California
Offline
Sr. Member
Karma: 2
Posts: 409
|
 |
« on: May 21, 2012, 02:05:21 am » |
I need to switch about 14v (before current limiting resistor) at about 490mA (actual calculated number without buffer) at PWM speed for one channel of LEDs. What are my best options, and what am I looking at for heatsinking needs?
Thanks
|
|
|
|
|
Logged
|
|
|
|
|
Sydney
Offline
God Member
Karma: 14
Posts: 717
Big things come in large packages
|
 |
« Reply #1 on: May 21, 2012, 02:31:14 am » |
I would say most tranhsistors will do the job.
|
|
|
|
|
Logged
|
|
|
|
|
Manchester (England England)
Offline
Brattain Member
Karma: 277
Posts: 25506
Solder is electric glue
|
 |
« Reply #2 on: May 21, 2012, 02:36:43 am » |
A logic level FET should do that without the need for a heatsink.
|
|
|
|
|
Logged
|
|
|
|
|
|
|
Manchester (England England)
Offline
Brattain Member
Karma: 277
Posts: 25506
Solder is electric glue
|
 |
« Reply #4 on: May 22, 2012, 04:02:17 am » |
Yes that one will turn on quite well with a logic level signal.
|
|
|
|
|
Logged
|
|
|
|
|
Anaheim CA.
Offline
Edison Member
Karma: 31
Posts: 2310
Experienced old Whitebeard with a Full head of Hair...
|
 |
« Reply #5 on: May 23, 2012, 04:19:14 pm » |
If one reads the data sheets for each, the relevant characteristic is Vgth and from the charts for that parameter the devices are similar and both are easily capable of > 7A Drain current with 3.5V Vgs and as 'best part' the IRL520 is capable of about 20% more current for the same level of gate enhancement, further the '3055 needs to be soldered to a heat sink whereas the IRL520 is a standard TO-220 case and thus (if required) is easier to mount. Were it my choice I would use the '3055 part because the design requirements ~.5 A wouldn't require a heatsink to be used and the "D-PAk" part is easier to use (smaller) and thus more devices will fit in the same PCB area. FYI, the 'Older' IR parts the L in the part number means logic level drive. IMO
Doc
|
|
|
|
|
Logged
|
“The solution of every problem is another problem.” -Johann Wolfgang von Goethe
|
|
|
|
0
Offline
Tesla Member
Karma: 71
Posts: 6624
Arduino rocks
|
 |
« Reply #6 on: May 27, 2012, 07:29:57 pm » |
super-beta NPN switching transistor will do nicely too (no heatsink) - go for Vsat <= 0.1V at Ic >= 1A.
What PWM clock rate were you wanting BTW?
|
|
|
|
|
Logged
|
|
|
|
|
California
Offline
Sr. Member
Karma: 2
Posts: 409
|
 |
« Reply #7 on: May 27, 2012, 11:43:55 pm » |
I don't know, whatever the arduino rate is. I was thinking about using the Arduino Tiny Library with an attiny84, is that the same rate?
|
|
|
|
|
Logged
|
|
|
|
|
California
Offline
Sr. Member
Karma: 2
Posts: 409
|
 |
« Reply #8 on: May 28, 2012, 12:02:05 am » |
I'm not really sure how to find a super-beta NPN, I'm using Newark.com and it has no filter for saturation voltage.
|
|
|
|
|
Logged
|
|
|
|
|
California
Offline
Sr. Member
Karma: 2
Posts: 409
|
 |
« Reply #9 on: June 08, 2012, 03:02:46 am » |
I just learned that I will require the IRL520 to sink about 1.5A at 5ish volts... How do I calculate how much power it will dissipate and whether or not I need a heatsink?
Thanks
|
|
|
|
|
Logged
|
|
|
|
|
0
Offline
Tesla Member
Karma: 71
Posts: 6624
Arduino rocks
|
 |
« Reply #10 on: June 08, 2012, 02:16:59 pm » |
For a MOSFET that is properly switched on the dissipation is I-squared-R, where R is the Ron value. For the IRL520 Ron = 0.27 ohms and at 1.5A that means the power will be 0.6W. A small heatsink will be needed probably.
When off the current is tiny so dissipation can be ignored. When switching the dissipation is much higher though, so you have to look at that if using high-speed PWM.
There are lots of far superior MOSFETs these days, with Rds(on) of 0.01 ohms or less, which wouldn't even get warm: 1.5^2*0.01 = 22mW
Hint - if you don't need a high voltage device the lower voltage devices will have lower Rds(on) for the same price.
|
|
|
|
|
Logged
|
|
|
|
|
California
Offline
Sr. Member
Karma: 2
Posts: 409
|
 |
« Reply #11 on: June 08, 2012, 05:26:08 pm » |
Hmm ok how about this one? http://www.newark.com/international-rectifier/irlb8721pbf/n-channel-mosfet-30v-62a-to-220/dp/10R3513I'm still not sure what spec tells me whether it is logic level or not... Its Rds(on) is less than .01ohms. The threshold is 1.8v, which I am assuming is the voltage required to switch it on so I need it to be less than 5v, correct? Can I switch it directly with an arduino pin? When you say that the dissipation is higher when switching, do you mean that actually switching off/on, on/off requires more dissipation than just having it switched on, or did you mean more dissipation switched on? I'm having a pretty hard time completely understanding these things, I appreciate your help.
|
|
|
|
|
Logged
|
|
|
|
|
Manchester (England England)
Offline
Brattain Member
Karma: 277
Posts: 25506
Solder is electric glue
|
 |
« Reply #12 on: June 09, 2012, 06:22:46 am » |
The threshold is 1.8v, which I am assuming is the voltage required to switch it on so No that is the voltage when the FET starts to turn on. You need to look at the on resistance when you apply 5V. In this case this is a logic level FET.
|
|
|
|
|
Logged
|
|
|
|
|
United Kingdom
Offline
Faraday Member
Karma: 131
Posts: 4674
|
 |
« Reply #13 on: June 10, 2012, 03:16:13 pm » |
For switching 490mA I would use a ZTX851 transistor, base driven from the Arduino through a resistor of about 220 ohms. Simpler (1 less resistor) and cheaper than a logic-level mosfet. But for switching 1A, I would use a mosfet.
[EDIT: I might be wrong about it being cheaper, depending on your supplier. However, for switching 490mA, a BC337 is just about adequate - and they really are cheap.]
|
|
|
|
« Last Edit: June 10, 2012, 03:19:29 pm by dc42 »
|
Logged
|
Formal verification of safety-critical software, software development, and electronic design and prototyping. http://www.eschertech.com
|
|
|
|
Anaheim CA.
Offline
Edison Member
Karma: 31
Posts: 2310
Experienced old Whitebeard with a Full head of Hair...
|
 |
« Reply #14 on: June 11, 2012, 01:49:58 pm » |
There is one thing not yet considered here and that is that 'pulsed' or PWM control heating effects are "Averaged" in the transistor (Mosfet or Bi-Polar) and dependent on duty cycle which is generally much lower than the steady state 'On' current. The Best choice IMO is the Lowest Rdson and least expensive Logic Level Mosfet you can find. Simple? Yes, IMO
Doc
|
|
|
|
|
Logged
|
“The solution of every problem is another problem.” -Johann Wolfgang von Goethe
|
|
|
|
|