Can this BSS1338 SMD MOSFET be used in the same way the RFP30N06LE is used in this power control kit SparkFun MOSFET Power Control Kit - COM-10256 - SparkFun Electronics
Keep the total dissipation below the Fet limits and you should have no trouble... The only real concern is the Vgth of the mosfet... which will usually allow 100+ mA @ 3Vgs which would be my worst case assumption. Based on the "Assumption" that you are using a "regular" Arduino and not one of the 3V3 variants... Then I would be REAL concerned that the Vgth be less than 3 volts... 2V - 2V5 being in the OK range.
Doc
That link is to a MOSFET that is nothing like the RFP30N06LE - it has a max drain current of 0.2A rather than 30A, an on resistance of 1.4ohms not 0.047 ohms, it requires 10V of gate drive to turn on properly (not 4.5V).
Having said that they function the same way except the surface mount one is underpowered for that load (the max drain current is not a value you go near in practice for continuous use - its usually the point at which the device is running as hot as it can with maximum heatsinking. For switching 0.15A go for a surface mount FET with about 0.5 ohm or less on resistance ideally - or at least a drain current rating of 0.5A
If the 12V light is a tungsten filament bulb it will draw about 1.5A at turn on if its 0.15A at temperature (the resistivity of all metals increases with temperature). If an LED it will just draw 0.15A.
You want a logic-level MOSFET anyway (Rds(on) quoted for Vgs=4.5V, not Vg=10V)
MarkT:
That link is to a MOSFET that is nothing like the RFP30N06LE - it has a max drain current of 0.2A rather than 30A, an on resistance of 1.4ohms not 0.047 ohms, it requires 10V of gate drive to turn on properly (not 4.5V).Having said that they function the same way except the surface mount one is underpowered for that load (the max drain current is not a value you go near in practice for continuous use - its usually the point at which the device is running as hot as it can with maximum heatsinking. For switching 0.15A go for a surface mount FET with about 0.5 ohm or less on resistance ideally - or at least a drain current rating of 0.5A
If the 12V light is a tungsten filament bulb it will draw about 1.5A at turn on if its 0.15A at temperature (the resistivity of all metals increases with temperature). If an LED it will just draw 0.15A.
You want a logic-level MOSFET anyway (Rds(on) quoted for Vgs=4.5V, not Vg=10V)
Thanks! I'll reread that a few times to try to digest it, but I think that tells me a little more about what some of those numbers are and what they mean. I'll post again when I find something that looks like it might be closer to what I'm needing.
Okay, looking around a little more, I see see sparkfun sells this darlington array made to switch loads, LEDs, relays @ 500 mA per channel. It's the ULN2803A
...so this soic chip looks to do the same. One chip will do all three of my lights (they'll be LEDs, not tungsten, but that info about tungsten lamps is something I didn't know).
It's the ULN2003A
Did you read the data sheet AT ALL?. The BSS138... Data sheet enclosed. Btw did you read what the OP wanted to do with the FET?... Something about a 150 Ma Light???
That link is to a MOSFET that is nothing like the RFP30N06LE - it has a max drain current of 0.2A rather than 30A, an on resistance of 1.4ohms not 0.047 ohms, it requires 10V of gate drive to turn on properly (not 4.5V).
Not quite it is a LOGIC LEVEL MOSFET...
The OP wasn't looking for a 30A SOT23 package... just a little bitty Mosfet for a pilot light...
I'm wanting to make a SMD version of this. I need to turn a 12 volt light on/off that draws about 150 mA.
(OP's FIRST POST... it's the caption... Under that BIG picture of the Mosfet.
This device would WORK PERFECTLY FROM A 3V3 Logic output... The first requirement for an "Expert..." is the ability to read a data sheet... Before you quote a worthless opinion. IMHSHO
Doc
BSS138.pdf (73.4 KB)
The datasheet listing for Rds(on) does not mention any other voltage than Vgs=10V and the phrase logic-level isn't present.
Delving further it turns out that there are some Rds(on) curves for other Vgs values in the graphs - but the graph section of a datasheet are usually "typical" not "guaranteed" behaviour and should always be treated with a grain of salt. This is the first time I've seen an important fact like a device being logic level buried deep in the datasheet like this - and the performance for the lower Vgs voltage is much worse than at 10V (certainly if you wanted to put 0.2A through it at for full temperature range).
But anyway the device isn't suitable for the intended use on power-dissipation grounds alone - given the abs max power dissipation is 0.2W and the current load 0.15A, any value of Rds(on) greater than about 1 ohm is bad news. Given the guaranteed worst-case Vds(on) is 3.5 ohms, that would mean 0.52W dissipation (even the typical value gives 0.21W which exceeds abs max.)
Also given the spread in Rds(on) at 10V (worst case = 2.5 x typical is a massive variation), the spread of threshold voltages (3:1 ratio), then you have to take those graphs at 2.75V and 4.5V with more than a grain of salt I think - I don't believe the manufacturer is actually claiming it guaranteed to work at 3.3V, just that some specimens are likely to most of the time. Otherwise they would put an entry in the table for Rds(on). Still it does look like it should be upto 5V.
What about the ULN2803A darlington array I mentioned in my last post here, it just be a good choice, won't it?
I Quote from the Fairchild sheet I enclosed...
N-Channel Logic Level Enhancement Mode Field Effect Transistor
However had you read the next 3 pages and paid attention to Vgth you would have noted that .8V was enough to cause a 1mA current to flow, and had you looked at the graph on page 3 top left side of the page you would have seen that 2.5V Vg @ 1V Vds was enough to cause 400 500 mA to flow... Clearly a good part for a 150 mA load... from 3V3 logic too.
The mistake iis easy enough to make If your electronics knowledge is spotty... The rest of the data sheet is quite clear however. That's why there are characteristic curves for the device on the other pages.
I was Very surprised that you chose to ignore the "Logic Level" Statement at the top of the page. And Yes certainly a ULN2003 will work, with 6 unused sections and some other drawbacks as well... Vcesat
is about 1.5V Vs 40 mV for the Mosfet... @ 500 mA for the '2003 and 150 mA for the BSS138 so they aren't quite in the same class... The '2003 was originally developed as a Print head driver for the old "Dot Matrix" printers. Many years ago... Thus only 7 drivers, the eighth pin was part of the character spacing. and not used except in a few of the real "Letter Quality" printers Panasonic KXP1135 is one and I still have one stored in my garage that is nearly new... One of the very few besides the Okidata model 82 that could do 7 part NCR forms.
Like C and C++ reading a data sheet requires that you understand what you read... at least a little.
If I was too hard on anyone I apologize. I read the data sheet carefully before I said it would work fine. The comments about the other Fet and the subsequent recommendations for "Driver IC's" were really out there
Much like using a 10 kiloton bomb just to kill the flys in the back yard, Workable but hardly desirable.
I have been doing Engineering for about 30 years total and I WILL NEVER make a recommendation that won't work... or for that matter one I haven't already done.
I would be really ashamed of myself for telling someone something that was clearly wrong.
If I ever say I don't know... it's because I couldn't get the information or figure out how to solve the question at hand...
Some questions I don't even bother with... because there are too many "Experts" involved already.
If there are any other questions... pm me, i Will answer as quickly as I see the email notification. I hope this helps... a little
Doc
ULN2003.pdf (331 KB)
Doc,
Ok, after reading what you said here, I'm starting to thinking the BSS1338 would work for me, (and I have them already). So, after reading a few wikis about MOSFETS, transistors, gates, sources, and drains, I've determined I will never fully understand this stuff!
Here's what my project is sizing up to be like: an ATtiny85 (programmed from an UNO). One sensor on a pin, and three output pins. Each output will. Have (3) Super bright LEDs like this:
They draw a max current of 20mA, so I'll probably run them around 15mA. So that's 45mA I need to switch. The datasheet for the ATtiny85 says 40 mA per I/O pin (and that's max, not typical), so I better use something to switch them.
You said the BSS1338 would work perfect for 3.3 volts... I could have my project work off a 3.3 volt SMD LM1117, since my maxbotix sensor will work at that voltage. So, do I connect the ATtiny85's output pins directly to the BSS1338? --like this?
Gate: ATTiny output pin
Source: Gnd
Drain: Cathode of the LEDs
...so I don't need any resistors between the output pins and the BS1338?
Thanks again.
