• AVR Freaks

Hot!speed response time of AN1078

Author
michael.li9
New Member
  • Total Posts : 25
  • Reward points : 0
  • Joined: 2008/08/27 12:57:14
  • Location: 0
  • Status: offline
2019/04/09 16:15:39 (permalink)
0

speed response time of AN1078

Hi Expert,
I am tuning speed PID of AN1078. It will take 100mS from one speed 900rpm to double speed 1800rpm. If I increase the P and I number. There will be a current spark when speed changes. 
Any one make the 100mS shorter? it is not acceptable in our application.
 
thanks
 
#1

14 Replies Related Threads

    Bob White
    Super Member
    • Total Posts : 260
    • Reward points : 0
    • Joined: 2010/11/06 19:52:38
    • Location: Denver, Colorado
    • Status: offline
    Re: speed response time of AN1078 2019/04/09 19:41:59 (permalink)
    4.67 (3)
    I am guessing that by "current spark" you meant to say "current spike"...
     
    If you want to increase the speed of your motor with its attached load you need to supply more power (power = torque x speed).  More power means more current.  When you step up the duty cycle to increase the average voltage on the motor the current will increase sharply until the motor speeds up an increases the back EMF.  There is no free lunch. If you want the motor to change from 900 rpm to 1800 rpm in less than 100 ms you will have to supply more current, both for the time the motor is speeding up and while it is running at higher speed (assuming the same load requiring the same torque - and the required torque is likely to be higher at the higher speed to overcome increased losses such as windage).
    #2
    qhb
    Superb Member
    • Total Posts : 9998
    • Reward points : 0
    • Joined: 2016/06/05 14:55:32
    • Location: One step ahead...
    • Status: offline
    Re: speed response time of AN1078 2019/04/09 21:40:09 (permalink)
    0
    michael.li9
    ... There will be a current spark when speed changes.

    You do not make clear if this is a problem, or just a comment.
    As Bob pointed out, this is entirely normal if you're trying to rapidly change the motor speed.
     

    Any one make the 100mS shorter? it is not acceptable in our application.

    In that case you will need to make sure your power supply can handle an even larger current spike...
     

    Nearly there...
    #3
    du00000001
    Just Some Member
    • Total Posts : 2889
    • Reward points : 0
    • Joined: 2016/05/03 13:52:42
    • Location: Germany
    • Status: offline
    Re: speed response time of AN1078 2019/04/10 02:29:06 (permalink)
    0
    michael.li9
    Any one make the 100mS shorter? it is not acceptable in our application.

    I don't know about acceptability, your motor, your driver - the only thing being the 100 ms.
     
    Your voltage might be too low to achieve faster ramp-up, your motor might disintegrate when trying to decrease ramp-up time, ...
    Basically, more power will result in a faster transistion. But more power might as well transform your motor into some kind of incandescent light bulb  LoL

    PEBKAC / EBKAC / POBCAK / PICNIC (eventually see en.wikipedia.org)
    #4
    michael.li9
    New Member
    • Total Posts : 25
    • Reward points : 0
    • Joined: 2008/08/27 12:57:14
    • Location: 0
    • Status: offline
    Re: speed response time of AN1078 2019/04/10 08:20:23 (permalink)
    0
    robertvwhite
    I am guessing that by "current spark" you meant to say "current spike"...
     
    If you want to increase the speed of your motor with its attached load you need to supply more power (power = torque x speed).  More power means more current.  When you step up the duty cycle to increase the average voltage on the motor the current will increase sharply until the motor speeds up an increases the back EMF.  There is no free lunch. If you want the motor to change from 900 rpm to 1800 rpm in less than 100 ms you will have to supply more current, both for the time the motor is speeding up and while it is running at higher speed (assuming the same load requiring the same torque - and the required torque is likely to be higher at the higher speed to overcome increased losses such as windage).


    You are right Bob. I means SPIKE.
    The transient spike is huge (more than 5 times) compared with stable high speed. for example, 1800rpm. 
    If that can't be solved, Is there any way to compensate, like during speed change back? 
    My application is control the distance (motor revolutions) in some time, for example 1s. There are two speed change, during time, like increase (double), decrease(1/2 or 1/3) then back to normal. Is there any way to control the distance (revolutions) in that time? 
    Only control the speed sounds like not reliable, because the load may changes and can't guarantee the PID take the same time when speed increasing, decreasing. is there any way  to count the motor turns (revolutions)?
    or any sugguest?
     
     
    #5
    michael.li9
    New Member
    • Total Posts : 25
    • Reward points : 0
    • Joined: 2008/08/27 12:57:14
    • Location: 0
    • Status: offline
    Re: speed response time of AN1078 2019/04/10 08:21:34 (permalink)
    0
    qhb
    michael.li9
    ... There will be a current spark when speed changes.

    You do not make clear if this is a problem, or just a comment.
    As Bob pointed out, this is entirely normal if you're trying to rapidly change the motor speed.
     

    Any one make the 100mS shorter? it is not acceptable in our application.

    In that case you will need to make sure your power supply can handle an even larger current spike...
     


    thanks qdh, the current spike is huge.
     
    #6
    michael.li9
    New Member
    • Total Posts : 25
    • Reward points : 0
    • Joined: 2008/08/27 12:57:14
    • Location: 0
    • Status: offline
    Re: speed response time of AN1078 2019/04/10 08:24:51 (permalink)
    0
    du00000001
    michael.li9
    Any one make the 100mS shorter? it is not acceptable in our application.

    I don't know about acceptability, your motor, your driver - the only thing being the 100 ms.
     
    Your voltage might be too low to achieve faster ramp-up, your motor might disintegrate when trying to decrease ramp-up time, ...
    Basically, more power will result in a faster transistion. But more power might as well transform your motor into some kind of incandescent light bulb  LoL


    Thanks du00000001. The transient time is shorter or known is better. I can compensate it in late.
    #7
    crosland
    Super Member
    • Total Posts : 1586
    • Reward points : 0
    • Joined: 2005/05/10 10:55:05
    • Location: Bucks, UK
    • Status: offline
    Re: speed response time of AN1078 2019/04/10 10:12:27 (permalink)
    0
    I don't know anything about that app note. What sort of control loop is it?
     
    If you use FOC then it's easy to apply a limit to the current demand. It will still slow down the acceleration of the motor. You cannot have everything in life.
    #8
    michael.li9
    New Member
    • Total Posts : 25
    • Reward points : 0
    • Joined: 2008/08/27 12:57:14
    • Location: 0
    • Status: offline
    Re: speed response time of AN1078 2019/04/10 11:12:03 (permalink)
    0
    crosland
    I don't know anything about that app note. What sort of control loop is it?
     
    If you use FOC then it's easy to apply a limit to the current demand. It will still slow down the acceleration of the motor. You cannot have everything in life.


    Crosland, Which algorithm has quick response?
     
    #9
    crosland
    Super Member
    • Total Posts : 1586
    • Reward points : 0
    • Joined: 2005/05/10 10:55:05
    • Location: Bucks, UK
    • Status: offline
    Re: speed response time of AN1078 2019/04/10 11:16:00 (permalink)
    0
    michael.li9
    crosland
    I don't know anything about that app note. What sort of control loop is it?
     
    If you use FOC then it's easy to apply a limit to the current demand. It will still slow down the acceleration of the motor. You cannot have everything in life.


    Crosland, Which algorithm has quick response?
     


    It would be better if you answered my question, rather than just asking another.
     
    Any algorithm can have a quick response but you need to decide on the parameters that are acceptable to you.
     
    As you have already found, a quick response will lead to a high current (torque) demand as the motor accelerates.
    #10
    NKurzman
    A Guy on the Net
    • Total Posts : 17610
    • Reward points : 0
    • Joined: 2008/01/16 19:33:48
    • Location: 0
    • Status: online
    Re: speed response time of AN1078 2019/04/10 12:43:34 (permalink)
    4.67 (3)
    No algorithm is going to give you fast response and low current.  Software does not change the laws of Physics.
    #11
    michael.li9
    New Member
    • Total Posts : 25
    • Reward points : 0
    • Joined: 2008/08/27 12:57:14
    • Location: 0
    • Status: offline
    Re: speed response time of AN1078 2019/04/16 15:13:53 (permalink)
    0
    crosland
    michael.li9
    crosland
    I don't know anything about that app note. What sort of control loop is it?
     
    If you use FOC then it's easy to apply a limit to the current demand. It will still slow down the acceleration of the motor. You cannot have everything in life.


    Crosland, Which algorithm has quick response?
     


    It would be better if you answered my question, rather than just asking another.
     
    Any algorithm can have a quick response but you need to decide on the parameters that are acceptable to you.
     
    As you have already found, a quick response will lead to a high current (torque) demand as the motor accelerates.


    Sorry, didn't get your question. I use AN1078, and the algorithm based AN1078.
    #12
    michael.li9
    New Member
    • Total Posts : 25
    • Reward points : 0
    • Joined: 2008/08/27 12:57:14
    • Location: 0
    • Status: offline
    Re: speed response time of AN1078 2019/04/16 15:17:13 (permalink)
    0
    NKurzman
    No algorithm is going to give you fast response and low current.  Software does not change the laws of Physics.


    OK, Rather select a fast response. But do you have any data how fast it can archived if don't care about current.
     
    #13
    NKurzman
    A Guy on the Net
    • Total Posts : 17610
    • Reward points : 0
    • Joined: 2008/01/16 19:33:48
    • Location: 0
    • Status: online
    Re: speed response time of AN1078 2019/04/16 16:30:53 (permalink)
    0
    That is up to your motor. Software can not make it move faster.
    #14
    PStechPaul
    Super Member
    • Total Posts : 2348
    • Reward points : 0
    • Joined: 2006/06/27 16:11:32
    • Location: Cockeysville, MD, USA
    • Status: offline
    Re: speed response time of AN1078 2019/04/26 01:41:36 (permalink)
    0
    Most induction motors have a locked rotor or peak current (and peak torque) about 3-4 times nominal. The time for acceleration from 900 RPM to 1800 RPM depends on the inertia of the components connected to the shaft, as well as the actual load.

     
    #15
    Jump to:
    © 2019 APG vNext Commercial Version 4.5