August 20, 2009
Purell
Here’s what we did for the Purell sanitizers (model 2720-01).
We desoldered the AA battery packs from the motes and set them aside. Instead we made a circuit with a battery that parasites from the 3 C cell batteries in the Purell.
There are two wire pairs connected to the Purell. The red and black wires go through a hole near the dispenser outlet, avoiding the moving portion, and into the black box. Avoid the cam, which moves, and solder to the battery terminals. The other pair of wires (yellow and blue) go to the motor through a hole which is drilled out of the motor case top. The other end of these wires goes to the 3-pin connector on our circuit board (vl parasite 3.1 on the server). The center pin of the connector goes to the positive side of the motor.
The circuit board was made in the electronics shop. There were two versions, but we think all the current Purells have the version marked 3.1 on the circuit board.
These components go on the circuit board:
1. Female MTA socket with wires and header .1 spacing, 2 pin
2. Female MTA socket with wires and header .1 spacing, 3 pin
3. Optocoupler Isolator Phototransistor 50 MA CTGR-20% DIP-6.
4. 330 ohm resistor
5. 2 resistors (resistance marked in the eagle circuit board file)
5. VL2020 horizontal mounted battery (BATTERY LITH COIN 3V 20MM HORIZ – about $5 from digikey).
6. 4 wires to connect with the mote. (Note: It would be nice to have some connector here so the mote could be detached from the Purell.)
7. IC Linear Regulator Voltage 3.3V, 300mA, (LDO TO-92?)
Two 4-40, 3/4″ screws connect the board to the Purell below the push button.
There are 4 wires from the board to the mote. Assuming that the board is mounted in the purell with the connectors on the right, towards the center of the device, up towards the soap dispenser, we’re labelling the wires from the top down on the left side of the board.
Top is ground (near the voltage regulator)
Next down is power.
3rd down is analog input (a voltage divider so the mote can read the voltage).
Bottom is user button interrupt.
The mote tucks into the case on the left.
Avant Design
Here’s the details for what we did for the Avant hand sanitizer.
We soldered a pair of wires to the motors. To do that we unscrewed the middle section of the box and soldered the wires onto pin headers that we soldered to the green PCB. We routed routed the wires through holes drilled through the separating platform between the lower case and the area holding the soap.
We separated the batteries from the mote and mounted both on the lower front of the inside of the box, facing the user — batteries on one side, mote on the other. We had to drill some holes to make a path for the wire.
These 4 wires — two from the motor and two from the batteries were routed to the circuit boards we made. The batteries go to a 2-prong connector (connector set MTA socket with wires and header .1 spacing, 2 pin), which we got from the electronics shop. The wires from the motor got to a three-prong connector (also connector set MTA socket with wires and header .1 spacing, 3 pin), also from the electronics shop, but we cut one of the wires off, so that one connector would be 2-pin, one 3-pin and they wouldn’t be accidentally plugged into the wrong place. We actually routed one set of wires from the solder points close to the boards, then soldered the wires to the appropriate trimmed wires coming from the connectors.
The eagle file for the circuit board is in the project file on the lab server. The parts that go on the circuit board are:
1. Female MTA socket with wires and header .1 spacing, 2 pin
2. Female MTA socket with wires and header .1 spacing, 3 pin
3. Optocoupler Isolator Phototransistor 50 MA CTGR-20% DIP-6.
4. 330 ohm resistor
5. A 2×5 male pin set with .1″ spacing
6. A 2×3 male pin set with .1″ spacing
The board was made in the electronics shop. The pin sets were designed to plug into matching female pin sets soldered directly onto the mote.
Prelim Testing
We’ve set up stands at fixed distances and are working out a few basic parameters, moving the motes a distance of 1 or 2 m at different power levels. The radios are positioned battery down with radios facing, on top of the 2.5″ PVC pipe stands. Channel 25.
Power level:
Power 1 m 2m 0.5m
Level mean std dev mean std dev mean std dev
15 46 1.8 52 1.8
10 45 2.5 46 2.5
20 55 4.0 52 3.0 54 7.0
25
We are now using a WI-SPY to monitor the signals operating around our lab in order to minimize interference with our experiment. Our motes are operating on channel 25, which has a frequency of 2485 MHz. Since the WI-SPY is showing a lot of noise on channel 25 and minor noise on channel 1, we are going to change our motes to operate on channel 1 which has a frequency of 2405 MHz.
Once again, the radios are positioned battery down with radios facing, on top of the 2.5" PVC pipe stands. This time, new batteries are being employed.
For our next experiment we are going to vary the orientation of the motes atop the PVC pipe at a constant distance of 1.5m and power level 20
Radio Spectrums
We’re checking for noisy signals with the wi-spy software v1.2. This says that the device covers the spectrum from
Technical Specs:
* Antenna * Internal PCB Trace Antenna
* Bandwidth * 2400 to 2482 MHz
* Frequency Resolution * 1 MHz
* Amplitude Range * -97 dBm to -50.5 dBm
* Amplitude Resolution * 1.5 dBm
* Sweep Time * 120 msec
It lists 14 channels (above 11 is in parentheses).
The technical specifications from crossbow explain that there is overlap between the IEEE 802.11.4 (wi-fi) and the motes.
The motes use IEEE802.15.4, which divides the spectrum between 2400 and 2480 into channels 11-26. From wikipedia “In the 2.4 GHz band there are 16 ZigBee channels, with each channel requiring 5 MHz of bandwidth. The center frequency for each channel can be calculated as, FC = (2405 + 5 * (ch – 11)) MHz, where ch = 11, 12, …, 26.”
(more…)
August 18, 2009
Radio Interference Test #3
In this experiment we will measure the disruption of the signal with various obstacles.
Position the three motes as before in a triangle, 1 m apart. Arbitrailiy assign one of the motes to be the target and the other two sensors.
SubExp 1: Take a base-line measurement
SubExp 2: Stand close to the target (as if wearing it as a badge). Take measurements for 2 minutes. Rotate your position relative to the target and the sensors so you are standing off to the side, still close to the target by about 30 degrees (as if you were wearing the badge and rotated your position, but don’t move the target). Continue to move around the target in 30 degree increments for a complete revolution.
SubExp 3: Stand between the target and sensor 1. Move backwards toward sensor 1 in increments of 10 cm to a distance halfway between the target and the sensor.
SubExp 4: Place a closed laptop facing vertically (as if carried in the hand on its side) so it is blocking the line of sight from the target and sensor motes. Move the laptop in increments of 10 cm to a distance halfway between the target and the sensor.
SubExp 5: Connect the sensor to the USB port of the laptop with a short (5-10cm ) USB extension cord and place the laptop horizontally next to the sensor. Move the laptop around the sensor 360 degrees in increments of 30 degrees.
Radio Position Sensitivity Test
The next test involves the same software and three compatible motes, defined in the first experiment.
Set up the three motes as before (in a triangle, 1 m apart). Two will remain fixed, one will be moved relative to the other two. In each position let the motes communicate for 2 minutes and record the number of attempted connections, the number of successful connections, the average signal strength and its standard deviation.
Move the mote in the horizontal plane 360 degrees in increments of 30 degrees (12 trials).
Move the mote in the vertical plane 360 degrees in increments of 30 degrees (12 trials).
Move the mote in a direction bisecting the other two motes in increments of 10 cm over a distance from the point between the other two motes to a point 2 m distant from that position along a line perpendicular to the line between the other two motes (20 trials).
Radio Compatibility Exp #1
Ted brought over an app that runs on three interacting motes. A sniffer picks up which of the motes received a message from another mote and the strength of the message received. The first test we want to perform is a compatibility test among the motes, to see if some just don’t play well with others.
Details below.
(more…)
Mission Statement
The purpose of this blog is to track the developments of the hand hygiene project in the GROK lab. We’ll post research notes and designs here as we go along.
Currently the hand hygiene project consists of several subtasks:
1. The instrumentation of automatic hand sanitizer dispensers to emit a radio transmission when they are used, using software developed for motes (small microprocessor/radios) written by Ted Herman of the CompEPI group.
2. Design and implement devices that trigger a radio transmission when a manual dispenser is used.
3. Redesign the mote badges worn by the health care practitioners using the system.
4. Use human factors techniques to improve the design of the components.
5. Perform human factors experiments to improve hand sanitation in hospitals.