(A&B) & (C&D)

[ranch varment]

Ok theres 4 batteries going in, and I only want the circuit live (Im not worried about series volt doubling yet.) when the batteries are in the exact configuration in the picture, so oposing ands are inverted batteries.

Note there is an inversion in the circuit when I and A&B with C&D.

[John Coloccia]

This is for the terahertz computer, right?

[ranch varment]

No.  this one has diodes in it, itll heat up like crazy.

So the circuit is only complete when the batteries are in this configuration,  change one battery and then itll force positive to positive, negative to negative, with no possible pathway from negative to positive - aka -circuit dead.

So I think,  unless I stuffed it up.

Reverse the diodes and then you get a NOR circuit.

[Rerouter]

Care to describe the function of this circuit? as no matter where you put a voltage, you will never see an output, by its own design, voltage can never flow out of any battery, and if you invert it, you short out all your batteries,

I've noticed a few of your other threads and you seem to want to avoid transistors, but i dont thik its going to work this time, this is the best i could come up with,

[ranch varment]

I wonder if you can care to follow the diodes, then youll see the circuits function A&B&C&D,  in simple talk, change any battery around then it wont have current anymore,  looks like none of you have even seen a circuit like this.

I dont need a resistor except to stop current flowing too quickly,  otherwise it has no bias on the logic at all.

Current attempts to get from battery to battery, but I stop every opportunity, except for the configuration I want, then I have an AND circuit.  Swap the diodes directions, and its a NOR.

The whole circuit is on when its pleased,  but you have to connect at the top if you want the logic to be correct when you add more to it.

thermally wise, its not better than your resistor & transistor,    its actually probably worse (even tho mine is 3 ands, as you can clearly see on the diagram.),  i wouldnt know.   But my circuit is easier to understand as its just one repeated part.

[LabSpokane]

Try wiring it like this instead:

[lapm]

After reading that confusing hand draw schematics? i have to say your circuit fails to do your and-operation. As its drawn, no power will flow except diode leakage current. If you reverse voltage sources they are suddenly shorted throw couple diodes, leading to very high current drain.

Your trying to get signal out when all batteries are present at given configuration?

[hamster_nz]

I have enough small signal diodes here that I dont mind using.

If I was to build this to understand what you are attempting what batteries would you recommend I use?

Also, the secret is not to  create an AND or OR gate but to build a NAND or NOR gate. You can't build anything complex without being able to invert a signal.

 If you had a nicely drawn design for a two-input XOR gate I would love to test it.... (a and not b) or (b and not a)..

[ranch varment]

oopsy im wrong.  i admit it,  B just constantly fires it.

hehehe,   ill try again next time....

[ranch varment]

Hey Hampster!  Thanks for the vote of confidence!

But I just traced the circuit and it doesnt work unfortunately.  :/

I think if you want to use batteries and wires, my only hope left is alternating polarities.
       
postive negative positive, in a triangle.

This way, the idea is its forced to go in one direction faced with 2 alternative polarities, one direction killed by the batteries being together,  so you dont even need a diode for this.

Since your being friendly with me,  1 bit of ram (sorta) looks like this->   (but this WILL NOT work yet,  until I figure it out properly.)



Im just being a tripper,   itll be great if it works!
Problems with it is the loop connects both ways...  the + and - negative in the dotted circle, are supposed to be both switching polarities, but the whole thing is just a foggy dream yet.

Also I drew it wrong, its supposed to be connected only to the positive terminal of the first battery, otherwise it wouldnt switch properly.

[nitro2k01]

What are you even trying to do? 
An AND operation would mean that you get a high, if and only if all 4 batteries are connected. Is that what you're trying to do? Are you even attempting to use batteries, or are they just used as a placeholder for logic signals?

[ranch varment]

The circuit doesnt work...    I looked at the problem again,  and trying to force 3 batteries together with diodes doesnt seem to even be possible...  i was just testing out the idea.

[nitro2k01]

Please answer my questions, though. I don't even understand what you are trying to achieve.

[ranch varment]

They are batteries,  and I thought it worked if they werent just some passive thing,  but unfortunately that doesnt work at all, and they need to be batteries, and it doesnt work when they are.  But it actually doesnt work when they arent either.  Just doesnt work at all.

I can force 2 batteries together,  but even just 3 is impossible with diodes,  because as soon as 2 batteries please each other,  The third isnt taken into account.  But there may be a way to do it!   I just cant think of it yet.

So,  the idea is,  you have to hit every + with a +,  and every - with a -,  and youve turned the circuit off.   But it just doesnt seem possible.

This new idea.  (that battery picture)  is me taking it lots of 2 batteries,  by only connecting one side to one side,  and alternating the polarity of the logic as I go, that way its a tri of two positives and one negative, and its forced to go in one direction, and its what im trying out now.

[ranch varment]

Ok heres another attempt at an and with diodes.




but i think it just stays dead constantly... doesnt work.

[nitro2k01]

So,  the idea is,  you have to hit every + with a +,  and every - with a -,  and youve turned the circuit off.   But it just doesnt seem possible.

This is a bad idea! If the batteries are perfectly matched, it will pretty much work as you intend it to. If they are unmatched, say one battery is 9 V and the other is slightly depleted so it reads 8 V, you will get a 1 V difference. If you connect a load to that, it will be supplied with 1 V. But what's more, for any current you get out of that 1 V, the less charged battery will have a current going backwards through it. Instead of recharging the battery, this will destroy the battery and eventually make the battery leak, even if this takes days, months or years depending on the load and difference in voltage.

I looked into your other thread and saw the statement "then i can piss at a terrahert and not even flinch". You idea doesn't at all address this problem. To build a super fast computer, you need to be able to switch a logic gate really fast. A semiconductor is just an electrically controlled switch. Nothing in your idea is electrically controlled. You'd still need to manually flip a battery to change the state of one bit. But what's more, you connection isn't instant. On the milli to microsecond timescale, you have contact bounce when one battery pole meets another battery pole and the voltage will (relatively speaking) slowly rise. You can see this on an oscilloscope if you touch the probe to one side of the battery while the other side is connected to the ground clip. On the nano to picosecond scale, the length of the wire becomes a problem. Even if you could connect the batteries instantly, it takes time for the wave front of the voltage to travel along the wire until the current starts steadily flowing. This happens at (max) the speed of light (or slightly slower for practical reasons) but this is still enough that in 1 nanosecond, the wave front can travel at most 300 mm. This is why CPUs are shrinking, to combat the propagation delay from the speed of light!

Sorry to crush your dream. I feel sort of like this:

[ranch varment]

I listen to everything you tell me!  And I agree with it.   That as a big thing, it wont go very fast, unless I can pump signal through it over long distances.

About the multi-battery problem-> I think it all gets converged away from the battery locations to one battery once its all working, so thats not my biggest problem, for the future of the project.

But until I get 1 bit of ram to work,  your right about the battery switching, im still stuck switching batteries by hand, or changing wires by hand,  Im trying to figure out ram, putting lots of thought into it!

[nitro2k01]

I'm pretty sure you fundamentally can't do an latch or even an inverter with the components at hand. The problem, again, is that you have no switching, or gain element and also no inverting element. As a thought experiment, try to think up a way that you have a chain of 2, 3 or any number of inverters, such that if you change the input bit, they all change in a cascade. I don't think you can. Without that, anything more complex than an OR or AND gate is impossible.

[Delta]

You are off your tits mate.

[ranch varment]

Delta hahahahaha your right.

[EDIT] i just read what you said,  thanks for helping.  yes i cannot think of what you say. [/EDIT]

I tried my best to come up with ram, but I cant do it without at least a mechanical switch.

But the logic language that works with positive and negative I think is still valid,  its just I have to make the whole thing without a single switch.


so it goes lightspeed but unfortunately you have to advance the state for it, for the time being.
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1. you can do it all from one battery,  but its broken up into readers, setters + constants.
2. you need a converse and inverse reader for the bit.
3. you need to add a logic redundancy convergence. (on negatives) that will fire when a bit wasnt set.
4. and im not sure about more than one positive, shorts it out.  im not sure how important it is to have 2 positives on one decision.

so every serial step in computation. (that it doesnt do properly, because it does it in parallel.)

+/-  +
     +
     -
     -
     -

so its like that every step, but im not sure if two positives short me out.

(note that two variables are being sent in here.)


--------------------------------------------------------------------------------------------------------------------------------------------------------------------------

[Mechatrommer]

(note that two variables are being sent in here.)
[img]

if you cant afford a schematic editor, get a decent cheap pc tablet...

[amyk]

This poster may win the "most WTF topics of the year" award...

[RickBrant]

 

[suicidaleggroll]

Two things:

1) STOP with these hand-drawn paint sketches.  They're impossible to read.  Download KiCAD, it's free, open the schematic editor, and make a REAL schematic if you want ANYBODY to take you seriously.

2) Gates need an output.  What is an "AND" gate if there's nowhere to output the result of the "AND" operation?  None of these "circuits" you keep drawing have any outputs, they're just a bunch of batteries wired to each other with diodes.

[RickBrant]

And with all of the diodes wired such that none of them will ever conduct. Good thing, too, since if they were wired the other way all the batteries would be shorted.