Author Topic: mosfet pwm  (Read 4118 times)

0 Members and 1 Guest are viewing this topic.

Offline made2hack

  • Frequent Contributor
  • **
  • Posts: 399
  • Country: 00
    • Made2Hack
mosfet pwm
« on: June 10, 2015, 05:31:06 am »
Hello,

So I want to use a mosfet to run a high power / high voltage led 31-36V @ 350mA - 1100mA.

I've chosen a KSP13 NPN transistor, a IRF830 Mosfet, a Nichia LED and some resistors. The power supply is 36VDC.

Here's the schematic.


Here are the datasheets:
Fairchild KSP13
IRF830

Some questions:
1 - Am I keeping the Vgs < 20 V by using the top 10K resistor and the bottom 10k + npn as voltage divider?
2 - The max Vce for the NPN is 30V, is this being exceeded?
3 - Will this work?


Offline TimFox

  • Frequent Contributor
  • **
  • Posts: 866
  • Country: us
  • Retired, now restoring antique test equipment
Re: mosfet pwm
« Reply #1 on: June 10, 2015, 05:36:47 am »
1.  When the transistor is ON, the two resistors should give about 18 V at the MOSFET gate.  However, this requires a drive voltage to the bipolar of at least 20 V, since the emitter would be at 18 V.  Probably better to ground the emitter and put R2 from the collector to ground.
2.  Yes, when the transistor is OFF, there would be 36 V from collector to emitter.  If you ground the emitter and put R2 from collector to ground, then the maximum voltage would be only 18 V.
3.  With these two suggestions, it might work.  It depends on the switching frequency and the device capacitances.
 

Offline PSR B1257

  • Regular Contributor
  • *
  • Posts: 241
  • Country: de
Re: mosfet pwm
« Reply #2 on: June 10, 2015, 05:37:47 am »
That will not work.
With the 63 Ohm resistor at the source of the MOSFET the voltage drop even at 350mA is too high.

You don't need the BJT to drive the gate. All you need is a series resistor to the gate and a 15V Zener diode from gate to GND (with no source resistor). This will work for frequencies <10kHz or so.

Quote
IRF830
Oh dear... This sucker won't stay cold even at a lousy 1.1A.
« Last Edit: June 10, 2015, 06:13:37 am by PSR B1257 »
In theory, there is no difference between theory and practice. But, in practice, there is.
 

Offline TimFox

  • Frequent Contributor
  • **
  • Posts: 866
  • Country: us
  • Retired, now restoring antique test equipment
Re: mosfet pwm
« Reply #3 on: June 10, 2015, 05:40:16 am »
My bad:  I didn't notice the 63 ohm value on R4.
If this is PWM, what peak current for what duty factor do you want?
 

Online Simon

  • Global Moderator
  • *****
  • Posts: 12578
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: mosfet pwm
« Reply #4 on: June 10, 2015, 07:04:41 am »
R4 should be the other side of the resistors whatever it's value. You would be best to use a zener in parallel with the mosfet gate and pull that high with a resistor, then ground the gate with the BJT or use a high side switch to the gate with a resistor devider so that when not powered it pulls low else it pulls high based on the resistor ratio, what you have will put 36V into the mosfet.
https://www.simonselectronics.co.uk/shop
Varied stock of test instruments and components including EEVblog gear.
Also, if you want to get ripped off: https://www.ebay.co.uk/usr/simons_electronics?_trksid=p2047675.l2559
 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 1824
  • Country: us
Re: mosfet pwm
« Reply #5 on: June 10, 2015, 09:30:52 am »
That will not work.
With the 63 Ohm resistor at the source of the MOSFET the voltage drop even at 350mA is too high.

You don't need the BJT to drive the gate. All you need is a series resistor to the gate and a 15V Zener diode from gate to GND (with no source resistor). This will work for frequencies <10kHz or so.

Quote
IRF830
Oh dear... This sucker won't stay cold even at a lousy 1.1A.

Yep. I'd use this method, and IRF3205.
The easiest person to fool is yourself. -- Richard Feynman
 

Offline made2hack

  • Frequent Contributor
  • **
  • Posts: 399
  • Country: 00
    • Made2Hack
Re: mosfet pwm
« Reply #6 on: June 10, 2015, 09:36:38 pm »
Ok,

so did I get this right?



Or do I still need to pull down the Gate separately?

Online Simon

  • Global Moderator
  • *****
  • Posts: 12578
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: mosfet pwm
« Reply #7 on: June 10, 2015, 09:37:44 pm »
put a 100K resistor across the gate to ground
https://www.simonselectronics.co.uk/shop
Varied stock of test instruments and components including EEVblog gear.
Also, if you want to get ripped off: https://www.ebay.co.uk/usr/simons_electronics?_trksid=p2047675.l2559
 

Offline Andy Watson

  • Super Contributor
  • ***
  • Posts: 1636
Re: mosfet pwm
« Reply #8 on: June 10, 2015, 09:42:50 pm »
Put a resistor between the fet and the led.
 

Offline made2hack

  • Frequent Contributor
  • **
  • Posts: 399
  • Country: 00
    • Made2Hack
Re: mosfet pwm
« Reply #9 on: June 10, 2015, 10:57:26 pm »
@andy - does it matter if I put the resistor from source to gnd? or does it hve to be on cathode of led to drain?

Offline Andy Watson

  • Super Contributor
  • ***
  • Posts: 1636
Re: mosfet pwm
« Reply #10 on: June 10, 2015, 11:08:54 pm »
If you put  the resistor in the source it will work against the control voltage on the gate - I think this is not what you desire in this case - i.e. you want the fet to operate as a switch - keep it out of its linear region to reduce its power dissipation.  Put the resistor in the drain circuit to limit the maximum current through the LED - in the anode or cathode, either will do.
 

Offline PSR B1257

  • Regular Contributor
  • *
  • Posts: 241
  • Country: de
Re: mosfet pwm
« Reply #11 on: June 11, 2015, 12:55:48 am »
Quote
so did I get this right?
Yes. But if you really have a PWM with an amplitude of 36V you have to choose a higher value resistor, 10k for instance.
What frequency are you applying?
In theory, there is no difference between theory and practice. But, in practice, there is.
 

Offline made2hack

  • Frequent Contributor
  • **
  • Posts: 399
  • Country: 00
    • Made2Hack
Re: mosfet pwm
« Reply #12 on: June 11, 2015, 01:48:19 am »
Quote
so did I get this right?
Yes. But if you really have a PWM with an amplitude of 36V you have to choose a higher value resistor, 10k for instance.
What frequency are you applying?

Amplitude will not be 36V. I did not bother drawing the PWM schematic. It will either be around 12V if I use a 555 or 3.3V if I use an MSP430.

Frequency? I don't really know. Whatever a 555 chip will do in terms of PWM in the range of duty cycle from 0% - 100% ?
As for using the MSP430, don't know what duty cycle. Whichever dims the LED properly  :-\



Offline PSR B1257

  • Regular Contributor
  • *
  • Posts: 241
  • Country: de
Re: mosfet pwm
« Reply #13 on: June 11, 2015, 02:26:50 am »
Quote
Amplitude will not be 36V.
O.k. Misunderstanding...

Quote
It will either be around 12V if I use a 555 or 3.3V if I use an MSP430.
In this case you don't even need the Zener. You can drive the MOSFET directly (100..1k gate resistor to prevent too high surge currents to/from the gate) from 12V. 3.3V however is a bit of a challenge.
If you gor for a IRLZ44N you're fine. This MOSFET can be driven from VGS=2.5 to 15V

Quote
Frequency? I don't really know. Whatever a 555 chip will do in terms of PWM in the range of duty cycle from 0% - 100% ?
Since you tell the 555 at frequency it has to work, you should know  ;)
1kHz will do the trick.
« Last Edit: June 11, 2015, 02:31:57 am by PSR B1257 »
In theory, there is no difference between theory and practice. But, in practice, there is.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf