[Prev][Next][Index][Thread]

RE: Motor PWM

Title: Message
The power chaps have some interesting dodges where 'flywheel' circuits can improve efficiency.  These are usually associated with current sensing loops, though.  But even with a simple H-bridge, the diode path for the current in an inductive load can actually recharge the supply battery! (Though not a lot!!)
 
Many years ago we developed a stepper drive method in Portsmouth that used a 30 volt 'speed up' to ramp up the current quickly.  The 30 volt supply was generated and sustained from the spike when the current was broken!
 
Cheers
 
John
-----Original Message-----
From: Derek Hall [mailto:derek@micromouse.me.uk]
Sent: Wednesday, 17 September 2003 7:51 PM
To: micromouse@cs.rhul.ac.uk
Subject: RE: Motor PWM

Thanks John, your final paragraph explains exactly how my motor drive works at the moment. A timer creates a 1 ms interrupt that turns on the motor and sets a second timer to a value between 0 and 1 ms that turns it of again. the calculations for the required power are performed during the first interrupt. my question is really is there anything to be gained from repeating the cycle to the motor several times before recalculating the command value? also the motor makes a lot of noise at this frequency that cant be efficient can it?
 
thanks
 
Derek
-----Original Message-----
From: owner-micromouse@cs.rhul.ac.uk [mailto:owner-micromouse@cs.rhul.ac.uk] On Behalf Of John Billingsley
Sent: 17 September 2003 02:48
To: micromouse@cs.rhul.ac.uk
Subject: RE: Motor PWM

It's easy to ague that the principle of PWM is wrong!
 
It introduces another variable into the system - the phase of the cycle when the decision is made.
 
Instead make a 'bang-bang' decision as often as possible (strictly bang-freewheel).  If the present situation calls for a drive that is less than maximum, the 'decision' will flap to and fro as the speed is seen to change - and you will effectively have PWM after all!  Of course you want the decision loop to run as fast as practical.
 
Another way to lock the drive to the computation cycle is to let each 'decision' determine the proportion of the time that the drive should be on before the next decision - so that the result of each computation is a single pulse of drive, locked to the data cycle and varying from never on to full on. 
 
Cheers
 
John
 
 
-----Original Message-----
From: Derek Hall [mailto:derek@micromouse.me.uk]
Sent: Wednesday, 17 September 2003 7:32 AM
To: micromouse@cs.rhul.ac.uk
Subject: Motor PWM

Thanks to every one who made Minos03 such a successful event. I enjoyed it enormously!
 
Anyone got any views on the best frequency to run a PWM drive for an ironless rotor motor. I currently run at 1Khz and was looking to increase that but it has been brought to my attention that others use lower frequencies of the order of 100 to 300 Hz. What are the pros and cons of higher or lower frequencies?
 
Thanks
 
Derek.