BJT vs MOSFET for logic gates help

[ratstar]

Ive got a slightly different route than this for my project. > : )

Having a good computer is nice, but you can get a second hand 2.5ghz 4gb ram computer for only 5 quid down the pawn shop, so that fact makes me head off towards a more custom hardware / analogue computer / fpga route than making an ordinary computer, and then I budgetted I can already equal a (latest) GPU's power with only a megahert clock, as long as the program instructions are all in physically.

Then i've actually gained performance out of a DIY computer. Actually did something decent for myself.

 

[Audioguru]

Why are you guys trying to re-invent digital logic when there is already a history of logic types:
1) RTL resistor-transistor logic.
2) DTL: Diode-transistor logic.
3) TTL: Transistor-transistor logic.
4) CMOS: Complementary MOSfets logic.

Here is a very odd Cmos circuit posted by Tristan in one of the posts here:

 

[ratstar]

I can do logic with just resistor- resistor logic, and capacitor- capacitor logic, and inductor logic, they dont seem to be the usual list, but its true you can.

 

[Harald Kapp]

but its true you can.

I have my doubts. You've been promoting this idea across several threads but haven't come up with any proof so far.

 

[ratstar]

I cant proove it to you, it seems that you'll have to work it out yourself in your own time, but I promise u it is possible!!

I could draw up the circuits, its not complicated. The logic is the easy thing, the engine that runs it (the oscillator) is harder in my opinion.

 

[Harald Kapp]

I cant proove it to you, it seems that you'll have to work it out yourself in your own time,

As long as you can't prove it, it's pure fantasy in my opinion.
If you had a working circuit, you could show it to us.

 

[ratstar]

But showing you the circuit would do me no good, (Im thinking your not going to like it.) youll have to think of it yourself.

 

[Audioguru]

But showing you the circuit would do me no good, (Im thinking your not going to like it.) youll have to think of it yourself.

I listed 4 types of logic circuits that worked perfectly and have been used in many circuits over the years.
But you are listing impossible logic that will not work.
The circuit types I listed were made into ICs and computer processor chips. You are not even using any ICs that are readily available and are very inexpensive..

 

[bertus]

Hello,

Here is a very odd Cmos circuit posted by Tristan in one of the posts here:

What is wrong with the symbols that Tristan is using:


Bertus

 

[Harald Kapp]

What is wrong with the symbols that Tristan is using:

Nothing.
The oddity is mentioned in the image:

But caption 1) is wrong: P is indicated by the little circle on the gate.
As for caption 2) : admitted, S and D are not discernible. I think these symbols are simplified to show the logic, not the physical setup. Plus one can swap S and D on a MOSFET if one accepts more or less drastically altered parameters - not that I recommend doing this

 

[Audioguru]

The symbol for a Mosfet should be standardized. I have never seen a datasheet of a Mosfet that has a circle on the gate like an inverter and that is drawn completely symmetrical so you cannot easily see the Source pin.
I have also never seen the arrows backwards.

Hee, hee. Just now I looked at the datasheets for two Jfets and they did not show a symbol.

 

[Harald Kapp]

I have also never seen the arrows backwards.

The direction of the arrows depends on where they are drawn for a reason:

• When drawn on the source pin they pretty much resemble the arrows of a bipolar transistor indicating the polarity of the source (arrow to source = N doped, arrow from source = P doped).
• When drawn on the backgate or substrate pin (although this connection is rarely if ever made available as a separate pin, hence the direct connection to the source in the symbol) the arrow indicates the polarity of the backgate vs. the channel. Thus it looks backwards to the arrow on the source but there's actually logic behind this. Take for example an N-MOSFET, then the backgate/substrate would be P doped and the channel, once brought into conduction is N with respect to the backgate. The arrow makes sense now?

I have never seen a datasheet of a Mosfet that has a circle on the gate

Neither have I. I think this symbol is for illustrating the operating principle of CMOS gates only.

 

[Nanren888]

Interestingly, https://en.wikipedia.org/wiki/MOSFET#Circuit_symbols shows

Which seems to talley with what Harald Kapp was saying.

 

[Audioguru]

A symbol should not need a label on it to what is what.