Author Topic: PWM current draw Formula  (Read 3112 times)

0 Members and 1 Guest are viewing this topic.

Offline macdad-Topic starter

  • Robot Overlord
  • ****
  • Posts: 131
  • Helpful? 5
PWM current draw Formula
« on: July 27, 2010, 12:12:52 PM »
Hey everyone,

For a few of my projects as well as upcoming robots I have been looking a PWM as a viable option for Motor Driving. The motor driver chips I'm using are the IXDN404PI chips with a peak(pulsed) current of up to 4 Amps.

My question is, how/what is the formula for determining current draw/maximum current output with PWM compared to Continuous current?

Thanks,
Nick(macdad-)

Offline dunk

  • Expert Roboticist
  • Supreme Robot
  • *****
  • Posts: 1,086
  • Helpful? 21
Re: PWM current draw Formula
« Reply #1 on: July 27, 2010, 04:42:18 PM »
so a quick glance at the IXDN404PI datasheet under the "Electrical Characteristics" section reveals:
IDC   Continuous output current   1 A
     
in reality it's not that simple though.
you will only get 1A out of it if you manage to keep it cool enough.
pay particular attention to anything the datasheet has to say about heatsinks and cooling.


dunk.

Offline macdad-Topic starter

  • Robot Overlord
  • ****
  • Posts: 131
  • Helpful? 5
Re: PWM current draw Formula
« Reply #2 on: July 27, 2010, 06:45:31 PM »
Alright, thanks dunk, I couldn't find anything on the power dissipation for the PDIP package of the IXDN404PI, but I'll do some trial and error to find a good threshold current since I won't be using much cooling for the chip.

But back onto my original question:
What is the formula for determining current draw of a PWM circuit?

For example, If I was using PWM on an LED, with the PWM Frequency at 1KHz and it's duty cycle at 50%, what is the current draw?

Offline dunk

  • Expert Roboticist
  • Supreme Robot
  • *****
  • Posts: 1,086
  • Helpful? 21
Re: PWM current draw Formula
« Reply #3 on: July 27, 2010, 08:10:50 PM »
For example, If I was using PWM on an LED, with the PWM Frequency at 1KHz and it's duty cycle at 50%, what is the current draw?
so at low frequency you would still need to spec the circuit for the full current draw of the LED.
but presumably as the Frequency increases you could get away with drawing more current and taking the average right?

in your example of 50% duty cycle at 1KHz the driver circuit for the LED should be able to handle more current than the "Continuous output current" figure.
the exact formula would be dependent on the various component's ability to recover. in particular, the driver component and the LED it's self.

for a MOSFET driver i was expecting to see a graph in the IXDN404PI datasheet that plotted max current/frequency but it is not there....
i'm not even seeing the time limit you can use the "Max current draw" value for.
i suspect the answer can be extrapolated from the "Supply Current vs. Frequency" graph but i'm afraid i don't have the skills.

maybe Soeren would know...


dunk.

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: PWM current draw Formula
« Reply #4 on: July 31, 2010, 06:51:42 AM »
Hi,

Quote
But back onto my original question:
What is the formula for determining current draw of a PWM circuit?

For example, If I was using PWM on an LED, with the PWM Frequency at 1KHz and it's duty cycle at 50%, what is the current draw?
Generally, it's quite straightforward to calculate the average current draw, as it is just the PWM percentage of the max. current. At 50% duty cycle it's half, at 10% it's 10% and so on, but... It still needs to be able to handle the max. current.

A regular 20mA LED can be used with that "formula", increasing the current, but don't go below 2% duty cycle (with an average draw of 20mA, i.e. 1A pulses), as the bonding wires will either melt or rip off of the crystal with too high currents, even if the average is low.

The IXDN404PI is a MOSFET driver, not a motor driver and the 4A will only be sustained some tens of ns and only with a large enough potential. The max. peak pulse current is very dependant on the voltage potential (in all drivers). At 1 kHz, the chip won't be able to sustain anywhere near 4A and, as dunk mentioned, even the 1A is dependant on your ability to keep it's cool.

For a simple, but quite effective, heatsink, a bit of aluminum curtain rail can be glued onto the IC with a small drop of super glue.

If something reliable (for driving motors) is wanted, use a motor driver instead.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline macdad-Topic starter

  • Robot Overlord
  • ****
  • Posts: 131
  • Helpful? 5
Re: PWM current draw Formula
« Reply #5 on: July 31, 2010, 10:56:22 AM »
Thanks Soeren. Didn't think of the percentages.

And the IXDN404PI was the best driver chip I could find, and for one in a DIP package. Even if it's not a motor driver chip, I won't be drawing more than 500mA at Max from the motors I'll be using with it. I'll definitely take into consideration the Aluminium rail, but it does seem odd that your recommending Aluminium, as isn't it normally a thermal insulator? But hey, can't judge until I try.

I'll look through for specific Motor Drivers.

Offline Razor Concepts

  • Supreme Robot
  • *****
  • Posts: 1,856
  • Helpful? 53
Re: PWM current draw Formula
« Reply #6 on: July 31, 2010, 11:26:04 AM »
Aluminum is excellent for heatsinking and is the material that is used for many many heatsinks (ie the ones in your computer are generally aluminum)

 


Get Your Ad Here

data_list