This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
paulsantangelo
Posts: 75
Joined: Mon Aug 29, 2005 1:01 am
Contact:

Hey there All:<p>I need a little help, At least a direction to point in. My background is a little limited in electronics DIY-fer and a little high scool electronics class.<p>I'm trying to build a circuit that has 8 sensors, nothing fancy just reed switches. When the reeds close, I need them to complete serveral circuits. <p>This is kind of difficult for me, I can just imagine what I'm trying to accomplish.<p>I have 8 reeds and 8 circuits, When a reed switch closes, I need 3 of those circuits to close. There is a pattern to follow and was thinking a logic circuit or some sort of array to accomplish my goal.<p>A quick notation of the pattern follows<p>reed 1 closes, circuits 456 closes
reed 2 closes, circuits 567 closes
reed 3 closes, circuits 678 closes
reed 4 closes, circuits 781 closes
reed 5 closes, circuits 812 closes
reed 6 closes, circuits 123 closes
reed 7 closes, circuits 234 closes
reed 8 closes, circuits 345 closes<p>There is some logic here, I just dont know a lot about logic circuits -unfortunatly. I started looking into it, have a few books that I borrowed, yellowed, he said yellowed books have the real story in them. I just dont know which way to leap.<p>Any thoughts would be great entertained and appreciated!!<p>-Thanks
-paul

Externet
Posts: 1855
Joined: Fri Feb 01, 2002 1:01 am
Location: Mideast USA
Contact:

Hello Paul.
That is acomplished with the use of diodes, or a diode array.
Example for the first situation Reed 1 closing:
The switched-on (circuit side) of reed 1 should have 3 diodes, one signalling to circuits 4, another to circuit 5 and the last to circuit 6 with the diode orientation depending if + or - is supplied.
That way, the closure signal of reed 1 activates circuits 4,5,6.
Repeat the same with all 7 other reed switches, feeding also 3 diodes each, and routed to the corresponding desired circuits.
You will end with an array of 24 diodes and your wishes will be served.
Miguel
- Abolish the deciBel ! -

philba
Posts: 2050
Joined: Tue Nov 30, 2004 1:01 am
Location: Seattle
Contact:

This is a very standard technique - encoding matrix.

paulsantangelo
Posts: 75
Joined: Mon Aug 29, 2005 1:01 am
Contact:

Thanks For all the responses. Remember when I said I was having difficulty imagining what to do. I got the pattern reversed. The circuits are on intially, and when the selected reed is activated the ciruits are opened.<p>I know this changes things but how?<p>Thanks again<p>-paul

cato
Posts: 366
Joined: Sun Sep 07, 2003 1:01 am
Contact:

You haven't fully described what you want to happen. For example, using your original chart for conversation purposes, what do you want to happen if BOTH reed 1 and 8 close? 3456 close or nothing closes because its not JUST reed 1 or JUST reed 8?<p>Thats a retorical question.<p>I suggest you change the focus of your thinking. Instead of writing the problem the way you have, focused on the switches, focus on the outputs. For example, it looks like you want (including your correction)<p>4 off if reed 1 or 7 or 8 is closed....which can be accomplished with a three input OR-gate and an inverter or a three input NOR-gate....<p>[ October 21, 2005: Message edited by: cato ]</p>

paulsantangelo
Posts: 75
Joined: Mon Aug 29, 2005 1:01 am
Contact:

Thanks.<p>You were right! Re-forcusing was what I needed to do. I got so wraped up in logic circuits that I forgot the real purpose of the circuit. I came up with a double switch. Basically as you know the reeds are a switch which when activated (lack of a better word) will allow voltage to flow. I needed that. At some given time, I need three ajacent circuits to open the cicuit, even though the reeds are still active.<p>For the second switch I came up with a PNP transitor, where three gates will be attached to one pole of a rotarty switch. The rotary switch will deterime which ciruits are opened.<p>The only thing I dont like about this idea, is that its kind of hokey. I mean three conductors are wired to a single pole, they alternate of course, one pole will control gates for circuits 456 while the next pole will be wired to control the gates for circuits 567 and so on.<p>It should work but I was hoping that I could come up with something a little more refined! I like the ideas of using logic gates, I'm just not sure that they would work here because the wires flow + Voltage to the PNP gate, I'm not sure how they could help!<p>If you got any ideas, I'm listening, If not I want to thank everyone that helped me with this, In one week, I went from crazy obsesive to getting a grip my situation, thanks for all the help! Focus, its important!!<p>-paul

Enzo
Posts: 276
Joined: Tue Sep 14, 2004 1:01 am
Location: Lansing, Michigan, USA
Contact:

Paul, tell us what you are trying to accomplish. What is the end product to be?<p>I had a man bring a circuit to me once that he was struggling to design. In involved some switches and a diode matrix - sound familiar? He wanted each switch to be able to independently turn on combinations of circuits without disabling the other switches ability to do so.<p>I looked at his drawings and asked him if he was trying to make a seven segment decoder. He thought it was some top secret idea only he had. He ran a gas station and wanted to be able to light up segments on a large "8" numeral on a sign to make the various digits of his daily price changes. Didn't want his competitors to get wind of his amazing labor saver.<p>I handed him a 12 cent TTL chip and said this did what he wanted in one part, and they have been around for years.<p>Point is that he brought his half-done solution to me rather than his problem. So tell us your actual problem, and we might have something real simple to do it.

paulsantangelo
Posts: 75
Joined: Mon Aug 29, 2005 1:01 am
Contact:

yeah:<p>One of the reasons why I picked this site for help, was that many of you (if not all) are more adept in electronic circuitry than I am. I mean, if I stayed with it when I was younger I probably would know what a decoder does, I'm just finding out about these chips now. I find no references to anything anymore, there used to be Radio Shack and please I dont want to start a whole thread of that. But I really can't find much out there in the way of semi conductor references (like the TTL cookbook).<p>Anyway, back to my issue. To you guys it will be stupid, thats why I havn't given you the whole forest as it were. Matter of fact this is my third design, on paper the first two didn't even work. This one works great, but it keep manifesting itself to be more. This time its my Uncles fault. And with that heres the idea.<p>I'm making a cutoff circuit based on wind speed and direction. The wind speed is all set! As I explained the premise of the circuit to my uncle who was a long timer DEC corp board guy, He said I should do the same thing for wind direction, Hence my quandry.<p>Based on direction I need the ciruit as well as the two next adjacent cicuits to be off when it is selected. With the reeds activated (the wind vane will have a magnet in it), the voltage will transfer down to the NG (No Go) cicuit. What I need is a zone which is selectable that will not allow the NG circuit to be activated. A second switch is exactly what I was after. This way, even though the reeds are active, the only ones that conduct are the ones not affected by the switch. I'm using a PNP transistor (something I just learned about (again very little reference - even at Barnes and Nobel)) as the second switch. By sending a postive voltage to the gate, I get the circuit to open, even if the reeds are active.<p>The way I have it on paper, seems to work, but I was hoping to refine it a little better. To select the proper gates on the transistors, I have three conductors wired to each pole on a rotary switch. I have to use a rotary switch for visual purposes. Thats the part I'd like to refine, if I could.
Maybe use a logic circuit to open the circuit to NG that would be great!<p>And then I could learn more about these logic gates, which is what I want to do!!!
-paul

jimandy
Posts: 570
Joined: Sat Dec 04, 2004 1:01 am
Location: Birmingham AL USA
Contact:

As I read your project it seems you would be better off dealing strictly with logic gates implemented in IC chips rather than mixing in discrete transistors (although obviously they can be used as logic gates). Even with the limited selection from Radio Shack you can build up more complex logic circuits. And yes, "TTL Cookbook" and more importantly "CMOS Cookbook" by Lancaster are essential guides for design at your level. If you have access to an EPROM burner and some EPROMs or EEPROMS you can design some pretty fancy "brute force" logic by figuring out the desired outputs for each set of possible inputs (essentially a refined way of implementing the diode logic pattern in an earlier post.).
"if it's not another it's one thing."

paulsantangelo
Posts: 75
Joined: Mon Aug 29, 2005 1:01 am
Contact:

WOW din't know you could "write" your own logic. What is the language? I do have access to an eprom burner. I might want to try this after I get this system down, unless you can suggest a quick tutorial that I can get.<p>The thing is, logic cicuits at basic, require 1 output and 2 inputs. Unless I'm sesning the wrong thing. The only thing I can use to turn off the transistor is a high. But your not talking about that are you! Your saying to replace the transistor with some logic, which is what I want to do, but I fear I'm nor knowledgable yet on just how to do it, especially when a logic gate needs at least two inputs and has one output. I mean the direction of flow from the reeds is through the transitor to the NG circuit. The high on the transistor opens the ciruit, well on paper anyway.

jimandy
Posts: 570
Joined: Sat Dec 04, 2004 1:01 am
Location: Birmingham AL USA
Contact:

(paul) <blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr> especially when a logic gate needs at least two inputs and has one output. <hr></blockquote><p>Wrongo. The simplest logic gate is an inverter - one input and one output. High signal in, low signal out and vice versa. Of course it is easy to make an inverter from a transistor but you can find a package of 6 of them in a logic IC #7404
"if it's not another it's one thing."

Gorgon
Posts: 325
Joined: Wed May 04, 2005 1:01 am
Location: Norway
Contact:

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>Originally posted by paulsantangelo:
WOW din't know you could "write" your own logic. What is the language? I do have access to an eprom burner. I might want to try this after I get this system down, unless you can suggest a quick tutorial that I can get.<p>The thing is, logic cicuits at basic, require 1 output and 2 inputs. Unless I'm sesning the wrong thing. The only thing I can use to turn off the transistor is a high. But your not talking about that are you! Your saying to replace the transistor with some logic, which is what I want to do, but I fear I'm nor knowledgable yet on just how to do it, especially when a logic gate needs at least two inputs and has one output. I mean the direction of flow from the reeds is through the transitor to the NG circuit. The high on the transistor opens the ciruit, well on paper anyway.<hr></blockquote><p>Hi,
To use an eprom for 'logic' you take 8 address lines as inputs from your reed switches. You may select active high or low input. The rest of the address lines is connected to 0V<p>If you activate the eproms chip select and output enable, the 8 data lines will provide you with the outputs from the 'logic'<p>For each of the 8(9) switch combinations you program the bitpattern in the eprom for the datalines to control whatever you control with this. I supose it is LEDs.<p>Example: switch inputs active level is high.<p>No switches: address - 0x00: Data-00000000
Sw 1-------------------0x01:------00111000
...2-------------------0x02:------01110000
...3-------------------0x04:------11100000
...4-------------------0x08:------11000001
...5-------------------0x10:------10000011
...6-------------------0x20:------00000111
...7-------------------0x40:------00001110
...8-------------------0x80:------00011100<p>(padding added to try formatting table)<p>This way you can make variants with more than one switch set with complete freedom.<p>Any eprom of 256 bytes or more will do.<p>TOK <p>The program you use is just your programming software where you edit the eprom data buffer.<p>[ October 22, 2005: Message edited by: Gorgon ]</p>
Gorgon the Caretaker - Character in a childrens TV-show from 1968.

paulsantangelo
Posts: 75
Joined: Mon Aug 29, 2005 1:01 am
Contact:

Gorgon:<p>That is something I think I am looking for. The logic chip would drive the gates on a PNP transitor or it could drive the b side of a NAND gate. It took me a while to understand it but your right all I'm looking to do is set a pin (3 pins) high based on a switch setting.<p>I guess the difficult part is the sw in your table. I need to use a rotary switch. If the chip you describe has 8 inputs and will make those outputs, Thats kind of what I'm looking for. Can you suggest a part number or is this just a programmable CMOS.<p>-Thanks
paul

paulsantangelo
Posts: 75
Joined: Mon Aug 29, 2005 1:01 am
Contact:

Could someone tell me if a mux/demux would work to select a setting on the rotary switch (1-8) and then on the demux side get the output described above?<p>example<p>Switch select 8 = 00111000<p>Thanks
-paul

paulsantangelo
Posts: 75
Joined: Mon Aug 29, 2005 1:01 am
Contact: