analogue memory system

[ratstar]

Hopefully this picture is neat enough to be understood.
You catch the value in an exiting capacitor each time, and u get to keep reusing the "signal level".

The idea is, you set the capacitor to a setting, and then it ping pongs with an ac signal, and it can keep repeating the same amount of capacitance into the machine, and then you dont need an adc! itll output the capacitance over and over again, until it leaks out of the capacitors. When it leaks out it actually gets louder, not quieter.

The cool thing, is that it comes out of a capacitor discharge, it means it will only be electrically attracted to itself, so you can do analogue arithmetic with it, and if you add rectification it can actually do conditional logic as well.

When it comes to setting the level, you do it from the logic, for the same amount of time that it pushed out to the logic, and then that should be the right timing for it to reset the value on the feedback phase. (when output is coming back to the state/input.)

Ive got my whole computer nearly worked out now, nearly time!
Im going to have an analogue ALU, and its going to be so much simpler than the log cost digital adders take! Why digital computers are such a waste of space. Being 32 times smaller than a 32 bit computer is really going to count.

 

[Harald Kapp]

What do you mean by

it ping pongs with an ac signal,

???

it can keep repeating the same amount of capacitance into the machine

Capacitance is an expression for the amount of charge a capacitor can hold at a given voltage (C = Q / V). You can't "repeat an amount of capacitance into a machine". You can't.

The cool thing, is that it comes out of a capacitor discharge, it means it will only be electrically attracted to itself,

What is "it" and how is "it" attracted? Do you understand how a capacitor works?

When it comes to setting the level, you do it from the logic, for the same amount of time that it pushed out to the logic, and then that should be the right timing for it to reset the value on the feedback phase. (when output is coming back to the state/input.)

??? voltage level = timing?

Frankly, I have years of experience with electronics but not the faintest idea what you are about

 

[ratstar]

Frankly, I have years of experience with electronics but not the faintest idea what you are about

Thats ok, I know that im a bit of an alien when it comes to doing technology, I definitely didnt get experience with years in HAM radio thats for sure, not that I even know what it is exactly. I have designs for wallpaper lcd screens, but I couldn't fix one that you probably could. (Like a normal cathode ray tube.) Maybe... i shouldn't be posting my crazy work, but I just feel so alone. Ive been banned off every forum on the internet, because I do electronics in a wierd way.

So, maybe ill stop bothering you guys, I used to be on programming forums, and I was more welcome there with my crazy way of programming, cause they dont mind as much BUT!! I found ALL!! the electronics forums I was completely lost with everyone, and I didnt find a single place where people would like what I posted. But it wont change me, I just do things my own way.

But if this, what I call "electronics", isnt actually "electronics" to u guys, I might have to just spit the dummy, and just know quite strangely to myself, I just cant do an electronics forum with it, because I dont have an outlet for myself, strange but true.

???

I just mean its going back and forth, excuse my silly talk...

Capacitance is an expression for the amount of charge a capacitor can hold at a given voltage (C = Q / V). You can't "repeat an amount of capacitance into a machine". You can't.

If you keep going back and forth through a capacitor, it will charge and backcharge, so im just keeping it at a definite position in the capacitor, instead of using the whole capacitance value.

What is "it" and how is "it" attracted? Do you understand how a capacitor works?

I just mean if another capacitor is discharging in the circuit with this one, they will have an independant current loop, they wont be going to each others negatives, and it just makes the alu easier to design.

??? voltage level = timing?

I just mean timing as in how long it takes to fill the capacitor, thats how fast the ac should be.

 

[Nanren888]

Sorry, also don't follow any of it.
Is what you are proposing like bucket-brigade devices?
32bits on an analogue signal is quite some impressive SNR. Digital, particularly binary, has benefits in terms of noise imunity.
.
Maybe you could talks us through it in a less abreviated form?
.
Some of us are old enough to have actually used analogue & hybrid computers. Can be great for differential equations. Multipliers, adders filters and such.
Not sure that you'll get even 16-bits accuracy without rather a lot of work and money.

 

[ratstar]

I dont mind the bad precision or the lossy state, having a 1 line bus is worth it, especially if you have physical restrictions.

Im going to be hooking it up to be an analogue physics engine, like the old differential analyzer simulations back in the olden days I imagine to myself, only being born 81 its a bit of a made up fantasy. I dont actually know for sure what went on back then. I ask my parents about Alan Turing and they just say they never knew about a computer at all back then, making me very confused child.

Lot of work - maybe, but ive spent bugger all on it sofar, its all in the head. $0 gets you real far if your prepared to do the homework.

Analogue delay is cool, I might be doing it, I'm not sure what I am doing with the state yet, but its definitely a possibility, if youve got a one core machine, delay delivering the state to it actually is nice seeming to me.

 

[Harald Kapp]

Thats ok, I know that im a bit of an alien when it comes to doing technology,

I agree. Have you ever read some basic electronics tutorials or books? I strongly recommend you do because your ideas of how electronics and electronic components works are (in mild words) "strange". In this specific case you need to undestand how capacitors work and how their operation is related to AC signals (voltages and currents).
It will also be a great help for you and for us if you get used to the correct expressions (maybe call it slang) used in electronics. Your using random words and descriptions makes it really hard if not impossible to follow your line of thought (admittedly this may be in part due to my not being a native English speaker , but is so, then only for a veeery small part).

I just mean its going back and forth, excuse my silly talk...

Going "back and forth" is not any more precise than "ping pong".

If you keep going back and forth through a capacitor, it will charge and backcharge, so im just keeping it at a definite position in the capacitor, instead of using the whole capacitance value.

This is nonsense. You always use the full capacitor's value. You can change the voltage across the capacitor by changing the charge within or change the charge by changing the voltage across it, It all boils down to the basic capacitor equation C = Q / V.
If you charge and discharge a capacitor in succession, neither voltage nor charge will stay at a "definite position" but will constantly change.

I just mean if another capacitor is discharging in the circuit with this one, they will have an independant current loop, they wont be going to each others negatives, and it just makes the alu easier to design.

There is no "other capacitor" mentioned in your first post. How are we do know what you mean if you don't tell us. Also, to check the validity of your statement we need to see the specific arrangement of components, i.e. a clear schematic circuit diagram - I showd you what this mean in your other thread about batteries in series.

I just mean timing as in how long it takes to fill the capacitor, thats how fast the ac should be.

AC isn't fast or slow. AC has a frequency, an amplitude and a waveform. If not mentioned otherwise or clearly deductible from the context, AC means a sinusoidal signal. If you connect this type of AC voltage across a capacitor, you will have an instantaneous current according to the equation I = C × dV/dt where dV/dt is the change in voltage over time. This may be what you mean by "fast", but you could also mean the frequency. See how important it is to use correct terms (or at least try to use as correct terms as possible).
When AC is applied to a capacitor, it will permanently be charged and discharged. There is no meaning in "how long it takes to fill the capacitor" because as soon as it is "filled" by the rising slope of the AC it will be drained by the falling slope of the AC.

I found ALL!! the electronics forums I was completely lost with everyone, and I didnt find a single place where people would like what I posted. But it wont change me, I just do things my own way.

Ive been banned off every forum on the internet, because I do electronics in a wierd way.

Do some introspection. I doubt all the others are wrong. If you want to be part of a community, adapt to their rules. Otherwise you risk being being banned from this forum, too.

 

[ratstar]

Going "back and forth" is not any more precise than "ping pong".

How about alternating between positive and negative, is that better language?

This is nonsense. You always use the full capacitor's value. You can change the voltage across the capacitor by changing the charge within or change the charge by changing the voltage across it, It all boils down to the basic capacitor equation C = Q / V.
If you charge and discharge a capacitor in succession, neither voltage nor charge will stay at a "definite position" but will constantly change.

What if the capacitor is half charged?

This is nonsense. You always use the full capacitor's value. You can There is no "other capacitor" mentioned in your first post. How are we do know what you mean if you don't tell us. Also, to check the validity of your statement we need to see the specific arrangement of components, i.e. a clear schematic circuit diagram - I showd you what this mean in your other thread about batteries in series.

The ALU isn't in the diagram, its only the analogue register.

This is nonsense. You always use the full capacitor's value. You can There AC isn't fast or slow. AC has a frequency, an amplitude and a waveform. If not mentioned otherwise or clearly deductible from the context, AC means a sinusoidal signal. If you connect this type of AC voltage across a capacitor, you will have an instantaneous current according to the equation I = C × dV/dt where dV/dt is the change in voltage over time. This may be what you mean by "fast", but you could also mean the frequency. See how important it is to use correct terms (or at least try to use as correct terms as possible).
When AC is applied to a capacitor, it will permanently be charged and discharged. There is no meaning in "how long it takes to fill the capacitor" because as soon as it is "filled" by the rising slope of the AC it will be drained by the falling slope of the AC.

So high frequency is better said "more frequent" rather than fast. Thankyou for the distinction.
I always thought AC can be any waveform, it doesnt even need to even go past 0, to be AC, even though maybe its not as nice to use?

Do some introspection. I doubt all the others are wrong. If you want to be part of a community, adapt to their rules. Otherwise you risk being being banned from this forum, too.

If it happens I've got no control over it. I guess I have to learn the savage lessons...

 

[Harald Kapp]

How about alternating between positive and negative, is that better language?

Starts to get better, but what is it that is changing polarity? Current, voltage, charge?

What if the capacitor is half charged?

It still has the same capacitance C, only Q and V are 1/2 of the max. value.

The ALU isn't in the diagram, its only the analogue register.

However your ALU is built: showing us only a part of a circuit and expecting us to guess what's happening will lead us nowhere.

I always thought AC can be any waveform,

It can. But without you telling us which waveform you want to use, it is again useless to discuss this as the behavior depends on the waveform, too. Plus, as I stated, without explicit declaration of the waveform usually a sine is assumed.

it doesnt even need to even go past 0, to be AC

In that case it is not pure AC but AC + DC. Pure AC has a mean value of 0 (volt, ampere). A signal which alternates e.g. between 0 V and 5 V is not pure AC but can be deconstructed into a DC signal of 2.5 V and an AC signal of 2.5 V (peak). The sum of both makes for a combined signal that changes from 0 V to 5 V. Both AC and DC parts of such a signal can vary in amplitude, they do not have to be of the same amplitude as in this example.

 

[ratstar]

Starts to get better, but what is it that is changing polarity? Current, voltage, charge?

The voltage.

It still has the same capacitance C, only Q and V are 1/2 of the max. value.

But u could also put in the mathematics, half the capacitance, because that is also true, if you say that a capacitor shrinks as it fills up, and becomes a smaller and smaller capacitor as it gets charged up. Thats why its non-linear, because it only gives out what breath its got left in it, like a balloon. And its the same when its charging too, eventually charge cant even fit in them.

However your ALU is built: showing us only a part of a circuit and expecting us to guess what's happening will lead us nowhere.

I'm frightened to even post it, once you see how different it is, I'm not sure your going to like it.

It can. But without you telling us which waveform you want to use, it is again useless to discuss this as the behavior depends on the waveform, too. Plus, as I stated, without explicit declaration of the waveform usually a sine is assumed.

Sometimes to machines, they are "Invarient" to some kinds of change, as in you can have a lee-way with a resistor, or the capacitors value doesnt matter so much and it still works. This "lee-way" is something I appreciate about circuits, and I feel I might be alone there, because people dont like my laziness it gives me when I describe circuits.
The wave form doesnt matter here, as long as its equal in both directions, could be square, could be sine, could be a discharge curve, I guess its better if its square because its a computer, and it would pump the capacitors faster.

In that case it is not pure AC but AC + DC. Pure AC has a mean value of 0 (volt, ampere). A signal which alternates e.g. between 0 V and 5 V is not pure AC but can be deconstructed into a DC signal of 2.5 V and an AC signal of 2.5 V (peak). The sum of both makes for a combined signal that changes from 0 V to 5 V. Both AC and DC parts of such a signal can vary in amplitude, they do not have to be of the same amplitude as in this example.

That sounds right in my book too.

 

[Harald Kapp]

that is also true, if you say that a capacitor shrinks as it fills up, and becomes a smaller and smaller capacitor as it gets charged up.

Capacitors do not shrink or grow with applied voltage - at least not the typical capacitor you use in electronic circuits. The "charge" in a capacitor results from the distribution of positive and negative charge carriers on the capacitor electrodes.
This is not to say that one can't construct a capacitor where the distance between the electrodes or the surface area of the electrodes facing each other varies. This has been done in mechanical ways (rotary variable capacitor) or electronic ways (voltage tuned capacitors), see this reference. But you have in no way indicated that you are going to use one of these.

I'm frightened to even post it, once you see how different it is, I'm not sure your going to like it.

Likely to happen

because people dont like my laziness it gives me when I describe circuits.

No small wonder considering this thread

 

[ratstar]

Am I right when I say, theres not one internet forum for electronics that handles experimental design.

So no-one is to do experimental design on electronics forums. That was my mistake.
Seems like a damn shame to me, It could be fun...

You better watch out for my ALU, cause it has a current divider in it to do the divide function. (WTH!?!?!?!?!?)

 

[Harald Kapp]

Am I right when I say, theres not one internet forum for electronics that handles experimental design.

Depends on what you designate "experimental" ...

 

[ratstar]

Could be anything really, just some-one with an original circuit, maybe... with a few mistakes here and there, and maybe some conceptual flaws, but just someone showing something that isnt just replacing a capacitor in a washing machine.

 

[Harald Kapp]

Sorry to say but your "designs" are far more "experimental" than that.

 

[ratstar]

I agree with you, I do electronics like a crazy wierdo dont I... if only someone would believe me that my stuff works as good as the "normal way to do it" but... oh well, freaky-tech dudes just have to live on, by themselves, lonely, shunned by the world.

If it works... doesnt matter how wierd it is does it?