Differential logic in steering DAC

[yefj]

Hello,till now i am used to the logic of as we increase the number of currnt sources going thew resistor.
It thermometer or binary weighed.
BUt here we have some sort of subtraction between out put voltages and i cant see how this subtraction could make the starecase form of DAC where as we increase sumthing we get a largers voltage value.
It seems that this subtraction makes one side "fight" with the other
Thanks.

 

[dick_freebird]

You would probably make two parallel R-2R ladders
to take the two "horns" of the differential current
"switch".

Settling time might be inferior to an emitter-steered
version, where one side of the pair is wasted to a
supply rail. If you need differential output then the
entire thing would be replicated with a logic phase
flip (e.g. the other side of the diff switch, this time).
The Miller kick from hard switching the base can
be a speed bump to get over, settling time wise.

 

[yefj]

This is not what i asked, i asked:
we have some sort of subtraction between out put voltages and i cant see how this subtraction could make the starecase form of DAC where as we increase sumthing we get a largers voltage value.
It seems that this subtraction makes one side "fight" with the other
Thanks.

 

[dick_freebird]

It doesn't, at the level you show. The arrangement
and its logic and summation method of N weighted
outputs, does.

 

[yefj]

the proffessor in the video lecture says it is a logical output its a special kind of dac

 

[Dominik Przyborowski]

I think, I guess what is your concern (hopefully).

m is some intermediate code and DAC ourput is taken differentially. Switches are differential, too.
So, for code 0,into branch Vop is flowing no current, while into Von full current. Opposite situation is for code 2^N-1, and LSB is then 2IR

 

[yefj]

So if i understand correctly N is constant and m is the code cool
thanks alot

 

[sutapanaki]

Yes, m is the code and N is the number of bits