Go Down

Topic: Great deal on logic mosfets... (Read 1 time) previous topic - next topic

focalist

Jan 24, 2013, 11:10 am Last Edit: Jan 24, 2013, 06:44 pm by focalist Reason: 1
I am not connected to these guys in any way!

However, was cruising ebay and came across these logic-level n-channel mosfets:

http://www.ebay.com/itm/300848355748?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

10 pcs  60v 12A n-channel power mosfet, Logic Level gate (I use the fairchild version for PWM for power LED's)
$1.00, free shipping, stock is in USA.  NOS looks like, for the price, who cares!

I got two sets (twenty mosfets) for two bucks, even I can afford to sock a few of these really useful power transistors away.

This is for the broke students and hobbyists.  No fair poaching for some company, they only have a small stock!

I am not promoting them in any way other than to say I think it's a great deal... Nuff said..
When the testing is complete there will be... cake.

focalist

Hmm.

Vendor bumped the price up, I bet the free shipping part wasn't thought through.. That's probably the price increase.  Still not a bad deal, imo... Forty cents each is still cheap.
When the testing is complete there will be... cake.

dhenry

Quote
12A


I would suggest that you view that spec with a huge grain of salt. For mosfets, you really need to look into SOA, thermal resistance, and derate aggressively.

For a dpak like this, you would be lucky if you sustainably dissipate 1w.

vasquo

For DPAK, D2PAKs, don't forget to use the appropriate sized area heatsink on your PCB, and not just use the minimum pad size.  (also helps to use the opposite side of the PCB, and use plenty of vias to connect each other)... and use of 2oz copper (or thicker)

retrolefty

Yes up to $2 for 10 now. I went ahead and bought a pack of 10. The Ron spec isn't great, so yea the 12 amp rating is probably a little 'optimistic' at best, but for handling up to an amp or two with no heatsinking, this 20 cent device should work great.

Thanks for the tip.

Lefty

focalist

No problem, I see this kind of stuff now and then.  I think the ones I bought before were RFD3055LE, those rate for 60v and 11A.

Since most LED PWM applications are a couple of amps at most, well below 60v, they work really well for that sort of thing.  Switching two amps at 12v, they don't even get warm, even with no heatsink.  The spec on these is similar enough that I expect them to behave in a similar fashion.
When the testing is complete there will be... cake.

focalist

They arrived today, I swapped in one on a circuit I was using the fairchild version with, performed beautifully.

In all honesty, if I thought there would be a heat problem, at this cost per unit it would probably just be simpler to split the load onto two or three mosfets rather than mess around with heat sinking those pads.

One stepper motor circuit I did used ULN2003 darlington arrays.  When the chips started getting hot, I did the super simple fixit... Stacked a second ULN2003 right on top, soldered the leads together, splitting the load.  Ran much cooler but I would not recommend this as a practice.. Lol
When the testing is complete there will be... cake.

dhenry

It is not unusual to stack chips, even in production products, particularly when they can be selected and output to a bus.

Memory chips are fairly frequently stacked, so are spi / i2c chips.

retrolefty


They arrived today, I swapped in one on a circuit I was using the fairchild version with, performed beautifully.

In all honesty, if I thought there would be a heat problem, at this cost per unit it would probably just be simpler to split the load onto two or three mosfets rather than mess around with heat sinking those pads.

One stepper motor circuit I did used ULN2003 darlington arrays.  When the chips started getting hot, I did the super simple fixit... Stacked a second ULN2003 right on top, soldered the leads together, splitting the load.  Ran much cooler but I would not recommend this as a practice.. Lol


Mine arrive today also, but I'm only about 60 miles north of San Jose. They are nice looking and I suspect will be very useful for switching loads larger then small TO-3 package transistors can handle, but not requiring a full size TO-220 higher wattage dissipation package.

Lefty

focalist

#9
Jan 27, 2013, 06:05 am Last Edit: Jan 27, 2013, 06:14 am by focalist Reason: 1
They have been performing great (the old ones, and assumedly these will the same) with my strobe project, and they also have been working great as PWM to 5 meter RGB strip light, 2A per color at 12v.  They stay pretty much ambient, I use a high value resistor gate to drain for discharge, and around 100 ohms on the gate to a logic pin.  I have used them with no gate resistor before, but been told it unduly loads the pin momentarily.

Considering that I had been intimidated by the scary symbol and only switched (ugh) to MOSFET for switching apps from darlingtons like TIP120 quite recently, the heat problems are much smaller with these!

I am about to take a foray into trying a MOSFET current regulator over a LM317 configuration, doing a bit of reading on current sources in general and better ways to drive the power LED's I love so much :)  If I torch a few transistors learning something new, it's a cheap lesson...

Gonna keep the vendor as one to check their inventory once in a while, cheap, good sources are always a great find..
When the testing is complete there will be... cake.

Go Up