Continue to Site

# Problem with output voltage of integrator (Spectre)

Status
Not open for further replies.

#### firsttimedesigning

##### Full Member level 1
so i am using spectre to simulate an integrator. (please see the attached file.)
The problem is that at the output of integrator, the output voltage is around 2 MV.
R = 300k and C = 26 pF (p = 1*10^-12)
I think the reason why that the output is around 2,000,000 volts is because spectre has initial conditions set up and it places a very large resistor on top of the capacitor and so the output voltage is large.
Does anyone know how to change the resistance of that large resistor? Thanks in advance.

integrator

can u provide ur test setup from spectre

Re: integrator

i would like to provide test setup..but i dont know how...
i am using transient analysis...

integrator

just give a sanpshot image of test-setup schematic

Re: integrator

There is a parameter gmin (normally I think it is set at 10^-12 pS-) which is used to simulate the resistance of off transistors.
I think you can change it in options in tran analysis.

Bupesh

Re: integrator

i see that option, but what should i changed it to?

Re: integrator

hai here iam attached some details of integrated circuits

integrator

hi , I am afraid that you made sth wrong. The circuit can work properly only when the op is powered by two power supply----a postive one and a negative one.Since Vout should be a negative voltage.

Re: integrator

well thank you for replying...i set the initial condition of the voltage at each node to be 0 and now i got 600V at the output instead of 2MV...it is an improvement but still it is not what i want...any idea why?

Re: integrator

firsttimedesigning said:
the output voltage is around 2 MV.

Did you use cadence's idle “opAmp”? what is you input signal?and how long the integrated time? I think you can reduce you input signal or integrated time, orther way,you can clamp the idle "opAmp" output swing (eg:0~5V).

Best regards!

Re: integrator

i am using voltage controlled voltage source to model the op-amp. so i consider the op-amp as an ideal op-amp. what i did was that i connect the negative output of the voltage controlled voltage source to groud. positive input, negative input, and positive output of the voltage controlled voltage source are connected just like what you would connect to an op-amp.

i also change the inital condiction of each node voltage to 0V. that was how i got the voltage down to around 600 volts. under simluator -> convergence aids -> initial condition, you can change the inital voltage of each node. if the initial conditions are not set, dc steady state values are used instead when finding the transiet reponse.

so yea...i still dont understand why the integrator is not working as it should.
the input is fine...but the output....sigh