Go Down

Topic: 4A current application (Read 1 time) previous topic - next topic



I apologize if I am posting this in the wrong thread.

I am working on a large scale display model that involves controlling a number of Shape Memory Alloy (SMA) wires. SMAs are a type of alloy that, when heated, revert back to a "remembered" shape. They are kind of like "programmable materials."

An easy way to actuate the SMAs is resistive heating. Running 6V DC across a 2 foot wire does the trick pretty well. The problem is that, due to the low resistance (around 1-2 ohms) of the SMA wires, this draws a lot of current from the power supply (4A is not uncommon).

When working with these wires in the past, I used a high power Solid State Relay to do the switching from the Arduino. While this was economical and convenient for a single wire, the application I am designing now involves perhaps 15 wires that need independent control from the Arduino. I would rather avoid the cost and space required of 15 Solid Sate Relays.

In order to do this, I have two questions:

1) I think that using a series of these MOSFETs: https://www.sparkfun.com/products/10213 would allow me to do the switching pretty easily. However, I have never heard of a MOSFET that will turn fully on at 5V such as these, so I want to make sure I'm not confused. Am I correct in thinking that I could control the wires from the Arduino with these MOSFETs?

2) If I were to use something like these MOSFETs, how would I go about safely implementing them with the high currents? I don't need a custom PCB or anything, I just need to be able to mount the parts. Although less desirable, a breadboard would even be ok. I am just concerned that using a normal breadboard, PCB or perf board would result in melted traces. Do I need to be concerned at this level and/or are there boards that I can use safely with 4A?

I would say my knowledge of the Arduino and electronics is just above the novice level, so I apologize if my questions are less than brilliant.



1) the crucial phrase is "logic level" - and an on resistance(*) quoted for a Vgs of 4.5V (or 5.0V).  Don't be misled by the threshold voltage.

2) Ones with a TO220 case: solder via stripboard to suitably heavy wires, if necessary bolt to heatsink with TO220 heatsink mounting washer (conducts heat but not electricity - this is vital if shared one heatsink).

Try and find better MOSFETs than those, they are poor performers by modern standards - aim for an Rds(on) of less than 0.02ohms and you wont need heatsinks for 4A.  Choose lower voltage parts such as 20 or 30V, and the on resistances will be lower than say a 100V part, in general.

(*) Rds(on)
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]


Thanks for the reply!

That sounds good about soldering the wires to the stripboard.

Thanks for the advice about the MOSFETs as well.

I would like to check my understanding of something:

Looking at the datasheet for the MOSFETs I linked to previously (http://www.sparkfun.com/datasheets/Components/General/RFP30N06LE.pdf it looks to me that at Vgs = 5.0 V, the drain to source on-resistance is 0.047 ohms.

At 4A, the power dissipated would be I^2*R = 16 * 0.047 = 0.752 W. This is far below the 96 W maximum. So although the on-resistance may not be as low as more efficient MOSFETs, would it still be ok to use these?

I ask more out of a desire to understand rather than an insistence on using these particular MOSFETs. I don't really care which ones I use, it just seemed that these had a lower on-resistance at Vgs = 5V than normal MOSFETs.

Thanks again!


At 0.7W you'll need some heatsinking really - otherwise they will run rather hot. Its just easier to keep things simple if possible.  Small clip-on heatsinks (or a little fan) would be enough at 0.7W I think.   However if the load is only intermittent for short periods maybe the heatsinking isn't necessary - personally I'd pick a 0.01ohm device and not even think about thermal issues.

To run at the 96W maximum you would need to mount on a water-cooled copper slab - no one goes anywhere need that limit if they don't have to!
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]


Thanks for the reply!

Wow I am surprised to hear that! I always thought that at the maximum power dissipation level the MOSFET would just get kind of warm or something, I didn't realize what an extreme case that was. :)

The load will likely be intermittent for a number of the wires, but really I would just rather not worry about anything overheating.

It seems to me that a lot of MOSFETs have a Vgs of 20V. How do people control these with the Arduino? I have been trying to to Google this, but most of the tutorials seem to just apply 5V and not worry about the on-resistance.

Thanks again!


Wattage is heat (at least the power dissipated by a semiconductor is heat)...    The temperature depends on how concentrated the heat is.  You can melt solder with a 25W soldering iron!  A heatsink spreads-out the heat and lowers the temperature.

A 100W incandescant light bulb puts-out about 10W of light and 90W of heat.

100 Watts is not enough to heat-up a room, but it's enough to make a light bulb too hot to touch!   Actually, when the light is absorbed by an object, the light is converted to heat also.   (I don't think you can melt solder with a light bulb because of the larger surface area...  I'll have to try that! :D  But, the filiment inside the light bulb is certainly hot enough to melt solder.)


Thanks for the reply!

That is very interesting, and makes sense. The lightbulb example definitely helps me understand why 96W for a small MOSFET would get very hot.


Practically speaking, TO220 packages start to get fairly warm with 2-Watts or so dissipation, and no heatsink. 

Also, in regards thresholds/etc, the most important thing is to look at the Id vs Vds curves for the device. This is figure 1 in the d/s for the FQP30No6L part from Sparkfun.

What you can see there is that the device is virtually turned-on as well for Vgs=5V as for larger Vgs values. For a non logic-level device, the Vgs=5V curve will be down much lower in comparison [similar to the Vgs=3V curve here].


That makes sense, thanks for the reply!


According to the data sheet for the smaller one "RFP30N06LE", it is fully enhanced at 3V G-S... 20A and a 1.5V VDS... It's on page 4, top right curve... AND it is CLEARLY Marked on Page One as a "Logic Level" Mosfet. A 10 A load and a .047 ohm Rdson will dissipate 5 watts at full load so..   a heat sink is in order. A 10 cm piece of 1.5 mm aluminium for each Mosfet bent into a u shape with a flat bottom where you will drill a 3 mm or better hole for the mounting hole... I'm not sure of where you are and I am totally unfamiliar with European hardware. Were it me I'd use a 1/4" 4/40 screw and nut, 3 mm or about .12" is fine and it should stay well below 100 deg C at full load (10A). Your stated requirements are somewhat less. The advice about lower Rdson is valid but unnecessary IMO for you... Unless you find a source of cheaper TO-220 Mosfets somewhere else. there are many sources of inexpensive Mosfets however the shipping time must be accounted for. I order parts frequently, sometimes daily even though I am retired but for projects that are month's ahead of actual use... or too insanely cheap to pass up... <BFG>

--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

Go Up