Can someone look at the specs on this MOSFET for me?

[Andy1845c]

Can someone take a peek at the datasheet for this mosfet and tell me if there is a reason it won't work to drive a small load (leds) directly off the I/O port of a PIC?

http://www.datasheetcatalog.com/datasheets_pdf/I/R/F/5/IRF510.shtml

I'm putting +5v on the gate and just under 9v scourc to drain and trying to light a small LED on a protoboard as a rough draft of a circuit. I can get it to glow dimly and my multimeter only reads a few volts at the LED where I expected to get around my 9 volts.

I have never used MOSFETS before and they are confusing the heck out of me! I have been googling them for a few days, I see some circuits use transistors to drive them, but I thought logic level fets could be driven off I/O ports directly.

Does the current from scource to drain make a differance in how it turns on?

 

[kchriste]

You should use a logic level FET for this. The IRF510 is NOT one. If your FET was made on a Friday afternoon then it will only switch 250uA with 4V of gate voltage.

 

[Andy1845c]

Ugggg.....:::bangs head on desk::: How did you figure the 250uA? Is it fairly simple or is it going to make my head hurt?

For some reason I thought these were logic level fets whe I ordered them like 6 months ago.....

What is a good logic level fet? I remember I had trouble finding a good choice in the Jameco catalog.

Thanks for your reply though, I apperciate it.

 

[Sceadwian]

You could 'patch' the problem by using a small transistor to drive the gate from the MOSFET's voltage source. kchriste is right though the IRF510 is horrible for logic level drive (I have a few I know) It's easier to replace the IRF510 with a logic level FET. The IRF510 only just begins to conduct at 4 volts. 5 volts or likley slightly less from a micro controller is only going to cause it to switch on in the very earliest part of it's amplification curve, so if you try to draw too much current from the fet it's effective series resistance increases fast.

 

[Andy1845c]

I thought about the transistor idea, but it seems to almost defeat the purpose of the MOSFET all together in this project. Plus, space in the enclosure is tight and I want as few components as possible.

I guess i'll just have to save the 510's for some other project.

 

[kchriste]

Ugggg.....:::bangs head on desk::: How did you figure the 250uA? Is it fairly simple or is it going to make my head hurt?

Besides max drain to source voltages and currents there are a few keys things to look at when selecting a MOSFET for simple DC switching applications:
1) VGS(th) or Gate to Source Threshold Voltage. This parameter will tell you the minimum and maximum voltage (best and worst case) at which the FET will just start to turn on. Here they'll also define what they mean by "just turn on" by stating how much drain current will flow (In your FETs case 250uA)
2) RDS(on) or Resistance between the Drain and Source for a given gate-source voltage and drain current. This basically tells you how to switch the biggest load that the FET is designed for. Here they tell you the gate voltage needed to switch the recommended maximum drain current with minimum power wasted in the FET. Your IRF510 needs 10V to switch 3.4A and the FET will act like a 0.54hm: resistor or lower when turned on.
3) VGS Voltage Gate to Source. This is the max voltage you can apply to the gate before damaging the FET. Your IRF510 will take a max of 20 of either polarity relative to the source terminal.

EDIT: As for logic level FETs, Jameco has the IRLZ44NPBF which can switch 21A at 55V with 4V of gate voltage. It sells for $0.59 in a TO-220 case. I found Jameco's search feature to be pretty lame. Try DigiKey's and then see if Jameco has the same if you prefer them. IRF's site has a pretty good search feature also.

 

[Sceadwian]

Andy, this is one of the reasons why IGFET's are not common in small LED driving circuits. Nowdays if I'm not mistaken most MCU output stages are mosfet and suffer the same fate. While the maximum drive current is limited to a low value due to VGS(th), the output impedance is much much lower for a comparable load.

 

[Andy1845c]

What I am trying to do, and I have another topic on this on here, is switch 8 400mA loads from a PIC. Darlingtons were my first though, but I have little wiggle room on the voltage. I want to drive the loads as close to 9volts as a LM350 will let me with a 12-14v supply (car). I was leaning toward mosfets because I understood they have less voltage loss then Darlingtons. If I lose 1.4 volts across the darlingtons, my project won't work.

Maybe i'll play with the darlingtons tonight and see what happens.

Logic level fets are expensive!

 

[Sceadwian]

How about using a comparator? The comparator can run from the motor supply lines and you just feed it's input from the PIC and a simple voltage divider set at about 2 volts, an opamp could be used in a pinch.

 

[Andy1845c]

How about using a comparator? The comparator can run from the motor supply lines and you just feed it's input from the PIC and a simple voltage divider set at about 2 volts, an opamp could be used in a pinch.

A comparator? Not a part I have worked with yet. can you point me toward a schamatic that uses a comparator in this fasion?

My brain is mush tonight. I've been washing out a hog barn for the past 15 hours.

 

[kchriste]

Maybe use something like the MC33298P instead. It'll save some IOs on the PICs but probably won't be cheaper than a bunch of logic level FETs. You could also use plain old NPN transistors such as the 2n2222 and drive them with apx 15ma each from the PIC. It's bending the rules a bit (Rule of thumb is estimate a beta of 10 for saturation) assuming the beta of the transistors used is better than 27 at 400ma.

 

[Sceadwian]

Not so good with schematics =) Comparators are pretty simple. A common comparator will have power supply lines, VCC and GND and three other pins. input, output, and reference. When the signal going to the input is above the reference voltage the output is high, when it's bellow the reference the output is low, that's it. The easiest way to do this would be to use an equal resistor voltage divider from the MCU's power supply lines, basically setting the refrence at half supply for the MCU. But using VCC from the supply that drives the FET, the comparator is going to have better drive characteristics than a single transistor driver.

 

[Andy1845c]

Okay, i'm a bit confused.

With the comparator, I would use the comparator to drive the MOSFETS, right?

What exactly can the MC33298P do? Can it directly drive a load, or is it made to drive a fet or transistor? There is so much info on the datasheet for it I was confused on what exactly it even is. It looks like I would have to be a better programmer then I am in order to make my micro work with it.

 

[kchriste]

With the comparator, I would use the comparator to drive the MOSFETS, right?

Yes, you could use it that way. You can't drive the 400ma load directly with it.

What exactly can the MC33298P do? Can it directly drive a load, or is it made to drive a fet or transistor?

It is like a fancy high powered shift register that uses the SPI protocol and has 8 logic level open drain FETs built in. It will directly drive loads up to 1Amp per output. There is a max limit depending upon how many outputs are active and what the ambient temperature is. Figure 21 at https://www.ortodoxism.ro/datasheets/motorola/MC33298P.pdf shows that you can have 8 outputs drawing 0.5Amp continuously if the ambient temperature is up to 100C but could draw 1.0 Amps out of 8 outputs at a chilly -50C if using a DIP package.
There may be simpler types of chips available. I just used it as an example.