05-24-2011 08:23 PM
well, it is a positioning control system, and the reason I trying to find a transfer function is because I need to look at the step response of the control system and get the performance parameters such as rise time, settling time etc... So this positioning control system is more like digital control system where you use a Z transform--so that's the reason I need a transfer function in S domain and convert that into z domain
05-25-2011 09:55 AM
It can be quite difficult to find parameters for all the mechanical and electrical components of a positioning system to create a model which is good enough to be meaningful.
I suggest that you measure the response to a step input and graph the result. Look at the graph and compare it to what you see in your text books. Is the response like a first order system, a second order system, or something very complicated? If it looks like a first or second order system, then use a model of that order and seek parameters which closely match the performance of your system. If it is complicated try to separate it into multiple first or second order systems with (hopefully) significantly different time scales.
05-25-2011 02:27 PM
I do not understand your question.
The transfer function is defined as the output response divided by the input command. So you will be measuring the transfer function. It will not give you an equation. You have to do that part. If your response to the step input is a nice smooth rise which gradually approaches the final value, then you say, "That looks like the 1-exp(-t/tau) behavior that I expect from a first order system." Then you use the model of a first order system and the equation along with the measured values to determine tau and an amplitude scale factor.
If you do not have any way to measure the response, that is a different issue.
05-25-2011 02:34 PM
well, you said that I can get the response to the step input... now here is the thing
what I have is this
some kind of feedback sensor
some kind of amplifier to provide pulses to the stepper motor
now when you are saying that I can get the response to the step input to estimate the transfer function...then i need to know what would be the procedure to get that response in the labview...
05-25-2011 03:15 PM
To measure the response of any system you need to be able to measure the output of system. For your case it is likely that you will need to measure the position, or possibly, the velocity. The static transfer function can be measured with a scale: N steps of the motor moves the system Y mm. To measure the dynamic response you will need to measure with distance resolution much smaller than Y/N and fast enough to see how the position of the system changes during one step.
Do you have any measuring equipment which can do this? Do you know how to use it with your program?
I am not trying to make this difficult for you. I have done a senior project and I have taught senior project courses. You have not clearly spelled out (on this Forum, at least) what you are doing, what specifications you have, and what constraints you have.
07-01-2011 04:57 PM
Now the transfer function is output response divided by the input command.. I understand this..
I was able to find the generalized transfer function of the stepper motor---but i just want to estimate all the parameters of this equation...
Now as you said in your previous email that the output response in my case would just be a position...
the way my system is set up is that the user would input the exact the position of the stepper motor in angle... then the labview would send out pulses through the daq digital output to rotate the motor to that particular specified angle....
Now Iam using a current driver to amplify the current to drive the stepper motor...
now since the transfer function is defined as output response/input command......so I know that the ouput would be exact position of the motor...what would be the input command in my case..
07-06-2011 04:14 PM
The input command might be the number of steps (and the step rate). Or it might be the current. It somewhat depends on how you define what you want to measure. The current will vary with time, both on/off according to the control signals and within the pulses due to the inductance and resistance of the motor and driver.
07-08-2011 07:29 PM
I appreciate for your response to my previous email...
Now just to give the further details of my vi system and how Iam control the stepper motor is this
Iam using an NI elvis board,, and it has some DIOs and I have connected these DIOs to the current amplifier LM18293-this is particulary suitable for running the stepper motor.
In my VI what I have is the steppers direction switches(Clockwise and CounterClockwise), the VI can run the motor in either half steps or full steps.
It has a knob that allows user to control the speed which in Steps/second.
So the input is the steps per second-I mean this what I would use, since the current is a constant 1 amp coming from current driver IC as I stated above..
Now in your previous what you said makes sense but can you be more thorough on what exactly will I need to to get the step response
I mean will I need gradually increase the speed and monitor the position( starting from 0 steps per second) to get the response.