ESP8266 Community Forum

- Sun Mar 20, 2016 9:42 am #43626 Now that you understand the principle of looking for a suitable transfer characteristic among the other parameters you understand, you are well-equipped to choose your own part(s) on the basis of price/availability/package etc.

I hope you also understand the consequences of using such high-current parts. You say 16Amps are not enough yet even this "small" amount of current is not trivial to handle properly. You might want to consider a modular approach that includes a dedicated Mosfet driver located close to the actual transistor. With large devices the higher gate charge has to be overcome within a sufficiently small amount of time such that the device is not lingering in the linear (half-on, half-off) region for too long.

To work out what's required look at the Gate charge parameter and establish the output impedance of the GPIO pin driving the Gate. The CR time-constant product then has to be taken into account - especially if you intend to use PWM control. If the PWM cycle time is comparable to the CR product your Mosfet will speed too long in the linear region.

- Sun Mar 20, 2016 9:47 am #43627 Of course the price is high if purchased from mouser/digikey, etc.
But if you're not scared by chinese clones, they are about $0.60 on eBay.

For smaller ones, in TSOP-6, there is the AO6408 with 7A drain current along with Vgs threshold below 1V.

- Wed Mar 23, 2016 8:47 am #43875 as i still in my way to get out the confusion from my head i get more details makes me more confused
i have read this question and the highest ranked answers says

Code: Select all The 12V and 6A is a good starting point. This tells me you need a mosfet with a max drain-source voltage capability greater than 12V so 20V would be a minimum criteria for this.

You want to switch 6A and you'll want it to do so with minimum volt-drop - just like a relay contact so you are looking for Rds(on) below (say) 0.1 ohms. This means at 6A it will develop a small voltage across the device of 0.6V (ohms law).

However, that will produce a power disippation of 6 x 6 x 0.1 W = 3.6W so if you are looking for a surface mount device you would prefer a lower disippation of maybe 0.5W max.

This means Rds(on) would be more like 0.014 ohms.

So far, your application needs a 20V transistor, capable of switching 6A with an on resistance no more than 0.014 ohms.

Vgs is "like" the coil voltage on a relay - it's how much voltage you need to apply to the coil to get it to switch BUT for a FET it's a linear thing and, if you don't apply enough voltage, the mosfet will not turn on properly - its on-resistance will be too high, it'll get warm under load and have a volt or two across it when you want a nice low resistance.

You then need to inspect the details of the spec to see how much you need to apply to guarantee the low on-resistance you want. A bit more on this further down.

The IRFZ44N has on the front page of the data sheet: -

Vdss = 55V, Rds(on) = 17.5 milli ohms and Id = 49A

It's not a surface mount device therefore a little more heat generated isn't going to matter too much (with a heatsink) so it'll do what you want it to do but I'd research a device with smaller Vds (say 20V) and you'll probably find one with a lot less than 10 milli ohms on resistance.

If you look at the electrical characteristics on page 2 you'll see that the 17.5 milli ohms on resistance requires a 10V drive voltage on the gate (3rd line down in the table). Less than this drive level and the on-resistance rises as would the heat produced.

At this point I can't decide for you any more but I think you might be looking for a device that will operate from logic levels. In which case the IRFZ44N won't do.

The STB36NF06L is a little higher with the on-resistance but the spec does suggest it will work from a 5V drive on the gate - see electrical characteristics (ON) but i'd still be tempted to find one that is more suitable.

I'd be tempted by this. The PH2520U is a 20V, 100A, 2.7 milli ohm device when the gate voltage is 4.5V. If your logic levels are 3V3 check figure 9 to see it will work well at 3V3.

One last thought about things - you are wanting to PWM a load and if the frequency is high you'll find that the gate capacitance takes some drive current into the gate to get it moving up and down quickly. Sometimes it better to trade off on-resistance to find a device with lower Vgs capacitance. You're into horse-trading now. Keep as low as you can on switching frequency and it should drive ok from a 5V logic pin.

mainly as i have read 5050 RGB LED strip power consumption is betwwen 1.2-2 Amps and that should be divided into 3 as we have three channels to control the PWM.

50 meter X 2A = 100 A for three channels in continuance Darin
Amps per channel 100 / 3 = 33.3 A

one of the parameters i start understanding is rDS(ON) and that about the resistance and the power desiption

if the above calculation is correct

1-what is puffer that should i keep on the maximum A of the MOSFET 20% 50%?
2-what is the Thermal impact - how much MOSFET will be hot ?
3- how i can minimize the temperature generated from the MOSFET as no heat sink planed ?
4- how to calculate the temperature that will generated based on the below parameters?

-LED Strip type is 5050 RGB common anode 12V
-i will use RFP30N06LEMOSFET
- Continuance drain current per channel is 33.3 A

i know too much questions usually from a student how want to learn (btw i am accountant and finished my collage almost 12 years programming and electronics hobbyist )

- Wed Mar 23, 2016 9:34 am #43881 You're not going to like this but from RFP30N06LE datasheet
1) Absolute Maximum continuous drain current 30A (not 33A)
2) Even with 5V on Gate, at 30A, RdsOn=0.047R
SO using nothing more than Ohms Law, Vds =30A x 0.0.47R = 1.41V
Power=VxI = 1.41x30 = 42.3Watts = smoke + molten mess

With a 1'C/W heatsink you might get away with it but with a 3.3V Gate drive even this huge chunk of Aluminium won't be enough.

Also, what PSU, cable and connectors are you planning on using