sstimuluss

I guess you mean the common connection point that exists when one person touches all four electrodes. This is why I've been representing skin as a four-resistor bridge when dealing with two pairs of electrodes..

Roff

What is the physical arrangement of the 4 electrodes?

__________________
Ron

Hero999

I didn't concentrate on the circuit to really try and understand it, however I can see that the differences between the two transistors will mess it up, you need a matched pair of transistors.

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez

sstimuluss

it hasn't _seemed_ to matter all that much how the electrodes are arranged, but basically using large metal plates placed on the floor. the leads connecting electrodes to circuit are quite long, about 4 metres each.

below is the circuit i've been using. one of the 1.8v LED's is mounted inside a small tube, opposite a LDR which affects a midi controller value. the optoisolator outputs are fed into a cmos circuit that eliminates bounce/false triggering and eventually triggers a midi note-on/off. so for each pair of electrodes, closing the circuit with light touch turns a midi note on, and then further contact (more surface area, less resistance) generates midi controller data.

[EDIT: the extra 1.8v LED and 2.2v LED are just indicators.. 1.8v matches whats happening with midi data, and 2.2v is normally on, turning off when a note is triggered]

the reason electrodes are on the floor is that contact dancers are to interact with them, essentially forming human chains between various electrode pairs to generate music, playing on amount of skin contact. but for the purpose of my query as to how to use just one power supply instead of a battery per pair of electrodes without losing the current independance, one only has to consider a single human body making contact with more than one pair of electrodes.

Attached Images

led fader-trigger touch circuit.jpg (43.4 KB, 6 views)

sstimuluss

the transistors aren't really the issue, excuse my possible ignorance (still learning ) I'm sure that matching them would be important somehow, but the problem of current crossing over still remains even without transistors in the circuits.

sstimuluss

bump.. please help, i've been looking for a solution to this problem for weeks now. Granted I can just use a battery per touch circuit, but there Must be a more elegant solution! My reasoning is that, apart from saving money, it would be better to be able to run all the touch circuits at equal voltage, so the current range for the LEDs generating midi controller data is within a defined range for processing on the computer, whereas using batteries results in variable ranges. I can't really use a 5 volt regulator on each touch circuit since 5 volts doesnt give enough current range throught the skin (because I'm using the first transistor as an amplifier rather than a switch) As I understand it, one needs to use a voltage source at least 2 volts higher than the regulator, so 5 volts seems like the highest option when using 9 volt batteries.
Ideally I would like to be able to power multiple touch circuits from a single AC-DC transformer with a regulator to smooth out ripple etc.
If my actual problem isn't explained well enough, please let me know and I'll try to clarify.

Roff

I finally understand your problem. The only solution I have thought of so far is much more complicated, involving either frequency or time selective circuits. How many pairs of electrodes do you have on the floor?

__________________
Ron

Russlk

The key is, that the negative and positive of the battery cannot be common. A transformer with multiple secondaries and rectifiiers (one for each circuit) is all that you need (or multiple transformers). That would be a custom design.

__________________
see my website: www.geocities.com/russlk

sstimuluss

ah ha.. thanks! i've never designed a transformer circuit so will have to do some research.. maybe a stupid question, but is it possible to use secondary DC-DC converters instead of secondary AC-DC? or would that still result in a common ground?

if a custom transformer design ends up being too much work for me (i have deadlines for show unfortunately), I'll have to use multiple transformers..

i recently picked up seven or so 8.5V 60mA AC-DC adapters at the dump, however when reading their voltages with a multimeter, an open circuit measures 14.2V, a 1M resistor measures 13V, and a 1.5K resistor measures 9V
.. this isn't the same as another adapter i have which measures 9.15V no matter what.. doesnt seem right.. and a bit scared to use them in a circuit designed for 9v.. i'm guessing they're old cellphone chargers, which may use different transformer design to regular adapters

sstimuluss

I'll be using maximum of 10 pairs, probably around 6 or so would be more managable in performance..

I've pondered over that approach as well, but not yet tried going down that route.. to somehow step through the touch circuits one by one to take readings eh? that would eliminate the need for exclusive V+ and Gnd..
would it be possible to take this approach and not have the LED's flicker on and off? even the slightest ripple causes oscilation in the midi controller data, not good

Roff

Frequency division multiplexing:
Each circuit has an oscillator (relatively low frequency) connected to one electrode, and a tuned amplifier connected to the other electrode, and a synchronous demodulator following that receiver, the output signal being proportional to the resistance between the electrodes. You would then use that output to drive your MIDI transducer. The frequency selectivity would cause it to reject signals from another electrode, which would be at another frequency.

Time division multiplexing:
Generate N pulse trains with duty cycle=1/N, where N is the number of transducer pairs, and each pulse train would be at a different phase (time slot). You could then synchronously demodulate the pulse trains (the demod acts as a sample-and-hold), and feed each output to its respsctive MIDI transducer.

Russ's separate transformer windings and regulators might be easier. Your idea of isolated DC-DC converters should also work. It still boils down to separate transformers (inside the DC-DC converters).

__________________
Ron

sstimuluss

Thanks for the info! fascinating.. will have to do a bit of research to fully understand what you're suggesting, but sounds worth my while to try designing such circuits regardless of which approach i end up using for this particular project.

seems like the quickest fix would be to use isolated DC-DC converters, if all that's required is an IC that is..

Russlk

The simplist solution is multiple transformers, this is the circuit. Parts are available from Mouser (www.mouser.com). The capacitor, C1, is probably not needed for your application. You could use the whole secondary but the output voltage would be double (about 18 volts).

Attached Images

PS.jpg (45.2 KB, 3 views)

__________________
see my website: www.geocities.com/russlk

sstimuluss

thanks, i haven't mentioned i live in new zealand.. would this be a good alternative?

12.6V CT, 150mA 2VA Centre Tap - Type 2851 Transformer
http://www.jaycar.co.nz/productView....=&SUBCATID=411

also, would there be much fluctuation in the output voltage?

Russlk

Yes, that transformer would work fine.

__________________
see my website: www.geocities.com/russlk