pepke@SCRI1.SCRI.FSU.EDU (Eric Pepke) (10/15/90)
The Spine-Chilling Saga of the Power Glove, part 3 Reminder: I am playing with a Power Glove, a toy version of a 3-D positioning device. It is made up of the glove itself, which has a knuckle box containing ultrasonic transmitters connected to a wrist box containing a microcontroller. There is a 180-degree rotated L-shaped array of ultrasonic receivers that sits on the TV; it is connected to a junction box which has a 9-pin connector to the wrist box and a Nintendo connector to the game. My shipment of 7400-series chips came in, so I took the big step this weekend: I cut the entire wrist box out of the system and tried to connect to the knuckle box and junction box with its array of sensors. The cable coming out of the knuckle box has 12 colored wires. The cyan wire is connected to the anodes of four diodes, each of which is connected to a resistive element on a finger. The four elements are in the thumb, index, middle, and ring fingers. The fifth little piggy gets none. The other ends of the resistive elements are returned on the brown, peach, orange, and white wires. I forget which is which. The red and black wires are, surprise! Vcc and ground. The violet and gray wires control a little minimalist switching power supply for the ultrasonic transmitters. Every time you kick the violet line, it charges a little capacitor a little bit. Gray returns the voltage through a 1Meg/56K voltage divider. In the Power Glove, this is controlled by the microcontroller. It only notices the positive spikes, and the frequency does not matter. Using the same frequency used to drive the ultrasonics works fine. The glove as it is maintains the supply at about 40 volts, so the return should be kept around 2 volts. At 50 KHz, it takes about 1/2 second to get up to voltage. That having been said, the transmitters function at 5 volts, and all of my tests were under these conditions. The yellow and green wires drive transistors that control the ultrasonic transmitters. I think that the green one is the left transmitter, but I'm not sure. Kick these at TTL levels at the desired frequency, and the transmitter beeps. The blue line goes through a 170 ohm resistor to one side of a piezoelectric beeper, the other side of which goes to Vcc. The ultrasonic receivers are resonant frequence devices made up of two little concentric parabolic dishes and a piezoelectric element. The three receivers, which I call A, B, and C in clockwise order, send signals to the junction box. The junction box has a nine-pin connector that goes to the wrist box. Pins 2 and 6 are ground, pin 5 is Vcc. Pins 1, 3, and 4 provide TTL-level outputs for receivers B, A, and C respectively. When the receiver is going, the pin goes up and down with the receiver's resonant frequency. Pins 7, 8, and 9 pass signals from the wrist box directly to the Nintendo connector, and pins 8 and 9 also go to the center receiver to light some LED's. My oscillator was a 555 timer with a 20-turn 5K pot, a 0.022 microfarad capacitor, and a 680 ohm resistor between pins 7 and 8. My frequency counter was a system made up of a series of 7490 decade counters, a little speaker, a stopwatch, and my brain and sense organs. All signals were at ordinary (none of this LS or C stuff) TTL levels, which seemed to work fine. All tests were done with the glove about a meter from the receiver array. Unlike the receivers, the transmitters are not tuned and can be driven at just about any frequency. The receivers responded well when the transmitters were driven at 25 and 50 KHz and poorly when they were driven at 12.5 KHz. I did not count the signals coming from the receivers, just tested for their presence, but I think the resonant frequency of the receivers is 25 KHz rather than 50, based on eyeball estimates of the size of the gap and back-of-the-head calculations. The receivers did skip the occasional beat. Perhaps driving the transmitters with the higher voltage would help protect against that. All this says is that I can get the ultrasonics to work. Setting up timings to see how accurate and precise the delay is is the next step. First, I think I'll try getting the 40 V supply to work. (Does anybody know of a good voltage comparator and/or op amp that works well with a 5 volt supply at this impedance? Either both of my 741's are fried or they just don't work with this setup.) Eric Pepke INTERNET: pepke@gw.scri.fsu.edu Supercomputer Computations Research Institute MFENET: pepke@fsu Florida State University SPAN: scri::pepke Tallahassee, FL 32306-4052 BITNET: pepke@fsu Disclaimer: My employers seldom even LISTEN to my opinions. Meta-disclaimer: Any society that needs disclaimers has too many lawyers.