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.

Strange Pulse Response of a Buffer

Status
Not open for further replies.

mburakbaran

Member level 2
Joined
Jan 18, 2010
Messages
44
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Location
leuven
Activity points
1,724
Hello,

I'm designing a high Input common mode range opamp in order to used it in buffer configuration in ams 0.35. I'm using some kind of a folded cascode topology. In order to realize the high swing at the input terminal, i am using complementary mosfets. I mean there are 2 parts of the opamp one is built with nmos transistors other is with pmos transistors. when one side is closed, other comes into play the supply the swing need. of course there is a region where both of the parts are working at the same time.

I'm applying 450m-4.6V ideal pulse into the input terminal.

The pulse response is quite strange.

The output signal does its normal slewing, before catching up with the input (about 20mV below the input level) it does some kind of an overshoot (but it is not an overshoot, as the output signal level didnt go over the input signal level yet), it falls about 10mV or so and then it starts rising again and finally catches the input with a 4mV error.


When i apply a small pulse like 2V-3V, it behaves strangely, too. It does a pure overshoot, but it does strange ups and downs till settling.

Phase margin is 64 degrees, i dont think it is about stability. I dont think it is about the damping coefficient either.

Any ideas?

By the way, i am using a flipped voltage follower at both complementary inputs.

Note:This strange behaviour is most significant when the temp is about -196 celcius.
 

Unless you have special models for your transistors, -196C is well below the standard range of operation. I believe your models are failing at that point. Is the behavior as expected at -40C?
 

Unless you have special models for your transistors, -196C is well below the standard range of operation. I believe your models are failing at that point. Is the behavior as expected at -40C?

I will check it tomorrow. By the way, i am not changing "tnom" at cadence, just changing the temp value. I dont think it is related to the failing of the models, i am trying to realize a new buffer to replace the old one which was designed by someone else previously. I am using exactly the same models and transistors. I never experienced such a tracking behaviour, that is why i asked if someone else has ever seen such thing. thanks in advance, i will get back to you.
 

same behavior for -40 celcius for that strange pulse response.

however, i managed to find the cause of oscillation like behavior. it was caused by the voltage difference i placed into the FVF in order to boost swing. when i use ideal vdc, that problem doesn't occur. when i use a basic current source and get benefit of another transistors Vgs to realize about 2.5V difference that oscillation-like behavior occurs.

what is the best way for realizing a voltage difference between two nodes?

thanks for your reply.
 

I am afraid without a schematic I can't really follow your design. Maybe someone else in the forum can give you some suggestions...
what is the best way for realizing a voltage difference between two nodes?
A current through a resistor is one.
 

is your rail-to-rail opAmp using a constant gm design? FVF? hard to understand what you are trying to do w/o a bit of a drawing? out of curiosity was the phase margin measured before or after you replaced the ideal DC source with the device that realized the 2.5V drop? at what OP was it measured?
 

is your rail-to-rail opAmp using a constant gm design? FVF? hard to understand what you are trying to do w/o a bit of a drawing? out of curiosity was the phase margin measured before or after you replaced the ideal DC source with the device that realized the 2.5V drop? at what OP was it measured?

FVF, flipped voltage follower.

Not a constant gm design.

Phase margin is measured with the practical realization of the voltage drop. Phase margin is even better when DC input level is 450mV or 4.6V compared to Vin(DC)=2.5

---------- Post added at 17:13 ---------- Previous post was at 17:06 ----------

I am afraid without a schematic I can't really follow your design. Maybe someone else in the forum can give you some suggestions...
A current through a resistor is one.

A current through resistor is not working for me as I need this specific voltage difference between a mosfet's gate and another's drain. Mosfet's gate voltage is almost fixed, but the other's drain can be manipulated. I want about 2.5 V difference between these two nodes.

Thank you very much again.
 

Hi mburakbaran,

sorry I have never worked with FVFs (very interesting though!) but I would think that having a constant gm design might still be a good idea for a wide input swing design...
 

Hi mburakbaran,

sorry I have never worked with FVFs (very interesting though!) but I would think that having a constant gm design might still be a good idea for a wide input swing design...

thanks a lot. i will consider it accordingly.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top