AVR-Chat

Hi Wagner -
I have been driving power FETs in 2 projects this past month, both
with lower voltage and power than yours.  I'm down in the 3V / 1A
region, and have had good results with the Fairchild FDG311N
and FDG312P, which were used in one project as a bridge driven
directly from a PIC18, and in the second project, only the 311N
is driven, for powering a laser diode; contolled from a Tiny26.

The LD needs current sensing and feedback, but little power
can be spared for the sensing resistor.  I thought that I was
cutting it too close to zero ohms with my .1 ohm resistor, but
you're even less with 50 mohm.  I would think with higher voltages
and a bigger power budget that you would use a higher value resistor.
These smaller R's are hard to use because the effective resistance changes
so much (proportionately) because of board placement and soldering
variations, etc.  My resistor is placed between the FET and the LD.

I use use the differential x20 A-D inputs on the tiny26 to measure the current
during pulses, and feedback a very slow correction voltage to the LD supply.

Ken Holt




You Wrote:


Hi, I am building a 4.2V @ 6A stepper motor (6 wires - unipolar) driver,
using the IRL530N or IRLZ24 power N FET.  I actually using the Alegro 2398
(twin bridge bipolar driver) dedicated chip for small motors, that chip
doesn't stand such 6A current.
  
As a protection scheme, I am designing a feedback current limit for the
FETs, so it won't fry the stepper in case of a mechanical lock or even
during steady holding current.  The main idea is to measure a 0.05 ohm
resistor small voltage drop, connecting FETs drain to ground.  The small
vdrop is to be compared with an adjusted trimpot voltage at a LM339
comparator.  The comparator will cut off the respective FET that had
overcurrent. This approach is being used with the 2398, since it has an
enable/disable input per side of each bridge.
  
I already came up with several possibilities, using opto-couplers, or PNP
or P Channel (BS110) small switching transistor to fire the FETs,
everything under control of an AVR of course, but also with the current
limit control by the LM339 and additional circuitry.
  
Then I thought about the possibility to use some already small dedicated
chip for the job of firing the external FET.  Something that has current
limit sensing, that I can connect to the FET drain resistors, and will cut
me some slack in the board and make things better.
  
Perhaps some stepper motor driver, or a multiple solenoid or small motor
driver, enough to have the output as open collector or floating emitter,
enough to deal with a pull-up or pull-down resistor at the gate of the
IRL530N.
  
Of course that those small dedicated chips that has "step forward,
backward, etc" do not serve, since all the control is being done already by
the AVR.  All I need is something that can deal with higher voltage at
output (20V) and has voltage sensing comparator for current limiting. 
  
I always have the other solutions, using LM339, etc, but why not try
something ready...?
  
  
  
+AVRDrive         +5V   +12V    +5V
-------.           |      |      |
       |           |      |      |
     R10k          |     R1k    COIL
       |        |¯¯¯¯¯|   |      |
 300mV o-----(+)|     |   |   |--'  IRF530 
       |        | LM  |---o---|--.  IRL530N    
     R620       | 339 |       |<-|  IRLZ24
       |   .-(-)|     |          |  w/clamp diode
      _|_  |    |_____|          |
           |                     |
           '---------------------o 
              300mV              |
                               R0.05
                                 |
                                _|_
  
R620 could be a trimpot adjusting full span current of 6A over R0.05.
One LM339 houses 4 comparators, so one chip per unipolar motor (4 phases &
FETs).
  
Any known dedicated IC to substitute the LM339 with advantages?
  
Thanks.
Wagner.