs for me
- predicts, plans, manages
- programming randomness within the controlled system
- automated systems: can become absolutes

Posted by min at 01:59 AM | Comments (0)

Op Amp　02.28.06

Dual Supply OpAmp: positive and negative voltages

7905= negative voltage regulator

- usually deal with single supply op amps, rail to rail

- opamp in piezos : small voltage change, very fast

- output: inverts signal and amplifies
- depending on the resistor, changes the amplification
- sensor sensitivity with time
- in opamps, time becomes the issue- will take time from output pin to go into the input pin
- may cancel out some readings

voltage out is 10 times the input

Wheatstone bridge : converts AC to DC
- a set of 2 voltage dividers so you can read both positive and negative voltage
- amplify signals as well as noise

Capacitors: slew rate changes and amplification rate changes (we can't use this for sensors)

- 0-3 Vs for OpAmps
- even when piezso when hit very hard, can produce upto 100Vs
- N4001 diodes : good for power circuits

Posted by min at 01:36 AM | Comments (0)

Datalogging Presentation　02.28.06

Zach
- high #s in reading maybe highByte and lowBytes were reversed

Therese
- getting Q3 value
2 bytes

getting the first half and not getting the latter values at all

- Z term only reads as ascii
- in ADCON1, 2 bytes

ADCON1=%10000010, right justify
ADCON1.7=1 (right justify)
ADCON1.7=0 (left justify)

- look at the processing code
- faster communicatino of 256 data / second for brain wave EEG sensors UDP instead?
- not http: can not do fast refresh

Doria
- processing - PIC : call and response process
ideal sequence of communication
- optimal timing to get both things to turn on at the same time, clear communication

Posted by min at 01:30 AM | Comments (0)

Datalogging　02.25.06

Datalogging
Data archiving, storing

Codes
PIC
Processing
Server side

- when reading multiple values, use a "marker" to make sure that all data came in and in the correct order
- in the processing part, (serialCount>4) for 5 different values, 2 pairs of bytes + marker (255), unique byte

Server end, debugging
- on server, we can do a check with tail
] tail -10/var/log/httpd/access_log

gives last ten lines of file, last ten hits of the server
grep (searching for something on the server)
] tail -1000/ var/ log | grep datalog.txt

- tells us when the hit came etc
- access log to see what hits they were

OR

- create our own log. create an error log separately. errorlog, delete once in a while so that the server does not crash
- listen more (look into more of data received)
- time request should be appropriate to get proper data (human time)
] $rm datalog.txt (to remove the log file to clear up space on server)
- server based back and forth
- ascii encoded: if not numeric, will try to interpret them as control keys instead (ex, next line, tab etc)
- even from the PIC side, use ascii if necessary
- multiple data sending, in processing, "&" to get 2 different values
- in php, "t" for tab

Xport serial pass through directly by ethernet
- PIC has to format "GET string"
- serial (baud rate, stop bit) to ethernet convertor
(Ethernet : IP address gateway, what port, netmask: which part of address should be paid attention to, 0= important address, 255=unimportant)
- http request to itp server
- telnet itp.nyu.edu 80
- ping to see if there are any connections for the server
$ ping.itp.nyu.edu, cntrl C to stop
- Xport: true serial therefore, true 9600. uninverted

SQL server database : structured query language
itp.nyu.edu/phpMyAdmin
create tables and fields

DEC= ascii fomatted decimal

Posted by min at 01:16 AM | Comments (0)

DataSheet Reports 2　02.21.06

Pinky - Tilt Sensor, GP1S036HEZ

- 5 pins
- interrupts. transmittor, photo emittor type
- emittor goes to PIC - output
- collector works with 5V
- testing sensors: do it in the appropriate environment with using multimeter readings
- when the sensor schematic does not show microcontrol- like part, this is more of a digital sensor, switch
- with light, detects then stops (the interrupts)
- binary
- not digital : no analog logic chip
- 2 bit sensor : readings based on the 2 pins
- triggered by light energy

- pull downs so that we can get better readings. otherwise, the pin may be floating
- switches also use a pull down resistor: 10k pull down resistor may be necessary in most cases
- physically, there is a small ball inside that detects the tilt: blocking / interrupting the light
- physical location, setting of the sensor may be required to get a better resolution
- to prevent back voltage, use a diode

Kate: BodyHeat IR sensor
- power, ground, out, reference voltage
- analog sensor
- detects motion of heat?
- senses pressure and distance
- works better in the dark, does not require a light beam
- passive IR sensor
- beam is quite circular

DL coupling: tying things together
ADD: input voltage from 0-5V. when are we getting this? VREF + and VREF -

- high impedance : less charge in frequency
- in parallel connection, ccurrent gets divided: the one with the lower impedance gets the higher current
- PICs: high impedance, therefore consumes low current
- when there is a voltage change = valid signal information
- to get faster readings, we need a lower impedance, max out the sampling rate on the chip
- person would reflect IR, blocking and interrupting the signal
- change in temp and sound waves
- reads only reflected light
- reads whether something has entered a field of sensitivity
- soft/ hard object detector
- ultra sonice detector
- one complex sensor may involve more than one sensor inside the component. a secondary sensor

Posted by min at 12:10 AM | Comments (0)

March 07, 2006

Networking at ITP, Nancy L Workshop　03.03.06

PC: putty as secure shell

the new server is now in place. Some general information on it:
• RedHat Enterprise Edition Application Server 4 (RHEL 4)
• SELinux for more info see this and that
• Apache 2
• PHP 4.3.9 (running in Safe Mode)
• Perl 5.8.5 (runs only in cgi-bin)
• MySQL 4.1.10 : REDHAT 4 allows this

ITP server is hosted by ITS- RED Hat Packages with the versions according.

Php runs as “user apache” – web server user and not user id make the file writeable

In putty, control u deletes the line in the prompt

[gml229@itp blog]$ quota
Disk quotas for user gml229 (uid 793994):
Filesystem blocks quota limit grace files quota limit grace
/dev/mapper/asvg-home
113276 125000 127000 2445 125000 127000

ls –l
control c : cancel out
top – shows what programs are being used right now

[gml229@itp ~]$ ps -U gml229
PID TTY TIME CMD
15243 ? 00:00:00 sshd
15245 pts/10 00:00:00 bash
17408 pts/10 00:00:00 ps

Man : manual
Q : to leave
D : down

Binary works best most of the times

Shows all hits on the site
tail -10/var/log/httpd/ access_log | grep

usage : 200 is successful load and read,

300 or 400(wrong password) is wrong access

Grep ’72.30.133.19’ /var/log/httpd/access_log

Posted by min at 07:42 AM | Comments (0)

March 04, 2006

UBi.ach Gmail Reader

Gilad, Tracy and I have begun to talk about our midterm for our NetworkedObjects class.

Lots have been talked about. We've already had three meetings, simply discussing about our concepts, the technicality that would be involved and the final object that we envision in the end. I have to admit, so far, the group dynamic has been really good. I've always wanted to work with both Tracy and Gilad and this is the most ideal setting. I'm really quite excited.

After much talk about applying RFIDs into a language translator on the floor at ITP, haptic objects that would give biofeedback etc, we've decided to create a "Gmail Reading Doll"
I think we've decided to explore in the lines of "Calm Technology" with emails in our daily lives.
Yes, it is true that this idea has been done before, on a parrot, but we would like to give this "doll" some preferences that would suit the user and give it a lot more character than the parrot.

We had originally thought of this idea of a ubiquitous doll that would read out emails, smses and the things to do in that day, but I think it would be of an enough challenge for us to get one part going. For the miterm, we plan on prototyping a very simple idea, demo and then hopefully carry on into the finals to realize our original concept.

Essentially, these are the underlying ideas behind our UBi.ach.
- whenever an email is sent out to ubiach@gmail.com, the perl script, cron tab will refresh and update the info that has been received. This then gets taken into PHP and the data will be separated and parsed into subject, sender, body etc of the email.
- this all then gets sent into our ITP SQL server and gets stored in a database. there is a system within that separates the important emails from the less important ones and will create a list of important contacts (eg. if my mums were to send me 10 emails per day to remind me to call, the UBi.ach will give me a constant reminder of the message)
- in the hardware side, the PIC, XPORT will communicate with the PHP script and the SQL server to check up for constant incoming email messages, events and reminders etc.
- the data will then be sent to a Text to Speech application and stored as a .wav or .mp3 file on the ITP server
- the sound files come into a flash card inside the doll
- PIC will read and play the sound files

Our first FlowChart for our project.

We had our next meeting again on Mar 3rd. After hearing some options from Patrick D, we have decided to go ahead with a single board computer that he recommended to Gilad.

Posted by min at 08:18 AM | Comments (0)

OutsideClass Collabs

Currently working on a project with Gilad Lotan on a ShadowBox.

Working with Todd, Alice and Tracy on the preamp modules for the SpeakerFilm. We've run into quite a few problems but these were obviously expected. We're experimenting with a few different audio amp chips from Analog Devices.

See links for more info.
Weekly updates.

Posted by min at 08:06 AM | Comments (0)

March 02, 2006

GumDropBlop Final Crit　03.01.06

The Blop in action

Overall, our project came to a good close. Jane and I had worked hard, realized what we had initially envisioned, taking and improving on the points that the class gave us.

First off, before beginning the final version for the midterm, we knew that we could not embed individual superbrights into all of the gumdrops. So we decided to use a white, powerful backlight to get a similar effect. It turned out even better than what we had expected. The glow was even better.

We tried to initially sew on all the gumdrops, thinking that they may fall off as the weight on the spandex white fabric got heavier. We attempted at using fishwires to keep the transparency. However, this became a big mess and we simply hotglued on the gumdrops onto the fabric which worked out just as well with less stickiness.

To continue on with the movement of these gumdrops, we built a stand with servos attached. But when we draped the fabric over, the gumdrop weight was simply too much for the motors to move in.

Overall, the crit went well and people liked the fact that the form was more organic and the subtle movements of these edible material

We went through millions of boxes of DOTS.

So pwetty

Preliminary inside structure of the Blop

Final interior of the Blop

The Blop- ordinary gumdrops turned into an amazing glow.

So far, from the floor, great feedback. People have approaced Todd about our project. the gumdrops began to melt off, so I had to take care of this. We needed to put some cups and a form of ventilation that would let the hot air out of the blop. We also need another light to give it a stronger glow.
Where to go next with this? No idea but I do have a concept that I'd like to work with. More to come after a round up meeting with Miss Oh.

Posted by min at 12:48 AM | Comments (0)

March 01, 2006

First day, AES internship Takes Off　02.28.06

I had fun on my first day of the AES internship. A little too much, since we were trying to cover everything in the time frame of 2.5 hours.
The PXCOMP site can be found here. Look for class plans, blogs and resources on this site.

So the first class. We didn't get to begin class until 30 minutes into the class, only because we had so many housekeeping admin related things to cover. Introduction, set up, flash intro, intro to a little bit of physical computing. We pretty much skimmed through everything, but it was necessary, being the first class.

It was an odd feeling, getting back into the mode of teaching. I've taught teenagers for a few years back in Korea, but I've yet to experience this with adults. The same core principles apply here also. Be patient, listen first before speaking and guide the students that would nurture them in a way that would make them as creative as possible in their learning process.

Admin Info (from our very first internship meeting with Gabriela and Heather a few weeks back)
Start Date : Tuesday, Feb 28th
End Date : Thurs, June 15th (May for us)
Time : Tues and Thurs, 4.30-7.00pm

Internship Requirements
6-8 hrs of commitment per week, and also outside of class

1. Regular Attendance
2. Assit in-class and after class with students : assisting with classwork in class
3. Follow along with lectures in class to stay up to date with topics
4. Assist with setup, breakdown of each class
5. Regularly check-in and respond to blog entries and discussion board topics on the website
6. Attend occasional assessment meetings with instruction team
7. Film portion of class sessions on an rotational basis
8. Prepare one short (10mins) lecture that is relevant to class topics: Mine will be held on March 21st on the PlasmaSpeakers I've been looking into recently- find relevant research (websites, photos etc)

Posted by min at 08:21 AM | Comments (0)

GumDropsWave

Jane and I have continued working with the use of GumDrops. We took everyone else's advice on creating an organic, life-form that would move on its own.

We knew we'd be using a lot more gumdrops. Hence the lighting part became an issue. We decided to use a big white light source instead of individual superbright LEDs in each of them. We used fluorescent light tubings right below our layer of gumdrops to get a similar effect. I still think that the individual superbrights inside them gives a better glow, but with using about 200 gumdrops, this is almost impossible.

Also, we began sewing them onto the fabric that Jane was able to find, but this simply got too messy and took forever. We ended up hot gluing the individual gumdrops onto the surface of the fabric.

We still wanted to keep a subtle movement on the surface of these gumdrops, so we decided to use 3 servo motors with sticks on each end to create this "living" movement. The weight of the gumdrops did become an issue. The servos were simply not strong enough to pulse them outwards for this effect, withholding the heavy weight of all the gumdrops.

Question: What's next?

Posted by min at 07:48 AM | Comments (0)