• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Error "time step too small in opamp"

Status
Not open for further replies.

Archit

Member
I don't have this opamp in my library so I am unable to run it. Most probably your opamp is going into saturation. What you can do is test this opamp in a new file with the required input voltages and power supply. The issue is likely with the power supply you are giving. Also, see from the datasheet its maximum operational frequency and whether you are within the specified range.
 

alec_t

Well-Known Member
Most Helpful Member
The problem may not lie with the opamp model. Sometimes giving voltage sources a finite impedance, or slightly altering capacitor or inductor properties, is enough to get round the problem.
 

MrAl

Well-Known Member
Most Helpful Member
hi
i have used op27A opamp part in converter circuit it shows time step too small error, so kindly help me to solve the issue.
Hello,

There is a model used in your simulation that is not in the library. You will have to supply the model if you want other people to be able ot run your simulation.

You will get time step errors when the derivatives in the numerical solution become too large or unstable. What this means is that even with the smallest time step being used the derivatives can not be contained properly or change too rapidly.

A solution that works sometimes is to make the maximum step size smaller. This sometimes works because it allows the simulator algorithm to be able to control the derivatives leading up to the problem derivative(s) and thus when it gets there again it finds a naturally smoother solution and so is able to solve it with a reasonable time step.

This doesnt always work because sometimes the solution changes too fast (could be a 'shock') no matter what we do with the algorithm. The alternative is to find the problem time constant and try to slow it down a little without changing the overall simulation by too much. The odds are that slowing it down slightly wont change the overall results too much because the time frame over which the solution exists is typically very small compared to the larger time constants.
 
Last edited:

dknguyen

Well-Known Member
Most Helpful Member
Hello,

There is a model used in your simulation that is not in the library. You will have to supply the model if you want other people to be able ot run your simulation.

You will get time step errors when the derivatives in the numerical solution become too large or unstable. What this means is that even with the smallest time step being used the derivatives can not be contained properly or change too rapidly.

A solution that works sometimes is to make the maximum step size smaller. This sometimes works because it allows the simulator algorithm to be able to control the derivatives leading up to the problem derivative(s) and thus when it gets there again it finds a naturally smoother solution and so is able to solve it with a reasonable time step.

This doesnt always work because sometimes the solution changes too fast (could be a 'shock') no matter what we do with the algorithm. The alternative is to find the problem time constant and try to slow it down a little without changing the overall simulation by too much. The odds are that slowing it down slightly wont change the overall results too much because the time frame over which the solution exists is typically very small compared to the larger time constants.
One specific case of this means that there is something that wants to change instantly in your simulation (vertical line on a graph) when the physical component does not allow instantaneous changes (like current through an inductor) and the simulation can't deal with it. Sometimes the way to fix this is to approximate the vertical step by a steep slope that the simulator can deal with.
 

MrAl

Well-Known Member
Most Helpful Member
One specific case of this means that there is something that wants to change instantly in your simulation (vertical line on a graph) when the physical component does not allow instantaneous changes (like current through an inductor) and the simulation can't deal with it. Sometimes the way to fix this is to approximate the vertical step by a steep slope that the simulator can deal with.
Hi,

But how do you tell the simulator to approximate the vertical step with a steep slope. You cant, all you can do is slow down the time constant.
 

dknguyen

Well-Known Member
Most Helpful Member
Hi,

But how do you tell the simulator to approximate the vertical step with a steep slope. You cant, all you can do is slow down the time constant.
You do it by modifying the circuit, not by literally telling the simulator to make treat a particular vertical slope as ramp. Same thing as slowing the time constant I guess, but slightly different mental approach which is what I did the last time I had to do this.

I have also received this error simulating certain types of feedback loops that resulted in race conditions because it was an ideal simulation environment with no noise or disturbances to unsettle the circuit and knock it into equilibrium.
 
Last edited:

ronsimpson

Well-Known Member
Most Helpful Member
time step too small error,
This is a common problem in LT SPICE. Now I wish I could remember......
I think it is from running a very long sim. (in your case 5 seconds) and having a very short Max Time Step.
Example; If you want a time step of 0.1nS and you want 5 seconds of those you have too many data points and run out of memory.

Google "time step too small error LTSPICE" That is what I did when I first saw this.

AND What Mr Al said is usually right.
 

eTech

Well-Known Member
hi
i have used op27A opamp part in converter circuit it shows time step too small error, so kindly help me to solve the issue.
Hi

A few of the symbol and model files are missing.
Some of the symbols will be in the autogenerated folder.
Share your symbols and model files and we'll take a look at it.

eT
 

unclejed613

Well-Known Member
Most Helpful Member
i'm having trouble running it because of missing symbols. however, the most often cause for "time step too small", is if you get positive feedback where you should have negative feedback. this makes it impossible for LTSpice to get a stable operating point. somewhere you have an inverting and noninverting input swapped.
 
Status
Not open for further replies.

EE World Online Articles

Loading
Top