using identifying constants
computon
Posts: 6
I am working on an application where I have a partial circuit. Any one of 48 different "items" can be used to complete this circuit, and I need to know which one does. I've tried using identifying resistors inside these "items" to complete the circuit, and capacitors to figure out which "item" it was, but it's been way too unreliable.
A friend of mine thinks resistors aren't the way to go, since if the contacts aren't touching just so, it could give a different reading, which may confuse one "item" with another, even with very widely spaced resistances.
Does anyone have any ideas on what I could do? Any help would be appreciated. Thanks.
A friend of mine thinks resistors aren't the way to go, since if the contacts aren't touching just so, it could give a different reading, which may confuse one "item" with another, even with very widely spaced resistances.
Does anyone have any ideas on what I could do? Any help would be appreciated. Thanks.
Comments
Yanroy, what you're suggesting is that I have 6 jumpers per "item" input into a shift register, and multiplexing the three output lines to multiple (how many?) shift registers. Is that correct? If I have the BS managing multiple circuits which are all multiplexed together, how do I know which circuit sent in the information? This is also important to my application. Also, I'm not sure if using six contacts per item is feasible.
Is there any other way of doing what I need?
If not, and I'm going to need to take an approach similar to that which Yanroy suggests, can anyone recommend a reputable electronics prototyping firm to do the job?
I'm not sure I get the picture of what an "item" really is, but for the sake of my suggestion below, we'll assume it's nothing more than a black box with external contacts on it. Everyone seems to be trying to identify the black box ("item") by its value, based on your lead-in. That may be unnecessarily restricting, only you will know.
Here my thought which ignores its inherent value. Stick an I-Button or small prom/eeprom inside, and every "item" will have an ID rather than a value. Now when it's inserted in the circuit you "read" a "sense" contact and determine from that what it is. Simple enough to put a reference table in the Stamp program to cross reference from ID to value(s).
Just a thought outside the box.
Regards,
Bruce Bates
From how you phrased the question about my proposal, I don't think you quite understand what a multiplexer is.· It doens't just combine signals from multiple inputs and pass them to a single output - It's more like an addressable switch.· You tell it which of the inputs to pass through by giving it different addresses.· Of course, this means you'd have to go through and poll each one of the shift registers in turn to read all the different possible places an item could be plugged in.
If 6 or 7 contacts per item is prohibitively high, perhaps an EEPROM in each one (as bruce suggested) is the way to go, though it strikes me as expensive and I doubt you could do with less than 4 contacts (power, ground, data, clock).
I hope I made myself clearer and didn't confuse you further [noparse]:)[/noparse]
Post Edited (Yanroy) : 7/19/2005 8:37:25 PM GMT
At first I imagined that you would need at least one contact. Not quite so. It just occurred to me that there is a NO external contact solution.
This is just a non-specific overview. Inside the "item" is a wound coil circuit which is not internally connected to the "item". Inside the "carrier" is an ocscillator. Each "item" will have a different coil, and thus a different resonant frequency. The carrier oscillator starts winding up, with ever increasing frequencies, until it "sees" the feedback from the resonant frequency of the "item" in a "receiver" loop coil. Not unlike RFID but much more localized, and I'd guess a good deal cheaper.
The (so called) big money is in the carrier which is a one time expense per unit. The coils should be cheap enough and one might even wind them by hand, if this is not a production effort.
Just another approach.
Regards,
Bruce Bates
My understanding (if Im not spot on with Bruce, I think this method may also work) is you are creating an air core transformer. The base has inductor near the surface the object will be placed. That inductor is energized with perferably a sinusoid AC signal. That inductor will have a natural reactive response to the energizing signal. An object contains an inductor and a capacitor in series, the combination of the inductor and capacitor is set to a specific frequency so that it's reactive resistance is a minumum at that frequency. The frequency can be set by changing the value of the inductor or capacitor. Perfereably the initial energizing signal will be a frequency that does not match any object, so that an object placed near the receiver coil will resist the coil's energization field. This should be detectable by the reciever as an effective change in the receiver coil's reactive resistance. Once this "object detection" has occured, the energization signal's frequency is ramped until the matching frequency for the object is found, and the object's identity is known.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·1+1=10
Post Edited (Paul Baker) : 7/20/2005 1:17:06 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·1+1=10