fake MOSFETs on Ebay?

Discussion in 'General Electronics Chat' started by strantor, Sep 12, 2011.

  1. tom66

    Senior Member

    May 9, 2009
    What I'm trying to say is $1.75 a MOSFET would be a steal, if you could get it. Saw a nice 30mOhm 400V MOSFET. How much each? $28!

    You'll need a high voltage driver, like IR2110, which has a built-in 500V driver.
  2. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    DOH! I get you now, yeah, 1.75 would be a steal, but remains to be see if He'll sell it to me. My offer was pretty low-ball.

    Ok, looking at this driver, the things I don't get:
    1. In the literature I've read, I have seen more than once that the gate of the mosfet needs a voltage higher than the voltage on the drain in order to turn on. Yet, I continually see 200 & 300V mosfets that specify a gate voltage of 10 or 12 volts. for example, The mosfet in post#1, Vdss 300V, Vgs 20V. So, would I need a driver that goes above 300V or not? why?
    2. Driver states it's a high & low side driver. I'm Not clear on what high & low side means. I assume That I am looking for a low side driver since the mosfets are N-type and I am switching ground with them. whatever the case, it seems this driver would work.
    3. How do I determine how many mosfets I could parallel with this driver? I know it has to do with Ciss of my mosfet, which in this case is 6300pF, and it also has to do with how fast I want to switch them, which in this case would be 20khz, and it has to do with the amp output of the driver, which is 2A the math escapes me. I have seen the formula (think) but was confounded by it, and then distracted by something else and never got back to it.
    4. how do I connect it to my μC? I assume the Lin would go to my digital pin, and I would send PWM here. Vss would connect to my uC ground, and simple as that. (plus 5-20V on Vdd)
  3. SgtWookie


    Jul 17, 2007
    Unless you are buying from the factory or a factory-authorized distributor, you may well be buying counterfeit parts, which simply won't perform up to the specs in the datasheet. "You pays yo' money and takes yo' chances". If you want to be certain you're getting the genuine item, check the manufacturer's authorized distributor list and buy only from that list.

    You're getting Vdss and Vgs confused.
    Vdss is the maximum voltage that can be applied across the drain and source terminal before the junction breaks down.

    Vgs is the voltage on the gate, referenced to the source terminal (ie: using the source terminal as the "ground" or 0v point). Most typical MOSFETs have a maximum Vgs of ±20v, but there are a number that have a much lower limit. To avoid smoke, verify the limits in the datasheet for the particular MOSFET in question.

    An IR2110 is a half-H bridge driver.
    If you are only switching the low side of a load (between the load and ground) then you only need a low-side driver.
    Read this application note:
    It covers a wide variety of techniques.
    Well, it's more complicated with an IR2110; the Hin and Lin need to go nearly to the IR2110's Vdd, which has to be somewhere above 10v (I think 12v is minimum) or the driver will go into shutdown; so you'd need a driver for the driver....

    Anyway, read that application note for starters.
    Then, if you're just going to be switching the low-side MOSFETs, there are quite a few other low-side drivers that would be easier to interface. Some have quite high current outputs (>6A). Then you also have things like trace inductances, etc. to worry about, but try to not take too much in at once or you'll get confused.
    strantor likes this.
  4. tom66

    Senior Member

    May 9, 2009
    Ah, I didn't know whether he'd be using a low-side MOSFET or a high-side MOSFET.

    A problem with placing MOSFETs in parallel is that the gate charge increases. The total on-resistance of the array falls. This causes lower static power losses (static = on or off, not switching between states) but higher switching losses.

    A good driver is http://www.microchip.com/wwwproducts/Devices.aspx?dDocName=en553187

    Heed my warning about using a single MOSFET for drive. (Or even multiple parallel MOSFETs.) You do not want a shorted MOSFET causing the motor to run out of control.

    How will you implement reverse, if necessary?
    strantor likes this.
  5. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    thanks, I think I will; digikey
    yes i was
    mine is +/-20V
    skimmed over all of it, read most of it in detail, understood some of it.
    Here's what caught my attention, on page 9:
    so after a bit of searching I found this:http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=FAN3121TMXTR-ND

    from the datasheet:
    Does that look like a good driver? Do you think it could switch a few in parallel easily? or should I get 5 or 6 drivers?

    It has 2 types available, one for CMOS, and one for TTL; not sure which one is better - I will probably be driving this with an arduino to start.

    also, it is available in inverting and noninverting; I assume that inverting would turn the mosfet OFF if the input is high, and noninverting would turn the mosfet ON if input is high?

    You got that right. The past few days have been very challenging. I'm a hands-on kind of guy, and all this reading bland writeups is not in my nature. I actually last night realized that I can't remember everything I'm reading and that my mind so "all over the place" that I haven't even really learned and understood anything since I started this quest, so I got a notebook and started writing down all my questions with notes and tying to answer them systematically one at a time. I'm a very scatterbrained person and until today, I've just been running around in circles.

    Yes I got a big 400A fuse to use; I would rather have a contactor as well... will see.
    I just started learning about drivers so pardon my ignorance; Why did you recommend the MCP14E7? Is it better than the one I linked to? All I know to look for is the Amp capacity and then the miller thing

    implementing reverse by swapping field wires probably