Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

How to get this differential voltage in analog circuit

Status
Not open for further replies.
M

mengly

Junior Member level 1
Joined
Jan 5, 2005
Messages
16
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
126
For V1=2.5v~5v, V2=2.3v~2.7v,
How can I get the V=V1-V2 by the circuit in standard CMOS technology (not using resistors)? Could it be done with a fully differential amplifier?

Actually, I want to get a single voltage output V=V1-V2.
 

Dear mengly,
You would be knowing how opams could be used to sum different voltages. A summer circuit.
We also have a circuit that would do differencing of signals. Look into any book on opam aplication. Opamps by James K Roberge has it. The book is uploaded in eda board.
The opamp u can implement in cmos.
 

since you're looking for V1-V2, you can do this with difference opamp topology, they're great neat tricks like this. It can also be done with differential amp only the R's would be different from positive and negative side but same principle.
 

If you can use capacitor you can connect terminal 1 to V1 amd terminal 2 to GND.
Then make terminal 2 floating and terminal 1 connect to V2. You will get V1-V2 on terminal 2.
 

Thank you for your reply.
I've seen someone esle used one capacitor to realize this function, but it gets a pulse voltage, not continuous voltage output, with more control timing.
I saw some ducuments present a symbol of differential amp as this sum circuit, and they explain it's realized on chip, so i am wondering if they are implemented by CMOS opamp with special configuration, .
 

see this circuit
 

But I must realize this function by digital CMOS technology,
and NO resistors can be used in it
 

As far as I can see, you have two options to solve the problem:

1) Digital solution: digitize these two inputs and perform the subtraction in digital domain, or
2) Use switched capacitor circuitry with two non-overlapping clock phase to get the V1-V2 function you required, if the sampled nature of signal processing is allowed.

Other pure analog solutions may run into some problems, if you do not have a good resistor available in your 'digital process'.
 

Any digital CMOS process include NWELL (in my case it has about 1K sheet resistivity) N and P Source/Drain, and POLY gate (if it is silicized it has about 10 ohm sheet resistivity).
You can use them to create resistors.
 

But I doubt if the precision of these type resistors can be enough for this circuit
Can the ratios of R are accurately equal?
 

In digital cmos, cant we use poly resistance?
 

To mengly

If layout engineer will care about ratio you will get it. Analog designers use the same resistors. As options they can use High-value POLY resistors and POLY-POLY or Me-Me capacitors.

To xuel

You can use POLY gate resistors, but for modern processes the resistivity will be too small: about 10 ohms per square. For large resistors it may be not acceptable.
 

I checked TSMC.18 Logic Design Rule, P+ or N+ Poly without silicide has a sheet resistance between 200 and 400 ohm/square. Also P+ or N+ diff without silicide has a sheet resistance bwtween 50 to 150 ohm/square. So I think you can choose these kinds of resistances.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top