On 9/22/05, Damon Hamm <damonhamm@gmail.com> wrote:
> I am attempting to fire 8 solenoids from a 16F628 and am inexperienced
> at playing with solenoids.
>
> The setup:
> One 16F628 connected to an optoisolator (necessary?) then to 8 power
> MOSFETs (with necessary resistors in tow).  8 mini 12V push solenoids
> (with flyback diodes) on the high side of the MOSFETs.
>
> Upon a trigger, I want to fire the solenoids (sometimes all 8 at once)
> and hold for 2 seconds or so.
>
> I am having somewhat success using PWM to control the initial current
> then drop it down once the coil is saturated to avoid burnout.
> Schematic:
> http://www.damonhamm.com/other/pic_solenoid_schematic1_pwm.gif
> (Initial tests do not utilize the #1 RC 'decoupling' circuit.)
>
> But, the MOSFET (1.5A? in a TO-220 case) heats up to be quite hot to
> the touch, which I wouldn't consider viable.  Using 5V at the MOSFET
> gate (full on or PWM at 100%) doesn't produce enough current to fire
> the solenoid, while using 12V with PWM at 20% for 100ms, then at %1 to
> hold for 2s, still causes everything to heat up much too much after
> several cycles.  I am still surprised that a duty cycle of 1:255 (~1%)
> still provides too much current!  Is there something wrong with my
> setup or are the MOSFETs just too wimpy?  I have some 6Watt DIP
> MOSFETs on the way to test out.
>

Damon,

I think the MOSFET shouldn't be getting that hot.  What MOSFET is it?

You can calculate the power dissipation of the MOSFET. Look up the on
resistance in the datasheet and multiply it by the current squared:
(I^2)R  Then multiply that by the duty cyctle of the PWM to get the
average dissipation.  If this figure is too high (more than half a
watt, say, if you're not using a heatsink), you can choose a MOSFET
with lower R.   This will mean choosing one with a higher current
rating than you would think you need.

The opto is not necessary.  The R/C is a neat idea, but I think the
voltage on MOSFET gate will be swinging between -2.5 and +2.5 V. 
That's probably not going to be enough to turn it "on" all the way. 
(I'm assuming the PIC is running on +5V).   I'd just put about a 10K
resistor between the PIC and the gate.

Actually, 5V may not be enough to turn some MOSFETs "on" all the way. 
You have to check the curves on the datasheet or just choose a MOSFET
with "logic-level" inputs.

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
  - fortune cookie

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist